QCW questions

[Rafft]

 hi

I have a bunch of 360uF 470uF 560uF 400v elcaps. if paralleled, will total almost 3000uF. can I use this for the DC BUS?

from the few coils I have seen on the net, all use single huge capacitor.

whats the theory behind that?

[davekni]

Yes, paralleled smaller caps works fine.  That's what my DRSSTC uses.  In my case, 96 parallel 470uF caps to handle high power pulses.

[Rafft]

thanks David

alright then, will parallel more

currently, this one is on 1800uF/400v using 560uF/370uF. charged by dcdc boost . 4S LiIon batt. bellow is at 250vdc. basically a battery-powered qcw. 

https://youtube.com/shorts/kzEzVNB3zBA?feature=share

[Rafft]

im in DR mode for phase lead adjustment. here is a scope shot of halfbridge output(yellow) and Iprim(blue). its the BEST ADJUSTMENT I could do (56uH and around 50 - 65R from 500R pot). im using IRG4PF50WD. according to calculation, its rated for max 250KHz.

Iprim has 1.7uS period(588KHz). too high. ive measured my primary inductance(LC meter) and no secondary to be around 8uH. resonant cap is 13.1nF. this should be around 493KHz. my secondary w topload(according to javaTC) is 420KHz.

so
secondary 420KHz
primary 493KHz
BUT this coil is stubborn and loves to oscillate at around 555-588KHz. I have tried messing with some caps for resonance(11nF and 15nF) but I gett smaller sparks.

IGBT is 'low frequency' but still operates

running out of ideas



and is this normal? looks like my coil is not DR'ing? looks like my Iprim is not 'amplitude modulating'?


more ON-time


info on secondary:
3.2" diameter x 4.85" Height
awg #33 684T(javatc)
609KHz javatc and actual Fres measurement seem close

topload (back to using 1)
1.3" x 6.4"

sec & topload = 420KHz (javatc & measurement)

I dont see anything else why this coil loves to oscillate around 588KHz, except like its resonating FROM secondary w/o topload(610KHz) 

[davekni]

and is this normal? looks like my coil is not DR'ing? looks like my Iprim is not 'amplitude modulating'?

QCW coils generally have enough arc loading to lower secondary Q so much that little energy is left to transfer back to the primary.  That's my guess as to what you are seeing, but I am not certain.  Hopefully members with direct QCW experience can contribute more information here.

I dont see anything else why this coil loves to oscillate around 588KHz, except like its resonating FROM secondary w/o topload(610KHz) 

Primary and secondary are coupled resonators.  There is an upper pole frequency and a lower pole frequency.  You appear to be running at the upper frequency (where most QCW coils run).

im in DR mode for phase lead adjustment. here is a scope shot of halfbridge output(yellow) and Iprim(blue). its the BEST ADJUSTMENT I could do (56uH and around 50 - 65R from 500R pot). im using IRG4PF50WD. according to calculation, its rated for max 250KHz.

Presuming typical UD2.X driver circuitry (at least phase-lead input part), maximum theoretical phase lead (with infinite inductance) is 90 degrees.  Practical upper limit is ~60 degrees.  At 588kHz, 60 degrees is only 284ns lead.  Driver + IGBT delay is likely more than 284ns.  To get phase lead at that frequency will require significantly faster driver and/or H-bridge, or different phase-lead circuitry such as a PLL.  (One somewhat-simple option is to reverse CT leads and add phase lag instead of lead.  In other words, start with 180 degrees and add lag such that driver/IGBT time and lag is not quite another 180 degrees.  Downside of this option is that phase lead increases as frequency decreases.)

Impressive that you have those IGBTs running at 588kHz and making sparks.

[Uspring]

I believe the primary and secondary resonance frequencies are too far apart to get visible 'amplitude modulating'. This amplitude modulation is caused by energy sloshing back and forth between the 2 tanks. A large difference in res frequencies impedes that.
Once you have breakout, the arc will add capacitance to your secondary tank, which will lower secondary fres and even increase the difference.

[Mads Barnkob]

Impressive that you have those IGBTs running at 588kHz and making sparks.

Do not underestimate the resonant switching capabilities of IGBTs. You are instantly handed like 4 times the speed and switching peak current (extremely rough estimate)

I believe the primary and secondary resonance frequencies are too far apart to get visible 'amplitude modulating'. This amplitude modulation is caused by energy sloshing back and forth between the 2 tanks. A large difference in res frequencies impedes that.
Once you have breakout, the arc will add capacitance to your secondary tank, which will lower secondary fres and even increase the difference.

This is why Steve Wards added a series of small capacitors inside his secondary coil. His first QCW versions simply needed more "self capacitance" to withstand detuning from the long sparks.

[Rafft]

hi guys

thanks for the input.

just had the "idea" to RUN the secondary to a MUCH lower Resonant Freq. like 200KHz, If I could achieve it. of course I will adjust the PRI res to be lower as well, 10% closer to sec res. I will try to let it run lower so it can be switched by the IGBTs limit.


currently


will TRY this out (350Khz JTC)

[Rafft]

I believe the primary and secondary resonance frequencies are too far apart to get visible 'amplitude modulating'. This amplitude modulation is caused by energy sloshing back and forth between the 2 tanks. A large difference in res frequencies impedes that.
Once you have breakout, the arc will add capacitance to your secondary tank, which will lower secondary fres and even increase the difference.

Thanks Uspring

copy that. will try to keep it closer next time. this is my first proper DR build, please forgive the n00b 

[davekni]

just had the "idea" to RUN the secondary to a MUCH lower Resonant Freq. like 200KHz, If I could achieve it. of course I will adjust the PRI res to be lower as well, 10% closer to sec res. I will try to let it run lower so it can be switched by the IGBTs limit.

will TRY this out (350Khz JTC)

Yes, lowering to 200kHz would be difficult using top load changes only.  Halving frequency requires 4x capacitance.
Standard UD2.7 drivers usually settle on the lower pole, so actual frequency may be a bit under 350kHz.  You may be able to squeeze enough phase lead at 350kHz or a bit under.  Optimizing IGBT gate resistors to get barely-enough dead time can help.  To get ideal switching, the final turn-on transition (after dead-time delay) should be roughly at zero-current.

Do not underestimate the resonant switching capabilities of IGBTs. You are instantly handed like 4 times the speed and switching peak current (extremely rough estimate)

Yes, these IGBTs can switch at 588kHz as you have shown.  Proper phase lead is difficult (requires PLL or ...), and switching losses will still be on the high side at that frequency.  Turn-off current would be significant in order to achieve turn-on at zero-current time.

Good luck with your 350kHz version!

[Rafft]

guys

did some minor adjustments on my coil

- refit variable inductor (12-50uH) to UD2.x because trimpot(for phaselead) OPENed! good thing there is the 5819 on feedback input. smelled something burnt. good thing comparator did not get damaged.
- reconfigured MMC 22nF/1000VDC MKP10 from 13nF to 27nF (4S4P). lets go for less L more C 

my setup. only place where I could connect EARTH GND. its only in DR mode. just wanna check sparks and phase lead. small dc fan is THERE to blow air into the 5V linear regulator(need to attach small heatsink), not the IGBTs 


checking long ON-time


and the good(?) thing is its oscillating around 312KHz and I havent touched anything on the coil, except for the MMC. where do you think this is going? south pole? Im really getting more confused  :'( or maybe its time to wind another secondary(?) maybe with Fres 200KHz w/o topload(?)


I do get better waveforms than before, and the variable inductor IS "finally" working(because Fres is lower , I  ASSUME).


--------------------------------------
primary coil TAP measured with LC meter with secondary mounted(unGNDed) = 14.3uH
resonant cap of 27nF & 14.3 primary = 258KHz (Resonance calculator)
Actual Resonating Freq = 312KHz
--------------------------------------

uhhhh why 


cheers
Ralph

[davekni]

- reconfigured MMC 22nF/1000VDC MKP10 from 13nF to 27nF (4S4P). lets go for less L more C 

22nF at 4S4P would be 22nF, not 27nF.  Is it actually 4S5P?  Or are the MMC capacitors actually 27nF/1000VDC?

--------------------------------------
primary coil TAP measured with LC meter with secondary mounted(unGNDed) = 14.3uH
resonant cap of 27nF & 14.3 primary = 258KHz (Resonance calculator)
Actual Resonating Freq = 312KHz
--------------------------------------

uhhhh why 

If MMC is 22nF, calculation gets a bit closer to 312kHz.  Also, a previous post listed your primary inductance measurement as 8uH.  If this is at the bottom end of your LRC meter's capability, that 8uH to 14.3uH change may be just measurement error.  What does JaveTC say your primary inductance is?  That may be more accurate than measurement.

I do get better waveforms than before, and the variable inductor IS "finally" working(because Fres is lower , I  ASSUME).

Does it work to get all the way to bridge switching just before zero current?

and the good(?) thing is its oscillating around 312KHz and I havent touched anything on the coil, except for the MMC. where do you think this is going? south pole? Im really getting more confused  :'( or maybe its time to wind another secondary(?) maybe with Fres 200KHz w/o topload(?)

Are you getting sparks at 312kHz?  That might be reasonable for the lower pole.  Now that I think about it, I recall some LTSpice simulations suggesting that UD2.7-style driver feedback is likely to start at lower pole if primary frequency is below secondary frequency (as is typical for DRSSTC designs), and start at upper pole if primary frequency is higher.  That would fit your experience here.

200kHz may be too low for this small coil.  Lower frequency requires more voltage to get breakout.  May have secondary arcing problems at higher voltage.

[Rafft]

David

thanks for the reply

Is it actually 4S5P?

sorry typo. MMC is 4S5P of 22nF 1000Vdc

Primary coil is 4.4" diameter. I removed it from setup and measured the inductance (single coil alone). the most spark length I could get is at 6th tap. it measured 7.16uH on LC meter. JavaTC shows 7.3uH at 6th tap.

Code: [Select]

5T = 5.36uH
6T = 7.16uH
7T = 9.15uH
8T = 11.15uH
8.5T = 11.2uH
9T = 13.4uH

 6th TAP 14.3uH from LC meter/measured WITH SECONDARY coil / 27nF with resonance calculator = 258KHz
6th TAP 7uH (from LC meter/measured PRI w/o SECONDARY)/27nF with resonance calculator = 362KHz,
 javaTC shows 359KHz
 and the actual resonance is 312KHz.

TEXT

Code: [Select]

J A V A T C version 13.6 - CONSOLIDATED OUTPUT
01/02/2022, 21:13:49

Units = Inches
Ambient Temp = 68ºF

----------------------------------------------------
Secondary Coil Inputs:
----------------------------------------------------
Current Profile = G.PROFILE_LOADED
1.6 = Radius 1
1.6 = Radius 2
1 = Height 1
5.85 = Height 2
683.1 = Turns
33 = Wire Awg

----------------------------------------------------
Primary Coil Inputs:
----------------------------------------------------
Round Primary Conductor
2.2 = Radius 1
2.2 = Radius 2
0.7 = Height 1
1.45 = Height 2
6.25 = Turns
0.06 = Wire Diameter
0 = Ribbon Width
0 = Ribbon Thickness
0.027 = Primary Cap (uF)
0 = Total Lead Length
0 = Lead Diameter

----------------------------------------------------
Secondary Coil Outputs:
----------------------------------------------------
417.21 [kHz] = Secondary Resonant Frequency
90 [deg °] = Angle of Secondary
4.85 [inch] = Length of Winding
140.8 [inch] = Turns Per Unit
0.00002 [inch] = Space Between Turns (edge to edge)
572.3 [ft] = Length of Wire
1.52 [:1] = H/D Aspect Ratio
117.4376 [Ohms] = DC Resistance
51585 [Ohms] = Reactance at Resonance
0.09 [ lbs] = Weight of Wire
19.679 [mH] = Les-Effective Series Inductance
19.753 [mH] = Lee-Equivalent Energy Inductance
19.032 [mH] = Ldc-Low Frequency Inductance
7.395 [pF] = Ces-Effective Shunt Capacitance
7.367 [pF] = Cee-Equivalent Energy Capacitance
13.716 [pF] = Cdc-Low Frequency Capacitance
4.28 [mils] = Skin Depth
5.49 [pF] = Topload Effective Capacitance
262.5383 [Ohms] = Effective AC Resistance
196 [Q] = Quality Factor

----------------------------------------------------
Primary Coil Outputs:
----------------------------------------------------
358.72 [kHz] = Primary Resonant Frequency
14.02 [% high] = Percent Detuned
90 [deg °] = Angle of Primary
7.2 [ft] = Length of Wire
20.74 [mOhms] = DC Resistance
0.06 [inch] = Average spacing between turns (edge to edge)
0.566 [ inch] = Proximity between coils
0 [inch] = Recommended minimum proximity between coils
7.29 [µH] = Ldc-Low Frequency Inductance
0.01996 [µF] = Cap size needed with Primary L (reference)
0 [µH] = Lead Length Inductance
107.959 [µH] = Lm-Mutual Inductance
0.29 [k] = Coupling Coefficient
0.127 [k] = Recommended Coupling Coefficient
3.45 [half cycles] = Number of half cycles for energy transfer at K
4.55 [µs] = Time for total energy transfer

----------------------------------------------------
Top Load Inputs:
----------------------------------------------------
Toroid #1: minor=1.3, major=6.4, height=6.75, topload

JTC

Code: [Select]

units=0,
ambient=0,
s_ws=0,
s_Al=0,
p_ws=1,
p_Al=0,
p_ribbon=0,
temp=68,
g_radius=0,
w_radius=0,
ceil_height=0,
s_radius1=1.6,
s_radius2=1.6,
s_height1=1,
s_height2=5.85,
s_turn=683.1,
s_wd=33,
p_radius1=2.2,
p_radius2=2.2,
p_height1=0.7,
p_height2=1.45,
p_turn=6.25,
p_wd=0.06,
p_vwidth=0,
p_rthick=0,
Cp_uF=0.027,
Lead_Length=0,
Lead_Diameter=0,
desired_k=0,
t.inner=1.3,
t.outer=6.4,
t.height=6.75,
TT=true,
TG=false,
x_Vin=0,
x_Vout=0,
x_Iout=0,
x_Hz=0,
x_Vadjust=0,
x_ballast=0,
rsg_ELS=0,
rsg_ELR=0,
rsg_rpm=0,
rsg_disc_D=0,
rsg_ELR_D=0,
rsg_ELS_D=0,
stat_EL=0,
stat_EL_D=0,
stat_gap=0,
SPE=true,
RGE=false


Does it work to get all the way to bridge switching just before zero current?

it auto 'phase leads' with higher interrupter ON-time. with low on-time, there is no phase leading.

Are you getting sparks at 312kHz?

yes. and the longest, starting from 5th tap to 9th tap.

Now that I think about it, I recall some LTSpice simulations suggesting that UD2.7-style driver feedback is likely to start at lower pole if primary frequency is below secondary frequency (as is typical for DRSSTC designs), and start at upper pole if primary frequency is higher.  That would fit your experience here.

Im not sure what to answer here, but its my best spark output and "cleanest" Iprim I have scoped, from all the testing I have done with 100Vdc. im guessing its around L=20uH'ish and 50R for the phase lead.

[davekni]

6th TAP 14.3uH from LC meter/measured WITH SECONDARY coil / 27nF with resonance calculator = 258KHz
6th TAP 7uH (from LC meter/measured PRI w/o SECONDARY)/27nF with resonance calculator = 362KHz,

What frequency is your LC meter using to measure?  Most LC meters measure at 100kHz or below, well below your secondary resonant frequency.  At low frequency, adding the secondary should have insignificant effect on primary inductance.  (If the secondary is shorted, primary inductance should go down a little.  If readings are precise enough, coupling factor can be measured this way.)  My inclination is to believe the w/o secondary reading, especially since that roughly matches JavaTC result.

it auto 'phase leads' with higher interrupter ON-time. with low on-time, there is no phase leading.

Sounds like phase lead is marginally sufficient.  Long ON-time matters most for IGBT power dissipation since that is where most time is spent.

Im not sure what to answer here, but its my best spark output and "cleanest" Iprim I have scoped, from all the testing I have done with 100Vdc. im guessing its around L=20uH'ish and 50R for the phase lead.

I wasn't looking for an answer.  I was just suggesting that perhaps the frequency reduction you achieved with more MMC capacitance seems reasonable.  Capacitance scaling alone would be only a factor of sqrt(13nF/27nF) = 0.69.  Since your actual frequency reduction was 312kHz/588kHz = 0.53 (If I'm understanding your post details correctly).  The additional frequency reduction is likely due to a change from upper-pole to lower-pole operation.

Overall, it seems like your 312kHz operation is a reasonable place to run.

[Rafft]

David

What frequency is your LC meter using to measure?

according to the circuit, 550KHz. this is the one I have built years ago. yes that JTC has close results with the LC meter readings (w/o secondary)
https://pic-microcontroller.com/very-accurate-lc-meter-based-on-pic16f628a-using-pic-microcontroller/

Sounds like phase lead is marginally sufficient.  Long ON-time matters most for IGBT power dissipation since that is where most time is spent.

does this mean phase leading worked because IGBT speed (250KHz)  got closer to the operating resonant freq? or its just the way UD2.x works?

The additional frequency reduction is likely due to a change from upper-pole to lower-pole operation

Its time for a better tuning then

btw here are a couple scope shots with it in QCW, indoors only. DC ramp(yellow) and Iprim(blue). this is the maximum amplitude setting(clean ramp up/down) before it flats out at the top(much thicker sparks) . ramp time is more or less around 8-10mS. any longer and the sparks go hissy and thick. Im using synchronous buck. switching freq of 20KHz. IRG4PC50FD IGBT for the switches. 

DC BUS its at 130Vdc.


and at 240ish Vdc


just scared sh*t If I crank up the DC bus more (Im too close to the coil)  I even forgot to check phase lead and measure Resonant freq. oh well, will do it next time... plus a short vid outdoors


actually , GOAL ACHIEVED = unTUNED QCW DRSSTC   

[Uspring]

I don't think your coil is untuned. The flattening of primary current while input voltage is rising is an indication of the secondary drawing power from the primary.
This is not really unexpected, since you are now running at the lower pole, i.e. a frequency below secondary resonance frequency. The arc capacitance will decrease the secondary fres and move it closer to the operating frequency. Power transfer to the secondary will increase.
That is different from upper pole operation, which you had previously. In that case the lowering of the secondary fres due to the arc will move secondary fres away from the operating frequency and move the coil out of tune.

[Duane B]

Deleted. I will start a different thread.

[davekni]

    What frequency is your LC meter using to measure?

according to the circuit, 550KHz. this is the one I have built years ago. yes that JTC has close results with the LC meter readings (w/o secondary)
https://pic-microcontroller.com/very-accurate-lc-meter-based-on-pic16f628a-using-pic-microcontroller/

That explains the change in reading with secondary in place.  Inductance reading is no longer valid at resonance.  Ignore that with-secondary reading.

    Sounds like phase lead is marginally sufficient.  Long ON-time matters most for IGBT power dissipation since that is where most time is spent.

does this mean phase leading worked because IGBT speed (250KHz)  got closer to the operating resonant freq? or its just the way UD2.x works?

250kHz is not directly relevant.  It worked because frequency was lower, so ~60 degrees of phase lead represents a longer time period, enough to almost compensate for total driver+IGBT delay.

    The additional frequency reduction is likely due to a change from upper-pole to lower-pole operation

Its time for a better tuning then

No, tuning is fine.  Lower pole is generally preferred for DRSSTC operation.  Both upper and lower pole frequencies exist when primary and secondary frequencies are identical.  The coupled system does not resonate at their center frequency.  Coupling splits the resonance into two frequencies, called the upper and lower poles.

[Rafft]

guys

just a small update: using upper pole again. added 5"dia x 5.8" H cylinder for topload. made another Primary coil also. 3 taps only. 6T(9.13uH) 7T(11.84uH) 8T(14.2uH) Cres = 14.66nF.   <=== measured with my LC meter and primary coil alone

Secondary Resonant freq with 2 topload is 343.6KHz (JTC) ....... freq drop from 420KHz to 343KHz

Primary resonant 8T(343KHz) 7T(385.3KHz) 6T(435KHz) (res calc)

I get biggest spark at 6T(435KHz)

I have also cleaned the wiring , UD is under the coil. buck and boost is outside. RAMP DC connected thru deans(T-plug) connector.


mey boost(4S input) and buck. this is a battery-powered QCW


checking the resonant freq and phase lead . Half-bridge output(yellow) Iprim (blue). 2.3us is 434.78KHz. wow. so close to the primary Fres (435KHz)  . measurement taken at peak of ramp output. resonant freq varies from start of ramp(438.6KHz), ramp peak(434.78KHz), ramp end(431KHz)




ramp out vs Iprim


half-bridge vs Iprim. Iprim still rising, need MORE juice perhaps    I have only 3100uF bulk cap.


video below. started with 130Vdc bus and 250Vdc........

=================================
so in summary:
Secondary Fres = 343.6KHz
Primary Fres = 435KHz
91KHz difference. IS this a BIG or small difference? is this NORMAL ? I mean the freq difference. I was thinking maybe a 10-20KHz would do, but I guess it needs bigger difference cuz its higher freq compared to DRs in the 50-150KHz (?)

though Im still not sure what the 'best' freq is YET. I aim to use the 8Th tap(or more) and re-adjust the MMC. high-impedance(more L less C) for primary.  still need to save up for BIG elcaps . and possibly make this into full-bridge driver.

cheers
Ralph

[Duane B]

so in summary:
Secondary Fres = 343.6KHz
Primary Fres = 435KHz
91KHz difference. IS this a BIG or small difference? is this NORMAL ? I mean the freq difference. I was thinking maybe a 10-20KHz would do, but I guess it needs bigger difference cuz its higher freq compared to DRs in the 50-150KHz (?)

Are you measuring these frequencies or just using javatc and calculations? I think your primary and secondary are too far off frequency from each other. It does not look like you have reached critical coupling yet between primary and secondary. Your waveforms look more like a single resonant system to me. Scan the primary resonant circuit and see if you get two resonant frequency peaks.

[davekni]

Secondary Resonant freq with 2 topload is 343.6KHz (JTC) ....... freq drop from 420KHz to 343KHz

What material is your "GND BASE REFERENCE" plate made of?  (The plate at the bottom of your coil.)  If metal (conductive), that will reduce both primary and secondary inductances as it blocks AC magnetic field.  Also, does anyone know if JavaTC includes such inductance reduction for the secondary coil due to top-load proximity?  Does your primary winding extend all the way from bottom to top?  Or, are there gap(s)?

[Rafft]

I have re-measured actual resonant frequencies of PRI and SEC

Secondary with toroid & cylinder topload, coil/topload removed from base
======================================
based on JavaTC = 353.3KHz Secondary and topload/s only
based on JavaTC = 343.6KHz sec/toploads/pri/mmc whole system
"series resonance test" signal gen / 10k res = 347KHz
"series resonance test" signal gen / 10k res =  289KHz w/simulated streamers , 40" wire upward

Primary with MMC(14.5nF) in parallel , Secondary coil removed
=======================================
6T = 410KHz most spark length I get
7T = 382KHz
8T = 340KHz

Primary with MMC(14.5nF) in parallel , Secondary coil INSTALLED
=======================================
6T = 410KHz
6T = 406KHz with simulated streamer

6T = 456KHz Secondary Coil GNDed, no streamers
6T = 447KHz Secondary Coil GNDed, WITH simulated streamers

What material is your "GND BASE REFERENCE" plate made of?

the white plate is plastic(from printer paper tray). rest of the thing/s holding the assembly, is wood. only large metal are the long bolts/nuts & aluminum tube(bolt support). there are no metal screen or plate to protect the electronics underneath(yet).

Does your primary winding extend all the way from bottom to top?  Or, are there gap(s)?

its elevated half an inch from "gnd base ref" . it has small gaps inbetween, zip ties near the 6T &T taps.

Your waveforms look more like a single resonant system to me. Scan the primary resonant circuit and see if you get two resonant frequency peaks.

ok, I assume youve built QCW before (?) . I have no idea how they would look like . 'these' waveforms are usually what I see online but sure, I will recheck it later 

[davekni]

Great data!  Thank you.

6T = 456KHz Secondary Coil GNDed, no streamers
6T = 447KHz Secondary Coil GNDed, WITH simulated streamers

Does this refer to bottom of secondary grounded (normal operation condition), or top also grounded (secondary shorted)?  I'm guessing the former.  If you took a measurement with shorted secondary, that would allow calculating coupling factor between primary and secondary.

Numbers look reasonable.  Here's a quick AC LTSpice simulation with your coil parameters.  I used slightly higher primary inductance based on 410kHz and 14.66nF.  Coupling factor is a guess at 0.32.  What does JavaTC say for coupling with your latest geometry?  Red is without arc.  Green is with arc.  Arc capacitance is derived from your frequency measurements with 40" wire.  Arc resistance is a guess.


I did try changing to your higher-inductance primary.  Indeed, arc voltage was slightly lower that way.  Only 1.5dB or so.  That is with constant input voltage.  With higher primary inductance, current is lower.  If your H-bridge were limited by current rather than voltage, the higher inductance 8T primary would likely perform better.

[Duane B]

I have re-measured actual resonant frequencies of PRI and SEC

Secondary with toroid & cylinder topload, coil/topload removed from base
======================================
based on JavaTC = 353.3KHz Secondary and topload/s only
based on JavaTC = 343.6KHz sec/toploads/pri/mmc whole system
"series resonance test" signal gen / 10k res = 347KHz
"series resonance test" signal gen / 10k res =  289KHz w/simulated streamers , 40" wire upward

Your waveforms look more like a single resonant system to me. Scan the primary resonant circuit and see if you get two resonant frequency peaks.

ok, I assume youve built QCW before (?) . I have no idea how they would look like . 'these' waveforms are usually what I see online but sure, I will recheck it later 

No, I have not built QCW before. I don't even know what QCW stands for. I assume the CW is for Continuous Wave, but I have no idea what the Q is for. Maybe I shouldn't have contributed to this post. I thought I might be able to help from what other experience I have had. Your waveforms, re-posted here, reminded me of a single resonant system. The current waveform is the exact thing you get with driving a single resonant system. The hump you see in the beginning transient indicates that the frequency is slightly off resonance. If it were in resonance the current would rise to a maximum and stay at maximum until the driving source stops. Again, this reminds me of a single resonant system.



As far as adding a 40" wire to simulate a streamer, I say that you are only adding capacitance to the secondary and by no means is it simulating a streamer, not by a long shot! I still stand by what I said in my branch off post: the supposed capacitance of the streamer is not changing the frequency.

[Duane B]

David, I must start out by saying that I have the greatest respect for you. And I have never used LT Spice before. However, I think you have a problem with your simulated diagram. If I were to draw this diagram it would not have the streamer load as indicated. It would be a secondary coil in series with a resistor in series with the top load capacitance, and that is it (a voltage source in series with an inductance in series with a resistance in series with a capacitance). The resistor is a dynamic resistor. It would be difficult to make calculations with this resistor because it changes resistance with and without sparks, and with how much energy is delivered to the spark. One thing is certain though, there must be a resistor in series with the secondary inductance and capacitance!

I suppose you could use a Thevenin's or Norton's equivalent circuit, and simulate with either a series or parallel circuit, but you diagram has components both in series and parallel.

[davekni]

David, I must start out by saying that I have the greatest respect for you. And I have never used LT Spice before. However, I think you have a problem with your simulated diagram. If I were to draw this diagram it would not have the streamer load as indicated. It would be a secondary coil in series with a resistor in series with the top load capacitance, and that is it (a voltage source in series with an inductance in series with a resistance in series with a capacitance). The resistor is a dynamic resistor. It would be difficult to make calculations with this resistor because it changes resistance with and without sparks, and with how much energy is delivered to the spark. One thing is certain though, there must be a resistor in series with the secondary inductance and capacitance!

Thank you for the compliment.
I did omit the wire resistance of secondary coil.  I could add that for a small change to the result.  The top-load capacitance does not have any significant series resistance.  It is just capacitance to ground.  Before breakout, Q is quite high.  JavaTC estimates that Q, 196 in your case.  That Q is not infinite because of secondary winding resistance (a bit higher than DC resistance due to skin and proximity effects).
Yes, arc load is dynamic.  I simulated some random place towards the end of the ramp with a long arc.  Power dissipation in the arc is not due to resistance in series with top-load capacitance.  It is from arc resistance.  Arc current returns to ground through the arc's capacitance.  That is the added 4.7pF in my simulation.  The only way the arc conducts current is through its capacitance, not through top-load capacitance.
Of course, reality is always a bit more complex than models.  Arc capacitance and resistance are both distributed along the arc.  And there is some capacitance from arc sections to the top load as well as to ground.  Picture more of a linear network of series resistors with a capacitor from each junction to ground, and capacitors across the resistors.  I have made some arc models with more than one resistor and one capacitor.  Don't have enough real-world data to know what values to use.  The simple single resistor and capacitor version can match basic arc parameters (power dissipation and secondary frequency shift).
Hope that makes things more clear.  Please feel free to question again if not.  I certainly do make mistakes.

BTW, here are a few links from this forum that discuss arc load modeling:
https://highvoltageforum.net/index.php?topic=670.msg4474#msg4474
https://highvoltageforum.net/index.php?topic=626.msg4107#msg4107
https://highvoltageforum.net/index.php?topic=1073.msg7715#msg7715
https://highvoltageforum.net/index.php?topic=798.msg9265#msg9265

[Duane B]

Dave, thanks for your description of arc loading, and for the links for more information. I will study the links and if I have more questions I will post to another thread or start a new topic. I do not want to hijack this thread any more than I have.

[Rafft]

David

Thanks for the simulations

Does this refer to bottom of secondary grounded (normal operation condition)

this exactly

If you took a measurement with shorted secondary, that would allow calculating coupling factor between primary and secondary.

yes, tried it on my other cheap LCR meter, 0.3uH reading. JTC shows 0.31x something. it has low decimal count unlike my DIY(but un-useable with secondary coil in place)




I have learned & finalized my hardware/values and settle for 450KHz operating freq.

Ive recorded a video outdoors. funny thing is "squeel" along with the spark at 300Vdc. none at lower voltages. I will recheck hardware again(maybe loose wiring)

time for a full-bridge, proper toroid topload (4"dia), and bigger bulk cap), newer primary for more k. QCW is just another level of beast 

cheers
Ralph

[davekni]

Ive recorded a video outdoors. funny thing is "squeel" along with the spark at 300Vdc. none at lower voltages. I will recheck hardware again(maybe loose wiring)

Nice performance and video!
My high-frequency hearing (2kHz and up) is down ~30dB, so I'm not hearing a squeel.  However, the sound is sharper on some pulses, like a sudden change in amplitude rather than a linear ramp.  Have you managed to scope anything at 300V to look for differences between the normal and sharp sounding pulses?

[Rafft]

David, Thanks for the compliment.

Have you managed to scope anything at 300V to look for differences between the normal and sharp sounding pulses?

unfortunately I havent saved the waveform(270v bus). too bad. anyways, it was just an intermittent connection from secondary coil "gnd", to outside GND. it was not a solid connection. now Ive soldered the "start" end of coil, to a wire going outside for GND connection. judging from some QCW vids of others, it sounded like the arc was hitting/getting a ground strike.

currently waiting for my 2200uF x 3pcs to arrive.I HOPE this increases arc length  else this is the max arc length the (small-ish? 3" x 4.8")coil can do

im also checking whats the difference with upper pole and lower pole operation.

Code: [Select]

my secondary[reference]: Series resonance test
347KHz with topload
282KHz with topload and 40" wire loaded

**video above shows straight arcs at 454KHz(2.2uS), way above secondary/topload resontant freq.

***Ive tried higher cap value for MMC (same primary TAP, MMC 27nF) ->> 279KHz(3.58uS), Very Clean Iprim, arcs are branchy now, this is close to the secondary(loaded) res freq. sorry no videos. it was just a quick test if it even arcs at that low freq

**** NEXT to try 22nF (LC res calculation) ->> 384KHz, slightly above secondary resonant freq



for QCW, any idea what is the difference between upper pole and lower pole operation aside from the branchy and straight arcs?


just a grab from your post, the green solid line. does the lower peak suggest the lower pole?? and higher peak as upper pole?? if so, it makes sense to my calculations 

[davekni]

unfortunately I havent saved the waveform(270v bus). too bad. anyways, it was just an intermittent connection from secondary coil "gnd", to outside GND. it was not a solid connection. now Ive soldered the "start" end of coil, to a wire going outside for GND connection. judging from some QCW vids of others, it sounded like the arc was hitting/getting a ground strike.

Only reason for a waveform was to figure out a cause.  Great job figuring out the issue quickly.  No need for waveforms.

for QCW, any idea what is the difference between upper pole and lower pole operation aside from the branchy and straight arcs?

I don't have any real QCW experience.  My knowledge is from my failed low-frequency experiment, simulations, and reading this forum.  Hopefully others with real experience will share details.  (Also, search the forum for QCW and upper/lower pole.  There is information here.)

just a grab from your post, the green solid line. does the lower peak suggest the lower pole?? and higher peak as upper pole?? if so, it makes sense to my calculations 

Yes, green is with simulated arc load.  Left peak is lower pole.  Right (higher) peak is upper pole.  Yes, upper pole generates 12dB higher arc power for this simulated load and constant bridge voltage.

[Rafft]

Yes, green is with simulated arc load.  Left peak is lower pole.  Right (higher) peak is upper pole.  Yes, upper pole generates 12dB higher arc power for this simulated load and constant bridge voltage.

David, awesome! didn't think LTspice could "see" those poles. makes for the PRIMARY resonant freq "hunting" faster. I do have some knowledge using LTspice but with -simple- circuits only, and NOT tesla coils.

**what where your input parameters for this?

**could you maybe post here your file?

**screenshot for simulation tab? etc etc

**small tutorial perhaps?   

this would really prove helpful once I make another primary coil (conical) to increase coupling between pri/sec and a 4"dia toriod(to make it look MORE of a traditional tesla coil). all I'm doing is just getting the sec loaded/unloaded resonant freq  and add 60 - 110KHz (from sec unloaded), for the primary resonance.

 

[Uspring]

for QCW, any idea what is the difference between upper pole and lower pole operation aside from the branchy and straight arcs?

I believe the branchiness of arcs to be the result of a low operation frequency. It does not depend on the choice of poles. Here's a short Q&A regarding poles:

1. What are poles?

Poles are resonant frequencies. A single tank, e.g. has a single resonant frequency, which can be calculated from its capacitance and inductance. Two tanks have 2 res frequencies, which each can be calculated the same way. If you bring the tanks in proximity, so that the magnetic fields of the coils begin to overlap, the system becomes coupled and the resonant frequencies of the system change. Initially, the lower pole is the lower of the 2 tanks res frequencies and the upper pole the upper one. Once the tanks are coupled, the lower pole will move downward in frequency and the upper pole upward. The larger the coupling is, the more the poles will move away from the tanks original res frequencies.
You can see this in Davids diagram (red lines). The secondaries resonant frequency is 347 kHz and the primary res frequency is 410 kHz. The poles are at 320 kHZ and 465 kHz. For the arc loaded coil (green lines), the frequencies are shifted somewhat downward due to the arcs capacitance lowering the secondary res frequency.

2. At what frequency does my coil run?

Standard drivers usually employ zero current switching, which is gentle on the bridge transistors. This means, that primary voltage and primary current are in phase. In order to produce a diagram I've shamelessly stolen the parameters from Davids simulation. It shows the primary current amplitude and its phase wrt the input voltage.

The green dotted line (phase) crosses 0 degrees at 452 kHz, which is close to the observed frequency. Note, that there is only one zero crossing. This is due to the large arc load and/or to a large difference between the tanks res frequencies.
You can read off this diagram also the gain at the operating frequency. This is 3 dB, which means a primary current of 1.4 A for each Volt of input voltage. With this information you can match your bridges output capabilities to the coil.

Here is another diagram, where I have increased the primary MMC to 27 nF.

You can see now 3 zero crossings of the phase at 263, 293 and 359 kHz. All 3 are, in principle, zero current switching frequencies, but the center one has an inbuilt instability. Useable are really only the upper and lower frequencies, which correspond to the peaks (i.e. poles) in the current amplitude line. Generally, with zero current swutching, the DRSSTC drivers will start with the primary res frequency and then end up at the frequency of the nearest pole. This turns out to be the lower pole in this case. It was the upper pole in the previous diagram.

At the lower frequency, where you are running with the 27n cap, the current amplitude is now at -8dB, i.e. 0.4 A per input Volt, so current draw is much lower than in the previous example. I'd expect considerably shorter arcs, if you haven't increased input voltage. And the shorter arc is probably the reason, why your measured operating frequency of 279 kHz is somewhat above the calculated one of 263kHz. Since I did not change Davids arc loading circuit, the calculated frequencies are a bit too small. Smaller arcs have less capacitance and don`t decrease the res frequencies as much.

3. What is a big primary/secondary fres difference?

so in summary:
Secondary Fres = 343.6KHz
Primary Fres = 435KHz
91KHz difference. IS this a BIG or small difference? is this NORMAL ? I mean the freq difference. I was thinking maybe a 10-20KHz would do, but I guess it needs bigger difference cuz its higher freq compared to DRs in the 50-150KHz (?)

That depends mainly on the coupling factor. Large frequency differences imply bad primary -> secondary power transfer efficiencies. But this can be compensated by a large coupling. QCWs often have a large coupling, while that of standard DRSSTCs is much lower. So QCWs can live with larger frequency differences.
Large couplings in standard DRSSTCs are difficult due to racing arcs and flashovers. QCWs often are built ot produce sword arcs. These arcs need to be slowly ramped up and run at high frequencies. Both of this makes them create relatively low voltage arcs. And this simplifies achieving a high coupling without undesirable side effects.

[Rafft]

Uspring

A big thanks for the explanations.  that explains most of what I'm seeing on my coil setup.


today's hardware update:

Epcos 2200uF 350Vdc bulk caps (6,600uF in total)


I Have gone Full-bridge of IRG4PC50FD. heatsink is from an old HDD


my coil. ditched the CAN top, replaced back to using 2 small toroids, so it looks more like a tesla coil


secondary with 2 toploads = 379KHz , 310KHz loaded with 36" wire (measured using sig gen and scope)
Primary is 11.8uH(8th tap) 9.9uH(7th) 7.8uH(6th) and MMC is 16.5nF(4s3p of 22nF)
k = 0.39
sec 3.2"D x 4.85"H #33 awg 683T 610KHz (jtc)
topload 1.3" 6.4"  423KHz (jtc) + 1.3" x 6.8" 389KHz (jtc)


calculating upper and lower pole
Fupper = F/sqrt(1-k)
Flower = F/sqrt(1+k)

Fupper = 485KHz
Flower = 350KHz

Ive tested 8T = 297KHz  , 7T = 431KHz , 6T = 450KHz, 6T giving the longest spark. I have only tested with 180Vdc(indoors) and it has already surpassed(more or less) the spark length compared to half-bridge on 250Vdc.

now my question is, my phase lead is different now , is this GOOD or BAD?


also, when I switch between 7T and 8T, it locks on to the upper pole(431KHz) and lower pole(297KHz), respectively. its like a toggle switch.. again , is this normal??  I had to take a second measurement just to see that I wasnt dreaming or making stuff up 


cheers
Ralph

[davekni]

I lost track of this thread, so here's some belated answers in case such is still useful.

David, awesome! didn't think LTspice could "see" those poles. makes for the PRIMARY resonant freq "hunting" faster. I do have some knowledge using LTspice but with -simple- circuits only, and NOT tesla coils.

Any of the spice-type simulators will show poles of coupled resonators using an AC frequency sweep.  Circuit isn't too complex, as you can see from the schematic I posted (as a JPG image).

**what where your input parameters for this?

All parameters are shown on the spice schematic I posted.  The simulation is for 1Vpeak drive as per the "AC 1" label on voltage source V1.  This could be changed to actual bridge output voltage.  Result would shift the plot in dB up (scale output), but not change shape at all.  R2 is a guess at total resistance of bridge IGBTs and primary coil, 0.1 ohms in my example.  L3 is a guess at wiring inductance from bridge through MMC to primary.  C2 is MMC.  L1 is primary coil, 10uH.  L2 is secondary, 19.68mH.  Directive "k12 l1 l2 0.32" defines the primary-to-secondary coupling factor.  C3 is top load capacitance, 10.7pF here.  C4 and R1 are a guess at arc loading for some point later in the ramp.  The "ac lin 4k 200k 600k" directive tells LTSpice to sweep frequency linearly from 200kHz to 600kHz with 4000 steps.  The "step param arc list 4.7p 10f" runs the simulation twice, first with 4.7pF for arc capacitance, second with 10fF for arc capacitance.  10fF is approximately 0, but avoids the issues with 0 value components.

**could you maybe post here your file?

Here's the LTSpice schematic file in ZIP format, as native LTSpice files are not accepted for posting here.
See end of this post for the file.  Appears that only JPG images are allowed in-line with text.

**screenshot for simulation tab? etc etc

Not sure if this is what you are after.  Here's a screen shot of the entire LTSpice window running this simulation:

**small tutorial perhaps?   

Hopefully the explanation below your first question covers most of the details.  Unzip the "sstc_ac1.asc" file and double-click it to execute it with LTSpice.  Click the run symbol (person running), then click on node "vo2" in the schematic to plot top-load voltage.

now my question is, my phase lead is different now , is this GOOD or BAD?

I presume the yellow trace is one side of H-bridge output.  What is the cyan trace?  An antenna?  If it is primary current, something is strange.  It appears to be switching at roughly peak current, not just before zero.  Wouldn't expect you to get much current if voltage is being driven almost 90 degrees out-of-phase with current.

also, when I switch between 7T and 8T, it locks on to the upper pole(431KHz) and lower pole(297KHz), respectively. its like a toggle switch.. again , is this normal??  I had to take a second measurement just to see that I wasnt dreaming or making stuff up 

Although I've never actually ran anything at the upper pole yet, I think this fits LTSpice simulations of such when approximating a UD2.7-style driver.  After you get comfortable with the one AC simulation here, I can dig up my transient (time) simulation of a simplified H-bridge into DRSSTC.

[Rafft]

David,

thanks. will have a go at that LTspice later 

I presume the yellow trace is one side of H-bridge output.  What is the cyan trace?  An antenna?  If it is primary current, something is strange.  It appears to be switching at roughly peak current, not just before zero.  Wouldn't expect you to get much current if voltage is being driven almost 90 degrees out-of-phase with current.

Yellow is Full-bridge output. Cyan is primary current (3rd CT). this happened when I switched from Half-bridge to Full-Bridge.

maybe probe for bridge( is reversed?)

btw, I have changed the variable inductor 14-50uH(old ckt) to 6-24uH(current ckt)...burden still the same 51R 3W(carbon comp)... could this smaller L affect the phase lead? but I do remember using THIS value while it was still in half-bridge, and it worked  as it should. or is this value too low for 250KHz-450KHz operation?

how is the phase lead inductor+Resistor working? is it when -in resonance- with that freq that it "shifts it"?

Although I've never actually ran anything at the upper pole yet, I think this fits LTSpice simulations of such when approximating a UD2.7-style driver.  After you get comfortable with the one AC simulation here, I can dig up my transient (time) simulation of a simplified H-bridge into DRSSTC.

what I meant was this:

secondary/topload/s = 379KHz unloaded, according to JTC
Upper/Lower pole = 485KHz/321KHz
k = 0.39

calculated :                                                     actual coil operation:
8T(11.8uH) & 4s3p(16.5nF) = 360.7KHz            298KHz
7T(9.9uH)                          = 394KHz               431KHz
6T(6.8uH)                          = 443.6KHz            450KHz

numbers in BOLD has really big freq difference. 7T is upper pole and when changing to 8T, it sort of flips down directly to lower pole. I still remember you said thats how the UD works, in lower pole.

everything with this full bridge is wierd. lol. back when I was in half-bridge, coil wont start unless I give the buck a standby voltage of around 40-50Vdc, but now that its in full-bridge, I can start the coil EVEN with buck output at almost 0v initial (and with a 0v - max volt rise RAMP, I get a "solid" spark. solid like SOLID..not hissy or hazzy etc etc call it a lovely sword spark if you will)
correction: it STILL needs that 'wick' but minimal voltage only. when I set the 'wick' to minimum(or zero), coil almost always misses to spark.


some javaTC data if you want

Code: [Select]

J A V A T C version 13.6 - CONSOLIDATED OUTPUT
18/02/2022, 10:28:07

Units = Inches
Ambient Temp = 68ºF

----------------------------------------------------
Secondary Coil Inputs:
----------------------------------------------------
Current Profile = G.PROFILE_LOADED
1.6 = Radius 1
1.6 = Radius 2
1 = Height 1
5.85 = Height 2
683.1 = Turns
33 = Wire Awg

----------------------------------------------------
Primary Coil Inputs:
----------------------------------------------------
Round Primary Conductor
2.9 = Radius 1
2.7 = Radius 2
1 = Height 1
3.35 = Height 2
8.462 = Turns
0.079 = Wire Diameter
0 = Ribbon Width
0 = Ribbon Thickness
0.0165 = Primary Cap (uF)
0 = Total Lead Length
0 = Lead Diameter

----------------------------------------------------
Secondary Coil Outputs:
----------------------------------------------------
349.83 [kHz] = Secondary Resonant Frequency
90 [deg °] = Angle of Secondary
4.85 [inch] = Length of Winding
140.8 [inch] = Turns Per Unit
0.00002 [inch] = Space Between Turns (edge to edge)
572.3 [ft] = Length of Wire
1.52 [:1] = H/D Aspect Ratio
117.4373 [Ohms] = DC Resistance
39257 [Ohms] = Reactance at Resonance
0.09 [ lbs] = Weight of Wire
17.86 [mH] = Les-Effective Series Inductance
19.29 [mH] = Lee-Equivalent Energy Inductance
19.032 [mH] = Ldc-Low Frequency Inductance
11.589 [pF] = Ces-Effective Shunt Capacitance
10.73 [pF] = Cee-Equivalent Energy Capacitance
18.142 [pF] = Cdc-Low Frequency Capacitance
4.68 [mils] = Skin Depth
8.078 [pF] = Topload Effective Capacitance
244.5588 [Ohms] = Effective AC Resistance
161 [Q] = Quality Factor

----------------------------------------------------
Primary Coil Outputs:
----------------------------------------------------
365.34 [kHz] = Primary Resonant Frequency
4.25 [% low] = Percent Detuned
85 [deg °] = Angle of Primary
12.41 [ft] = Length of Wire
20.77 [mOhms] = DC Resistance
0.2 [inch] = Average spacing between turns (edge to edge)
1.057 [ inch] = Proximity between coils
0 [inch] = Recommended minimum proximity between coils
11.502 [µH] = Ldc-Low Frequency Inductance
0.018 [µF] = Cap size needed with Primary L (reference)
0 [µH] = Lead Length Inductance
182.911 [µH] = Lm-Mutual Inductance
0.391 [k] = Coupling Coefficient
0.127 [k] = Recommended Coupling Coefficient
2.56 [half cycles] = Number of half cycles for energy transfer at K
3.16 [µs] = Time for total energy transfer

----------------------------------------------------
Top Load Inputs:
----------------------------------------------------
Toroid #1: minor=1.3, major=6.4, height=7, topload
Toroid #2: minor=1.3, major=7, height=8.3, topload


Code: [Select]

units=0,
ambient=0,
s_ws=0,
s_Al=0,
p_ws=1,
p_Al=0,
p_ribbon=0,
temp=68,
g_radius=0,
w_radius=0,
ceil_height=0,
s_radius1=1.6,
s_radius2=1.6,
s_height1=1,
s_height2=5.85,
s_turn=683.1,
s_wd=33,
p_radius1=2.9,
p_radius2=2.7,
p_height1=1,
p_height2=3.35,
p_turn=8.462,
p_wd=0.079,
p_vwidth=0,
p_rthick=0,
Cp_uF=0.0165,
Lead_Length=0,
Lead_Diameter=0,
desired_k=0,
t.inner=1.3,
t.outer=6.4,
t.height=7,
TT=true,
TG=false,
t.inner=1.3,
t.outer=7,
t.height=8.3,
TT=true,
TG=false,
x_Vin=0,
x_Vout=0,
x_Iout=0,
x_Hz=0,
x_Vadjust=0,
x_ballast=0,
rsg_ELS=0,
rsg_ELR=0,
rsg_rpm=0,
rsg_disc_D=0,
rsg_ELR_D=0,
rsg_ELS_D=0,
stat_EL=0,
stat_EL_D=0,
stat_gap=0,
SPE=true,
RGE=false


edit:
been doing the LTSpice sim. Ive edited the primary inductance and MMC. secondary and topload. and the coupling as well. values used was from above jtc. I hope its correct?

btw HOW do you make inductors (like L1 and L2) be coupled together? this is based on the directive "k12 l1 l2 0.32", right?  though I dont understand what 'k12' is.



edit2:
on second thought (regarding strange waveform between bridge and Iprim), I will just have to RE-wind that GDT. I could have messed up the wiring(probably) ... it was made 12yrs ago w/c never got used. and this is my 1st time using full-bridge..if memory serves me right. better late than never 


=========================
edit3:
I have replaced the big GDT with my smaller toroid(wound with 5-filar, 100uH), yellow (bridge) &  cyan (Iprim)

result still the same... tried with the variable inductor and the result below

testing with 50Vdc bus(DR only) , with slug removed(6uH) on the input LC


with slug installed, around 24uH

[davekni]

Yellow is Full-bridge output. Cyan is primary current (3rd CT). this happened when I switched from Half-bridge to Full-Bridge.

maybe probe for bridge( is reversed?)

Probe reversal would shift the waveform by 180 degrees, so not explain a 90 degree offset.  It's surprising that you can get much primary current at all with almost 90 degree phase shift.  Perhaps the phase shift is closer to 0 (or 180) earlier in the burst, allowing primary current to build.

btw, I have changed the variable inductor 14-50uH(old ckt) to 6-24uH(current ckt)...burden still the same 51R 3W(carbon comp)... could this smaller L affect the phase lead? but I do remember using THIS value while it was still in half-bridge, and it worked  as it should. or is this value too low for 250KHz-450KHz operation?

edit3:
I have replaced the big GDT with my smaller toroid(wound with 5-filar, 100uH), yellow (bridge) &  cyan (Iprim)

result still the same... tried with the variable inductor and the result below

testing with 50Vdc bus(DR only) , with slug removed(6uH) on the input LC

It looks like delay through the driver and H-bridge is more than 90 degrees.  That can work with feedback inverted (CT polarity, GDT polarity, etc.).  With more phase-lead (24uH), voltage and current are almost 90 degrees off (not desired).  With less phase-lead (6uH), voltage is delayed (not advanced as much), so closer to current phase.
The change to full-bridge could be related to this.  Twice the load on driver and GDT will slow down gate waveforms (add delay).

calculated :                                                     actual coil operation:
8T(11.8uH) & 4s3p(16.5nF) = 360.7KHz            298KHz
7T(9.9uH)                          = 394KHz               431KHz
6T(6.8uH)                          = 443.6KHz            450KHz

After some thought, I have experienced this on my DRSSTC.  When arc loading capacitance gets high enough (long arcs), it reduces secondary frequency below primary frequency.  I have scope traces showing this transition from lower-pole to upper-pole operation at the end of a long arc enable pulse:
https://highvoltageforum.net/index.php?topic=798.msg9200#msg9200
Difference is that frequency remains similar.  Mostly the pole locations are moving due to arc capacitance.

https://highvoltageforum.net/index.php?topic=798.msg9200#msg9200

Phase lead of UD2.7 (L+R series circuit) works whether at resonance or not.  You can demonstrate this in LTSpice with a very simple circuit.  Current source (instead of voltage source) simulated CT output.  Feed that current source through a series inductor and resistor.  Plot voltage of the current source (voltage of series R+L) over a sweep of frequencies.  Phase lead does change some with frequency, but no sudden change around Tesla coil resonances.  (There is no capacitor in this phase-lead circuit, so that circuit is not itself resonant.)

btw HOW do you make inductors (like L1 and L2) be coupled together? this is based on the directive "k12 l1 l2 0.32", right?  though I dont understand what 'k12' is.

Yes, just add the "k" statement and inductors are coupled.  "k12" is an arbitrary label, but must start with "k" so LTSpice knows it is a coupling factor.  The two (or more) inductors that are coupled are defined by the "l1" and "l2" strings.  Those must match the instance labels of the inductors.

edit3:
I have replaced the big GDT with my smaller toroid(wound with 5-filar, 100uH), yellow (bridge) &  cyan (Iprim)

You could try the 8-filar version (four paralleled primary windings) for even lower parasitic inductance:
https://highvoltageforum.net/index.php?topic=1854.msg13949#msg13949

However, my guess is that you are running past 90 degrees of phase lag.  Rather than attempting to reduce delay, it may be better to live with that delay and perhaps add a bit more to achieve close to 180 degrees.  Slightly less than 180 is as good as the usual slightly less than 0 degrees.  Only down side is that phase will change more as frequency changes.  (I've build resonant H-bridge systems using phase lag instead of lead, though not for a Tesla coil.)

[Uspring]

secondary/topload/s = 379KHz unloaded, according to JTC
Upper/Lower pole = 485KHz/321KHz
k = 0.39

calculated :                                                     actual coil operation:
8T(11.8uH) & 4s3p(16.5nF) = 360.7KHz            298KHz
7T(9.9uH)                          = 394KHz               431KHz
6T(6.8uH)                          = 443.6KHz            450KHz

Usually the coil will operate at the lower pole, if primary fres is lower than the secondary fres and at the upper pole if primary fres is larger than the secondary fres. This is, what you see here.
If res frequencies are similar, the coil will, for a while, run at both pole frequencies. This shows up as beats in the primary current waveform.

[Rafft]

guys

Ive converted my synchro buck, to non-sync. so that instead of using 2 igbt(hi & lo), I can just use both for HI side, and a diode for low.

it all works on bench test. 10v & 20v input, with 2 2.2R 10w(wirewound) in parallel(as my load)

now in actual, I have the output LC loaded with 3kR 5w (wirewound, three 1k5w in series). DC bus is 113Vdc. when I trigger it, its in DR mode sparks(and not sword sparks).

when I connect scope probes(set to x10), spark output goes sword. If I remove probe connection, goes back to branching sparks.

Ive even used 3k9 3w(carbon comp) and 680R 5w wirewound.. still same result



ideas?


edit:
I just remembered seeing a video on ytube where he had bulbs (near the buck).. hmmm I wonder if this would work to drain the output quickly(?)

BUT why does connecting a scope probe 'fix' the ramp output? my coil is battery-powered and no connection to ac line (only scope is connected to ac line, for power)

edit2:
maybe this is the downside of being an async. compared to a sync w/c never uses any 'load' to discharge because its already being handled by the low-side switch.

edit3:
  https://4hv.org/e107_plugins/forum/forum_viewtopic.php?156071
based here (Steve Ward), maybe I just go back to synchro and use a mosfet for the lo side? better switch than an IGBT?

[davekni]

BUT why does connecting a scope probe 'fix' the ramp output? my coil is battery-powered and no connection to ac line (only scope is connected to ac line, for power)

My guess would be a change in grounding rather than a change in ramp shape.  Scope is adding a ground connection path.  If it is ramp related, I'd suspect that without scope ground something in your buck control circuit is getting confused by noise.

[Rafft]

BUT why does connecting a scope probe 'fix' the ramp output? my coil is battery-powered and no connection to ac line (only scope is connected to ac line, for power)

My guess would be a change in grounding rather than a change in ramp shape.  Scope is adding a ground connection path.  If it is ramp related, I'd suspect that without scope ground something in your buck control circuit is getting confused by noise.

come to think of it, heatsink for buck is floating. will add bypass cap to buck GND

David

would the lo-side benefit from a mosfet(as switch) + another diode in parallel? my thinking is the diode will be the 1st one to catch , followed by the fet(so it does not have to carry the bulk of the discharge). would this work?

[Rafft]

update:

 *seems*  Ive forgotten to reconnect the output LC GND to  buck control GND. this connection was on the LO out going to IGBT!  sheeesh! that was the reason why async did NOT work before!

anyways, problem solved! buck now more robust. only high side switch to worry about 

now on to higher coupling! (k=0.59) I hope I dont get flashovers . lol

[davekni]

anyways, problem solved! buck now more robust. only high side switch to worry about 

Great to hear that you've found and fixed the issue!

would the lo-side benefit from a mosfet(as switch) + another diode in parallel? my thinking is the diode will be the 1st one to catch , followed by the fet(so it does not have to carry the bulk of the discharge). would this work?

Presuming I'm understanding correctly, this is often called "synchronous rectification".  An external diode isn't required, but can be useful if lower Vf and faster than the MOSFET's internal body diode.  Synchronous rectification is common in low-voltage converters where diode Vf is a significant fraction of output voltage.  For your buck, I wouldn't bother with synchronous rectification.  Diode(s) alone as you have now should work well, presuming fast diodes and sufficient current capability.

[Rafft]

Hello again

currently im using (slow) IGBTs IRG4PC50FD in full-bridge for the inverter and a slightly modded UD2.1b.

1:32:31/10R for OCD and I have set LM311 comparator to trip at 1.1v (110Amp). ramp from buck ranges from around 4mS to 8mS. anything more than that, its just adding *hiss* and no more spark length. my boost can generate up to 300Vdc. storage cap 6600uF total.

testing some lower taps on my primary coil CAN trigger the ocd. my coil setup loves to run on 450KHz. resonant freq of secondary/topload is 377KHz(15.34mH/8.135pF k=0.374), so basically its running on upper pole.

am I conservative at 110Amps? what could possibly be my 'max' amp setting from these IGBTs?

thanks

[davekni]

am I conservative at 110Amps? what could possibly be my 'max' amp setting from these IGBTs?

I'd guess you could go slightly higher.  Depends on your risk tolerance for fried parts, and on how well phase-lead is working (how significant switching loss is).  Specified peak current capability for IRG4PF50WDPbF is 204A.  I don't see a time duration for that peak.  1ms is typical for such specs.

Consider the case of 200A peak.  Average current for 200A sine wave is 2/PI * 200A = 127.3A.  Each IGBT conducts half the time, so 64A average.  Extracting from graphs, guess of 5V average forward drop for 320W conduction loss.

Switching loss is the harder part to estimate.  Even with "ideal" phase lead, switch-off needs to occur when some current is still flowing to cause H-bridge output voltage transition.  Switch-off also needs to be early enough so that opposite IGBTs turn on before significant current reversal after dead-time.  Highest turn-off loss spec. is at 28A and 25C, 1.38mJ.  No data on turn-off energy vs. temperature.  Also would be hard to get phase-lead perfect enough to be down from 200A to 28A, especially at 450kHz.  I'll take a wild guess at 5mJ for turn-off energy with hot IGBT die and whatever current switch-off occurs at.  That's 450kHz * 5mJ = 2250W per IGBT, much more than conduction loss.  Total for this guess is 2570W.

If IGBT die starts at 50C and rises to 150C, 100C delta, tolerable thermal resistance is 100C / 2570W = 0.039C/W.  Spec. thermal impedance plot hits that at ~250us pulse width.  Since you need at least a couple milliseconds close to limit current, 200A is likely too high.  Of course, this is a wild guess at switch-off loss, which is dominant for your high frequency use of medium-speed IGBTs.  200A could work for short bursts of a standard DRSSTC per this guess calculation.

[Rafft]

David

I'd guess you could go slightly higher.   

thanks! I really like this answer  I was thinking maybe I could push this to 120A-130A . Ive read some old posts that this specific IGBT could go 4x its Ic rating(or was it the Icm x4) but that was for DR and 50KHz.

with regards to your explanation,  this still goes down to using a better (faster) IGBT then. was thinking of using THIS
https://www.google.com/url?sa=t&source=web&rct=j&url=https://uelectronics.com/wp-content/uploads/2021/02/MBQ60T65PES-Datasheet.pdf&ved=2ahUKEwiAouKrosf2AhWRslYBHZb9CnsQFnoECDEQAQ&usg=AOvVaw271Wb2Nv5obmTWlhJivFyC

hope this will be better

[davekni]

thanks! I really like this answer  I was thinking maybe I could push this to 120A-130A . Ive read some old posts that this specific IGBT could go 4x its Ic rating(or was it the Icm x4) but that was for DR and 50KHz.

For low frequency, my rule-of-thumb is that IGBTs fry at 4x continuous current rating or 2x peak current rating.  Margin is needed below frying current.  I've ran destructive testing on two of my STGW60H65DRF parts, getting a ways past that rule-of-thumb.  Ran second one to 650A peak, more than 2x it's 240A peak rating.  Diode fried first at 600A peak, but still over the diode's 240A peak rating.  This was at 80kHz to match my DRSSTC.  Short ring-down bursts, so likely to fry at a bit lower current in real DRSSTC use.

with regards to your explanation,  this still goes down to using a better (faster) IGBT then. was thinking of using THIS
https://www.google.com/url?sa=t&source=web&rct=j&url=https://uelectronics.com/wp-content/uploads/2021/02/MBQ60T65PES-Datasheet.pdf&ved=2ahUKEwiAouKrosf2AhWRslYBHZb9CnsQFnoECDEQAQ&usg=AOvVaw271Wb2Nv5obmTWlhJivFyC

hope this will be better

Yes, spec's look great!  Very low Eoff and very low gate charge.  (And Eoff is specified at 60A rather than 28A as your existing part.)  Have you found a place to purchase these?  A quick look at oemstrade.com shows little stock availability.  I may buy some if/when they are available.

[Rafft]

Have you found a place to purchase these?

yes. its sold by an online company/seller who does COD.

ordered a few. I do hope I get the real deal.

edit 1:
and from your last post, IRG4PF50WD(900v device) is different from IRG4PC50FD(600v Ic=70A@25degC 39A@100degC Icm=280A and much slower than PF50WD)

[Rafft]

ok. making progress! 

tried 'high impedance' PRImary setup. and after a few more calculations regarding HOW to tune it(almost), im Loving the result!  old logs/build notes really help!

didnt think(!) I would get a goooood spark output with LESSER amp draw. coil freq=450KHz

this is from 3rd CT 1:100T/1R burden. 100A per division. blue trace(Iprim) yellow is full-bridge output. its about 50Amps only. almost half of my usual.


my limiter is currently set to 1.1V(110A) so I guess im -safe- to push it farther a bit to maybe 150A 

bridge output does not look 'sharp' at the top edge. nothing to worry, Ive just overdriven the ramp gen to do THAT 

[Rafft]

hi guys

how is it easy(or hard) to implement a frequency change-over switch for the UD2.x?  an adjustable free-running oscillator will initiate the spark then natural feedback would take over in a few clock cycles. doable with logic gates? arduino?

[davekni]

Haven't tried this for QCW, but one option is to modify UD2.7 to be self-oscillating.  Here's a schematic of my preferred variation:
https://highvoltageforum.net/index.php?topic=1336.msg9894#msg9894

A variation can be done with fewer modifications to UD2.7 as described in third paragraph of this post:
https://highvoltageforum.net/index.php?topic=1373.msg10197#msg10197

These are intended for more reliable DRSSTC startup in cases where feedback is weak or initial H-Bridge state is wrong (same output voltage already as what the driver initially forces after enable starts).  However, by adjusting component values, the self-oscillation frequency could be made to match a QCW upper-pole frequency and made a bit stronger so that feedback doesn't overtake it for several cycles.  If you are interested in trying this approach, I'd be happy to help further with figuring out best component values etc.  Would be an experiment, not a guaranteed solution.

[Rafft]

Hi David

remove R7(100k), add 50k and adjust C33. seems doable. what does SV1 mean? short the inductor? if so, no more phase lead?

If I decide to do this, should I 'tune' my coil near lower pole(as calculated together with coupling coeff)?

so what happens is coil(tuned to lower pole) will start oscillating at upper pole, then afrer a few cycles, oscillate to lower pole, right?

[davekni]

Part number references are for this UD2.7 schematic:
https://www.loneoceans.com/labs/ud27/UD27Cschematic.png

SV1 pins 2-3 are shorted to include phase-lead, pins 1-2 for no phase lead.  If the jumper is left off, there is no burden resistor for feedback CT, so it won't work correctly.  Sorry for the confusion.  I should have listed the schematic reference.

If I decide to do this, should I 'tune' my coil near lower pole(as calculated together with coupling coeff)?

Not sure exactly what this means.  I think that QCW designs using UD2.7 without any special features need to tune uncoupled primary frequency slightly higher than uncoupled secondary frequency.  That way UD2.7 tends to start at upper pole (presuming sufficient phase lead).  With a PLL design or UD3.x, it is possible to force upper pole operation even with primary frequency tuned lower than secondary.  I believe this improves performance.  Arc loading lowers secondary resonance, keeping frequencies close over entire arc growth.  That should improve performance.

(One qualifier on any QCW suggestions I offer:  My only personal QCW experience is my low-frequency QCW experiment.  It was more like a QCW SSTC with coupling of 0.9, way higher than other designs.  All other QCW knowledge is from simulations and analyzing designs posted here.)

so what happens is coil(tuned to lower pole) will start oscillating at upper pole, then afrer a few cycles, oscillate to lower pole, right?

I'm guessing that "tuned to lower pole" means primary tuned lower than secondary, in which case UD2.7 starts at lower pole.  I think the goal is to keep operation at upper pole for the entire ramp.  (Both pole frequencies will drop with arc loading.)  I am not at all certain that UD2.7 will stay at upper pole once feedback takes over.  Hopefully someone else can offer insight here.  (Or perhaps I can find time in the next week or two to run simulations.)

If we make the UD2.7 self-oscillation strong enough, perhaps it can still influence frequency even after feedback is strong enough to provide most of the signal.  Don't know.  Should have a better chance of this if D1 and D2 are silicon signal diodes (1N4148 or similar) rather than schottky diodes, and R2 increased to 5-7k.  With that, I'm guessing (without simulation yet) that C33 value of ~560-680pF would result in 450kHz operation before feedback.  (Test with no power to H-bridge.  Measure frequency on UD2.7 output or GDT or IGBT gate.)

Good luck!  Remember, no pressure to try this.  Only if you feel like experimenting.  I am quite curious how it goes if you do try.  Or wait a bit for me to try simulations first.

[Rafft]

David
primary current

I'm guessing that "tuned to lower pole" means primary tuned lower than secondary, in which case UD2.7 starts at lower pole.  I think the goal is to keep operation at upper pole for the entire ramp.  (Both pole frequencies will drop with arc loading.)  I am not at all certain that UD2.7 will stay at upper pole once feedback takes over.  Hopefully someone else can offer insight here.  (Or perhaps I can find time in the next week or two to run simulations.)

yes. exactly. kindly check my javaTC file. MY coil resonates from 450KHz(ramp start) going down to 431KHz(ramp TOP). 5mS total, but not including ramp-down(roughly 1mS). I tried higher tap (13.2uH) and it went down to lower pole (316KHz-312KHz @ 45Apk only, low spark output). btw, Im using the MBQ60T65PES (full bridge - 4pcs IGBT). I doubt they are the real deal(they look ugly). not sure also if this is(was) developed in china. havent pushed it yet, just limiting to 120Apk.

If we make the UD2.7 self-oscillation strong enough, perhaps it can still influence frequency even after feedback is strong enough to provide most of the signal.  Don't know.  Should have a better chance of this if D1 and D2 are silicon signal diodes (1N4148 or similar) rather than schottky diodes, and R2 increased to 5-7k.  With that, I'm guessing (without simulation yet) that C33 value of ~560-680pF would result in 450kHz operation before feedback.  (Test with no power to H-bridge.  Measure frequency on UD2.7 output or GDT or IGBT gate.)

I will see what I can do. my UD is built on pad-per-hole board. almost every pcb trace, is a component leg.  im using MAX913 for the comparator.

what Im seeing here is that start-up is 450Khz(this is close to 430KHz=big sparks), coil (feedback to be 'lower pole', as set by PRI LC), would the output go from big spark to 'smaller' since it will get feedback from lower pole freq?

ok let me wait for your SIMs 

my coil currently (javaTC)

Code: [Select]

units=0,
ambient=1,
s_ws=0,
s_Al=0,
p_ws=1,
p_Al=0,
p_ribbon=0,
temp=20,
g_radius=0,
w_radius=0,
ceil_height=0,
s_radius1=1.6,
s_radius2=1.6,
s_height1=9.25,
s_height2=13.7,
s_turn=626.8,
s_wd=33,
p_radius1=2.9,
p_radius2=2.8,
p_height1=9.25,
p_height2=10.33,
p_turn=6.834,
p_wd=0.079,
p_vwidth=0,
p_rthick=0,
Cp_uF=0.0142,
Lead_Length=0,
Lead_Diameter=0,
desired_k=0,
t.inner=1.3,
t.outer=6.4,
t.height=14.7,
TT=true,
TG=false,
t.inner=1.3,
t.outer=6.8,
t.height=16,
TT=true,
TG=false,
x_Vin=0,
x_Vout=0,
x_Iout=0,
x_Hz=0,
x_Vadjust=0,
x_ballast=0,
rsg_ELS=0,
rsg_ELR=0,
rsg_rpm=0,
rsg_disc_D=0,
rsg_ELR_D=0,
rsg_ELS_D=0,
stat_EL=0,
stat_EL_D=0,
stat_gap=0,
SPE=true,
RGE=false

screenshot grab of the igbt internals (taken from buyer comments, same seller I got these igbt from)... small? bad? legit?




-Ralph

[davekni]

Ralph,

yes. exactly. kindly check my javaTC file. MY coil resonates from 450KHz(ramp start) going down to 431KHz(ramp TOP).

Simulating with your JavaTC results starts at 470kHz.  That discrepancy is within reason.  Perhaps a bit more top-load capacitance due to non-infinite distance to surroundings or some such reason.

I will see what I can do. my UD is built on pad-per-hole board. almost every pcb trace, is a component leg. my C33 is 1nF(is this ok?), still based on orig schema.

ok let me wait for your SIMs 

Short answer:  Probably not reasonable to try with your driver.  I did manage to get simulation cases to work, but with significant circuit changes.  Would likely require experimentation (adjusting values) to get it functioning in practice.  In case it is of interest, here is a working simulation example:

 [ You are not allowed to view attachments ]
Attachment ended up at end of this post.

Required dramatically reducing feedback CT signal strength and increasing comparitor positive-feedback amplitude, along with adjusting self-oscillation frequency to slightly above upper pole frequency.  With those changes, remained at upper pole with change of primary MMC from 14.2nF to 17nF, so primary frequency below unloaded secondary frequency.  Had some cases working even at 20nF MMC, but tricky there.  Narrow range of component values that worked.

BTW, B1 is a voltage-controlled voltage source.  It is replacing (for simulation) gate driver chip through IGBTs.  Since it has zero delay, phase lead is set very low here.  Would need to be increased to match real circuit delays.

and im using MAX913 for the comparator.

Should be fine for normal UD2.x, but would make above changes even more tricky.  Has higher input bias current and lower output voltage swing than TL3116 , affecting values needed.

I doubt they are the real deal(they look ugly).

screenshot grab of the igbt internals (taken from buyer comments, same seller I got these igbt from)... small? bad? legit?

Certainly can't tell for sure, but could be genuine.  Would be hard to fit much larger die in a TO247 package.  (The upper die closer to the gate lead should be the IGBT and the lower one the anti-parallel diode.)
MagnaChip must sell directly to equipment manufacturers mostly.  I find relatively little about the company and few of their parts in normal electronics distribution.  Appears to be primarily a South-Korean company, though not officially incorporated there.  Quite possible that IBGT/diode die are made there and packaged in China.  Could even be die made in China by or for MagnaChip.  I ordered some from EBay (from China).  AliExpress doesn't work for me for some reason.  Cheaper there, though.
In online pictures I noticed that some parts have numbers or letters molded into the two small disks on package front, while others are blank.  That could be some indication of genuine or not.  I don't know.
Next similar but smaller MagnaChip die is MBQ50T65DSC.  Slower IGBT with higher gate capacitance.  Do you have a meter than can read capacitance?  Try measuring gate capacitance to drain/source shorted together.
Perhaps the most likely scenario is that these IGBTs are rejects from the packaging line (or parts from molding machine setup runs).  In that case they may have genuine die but not complete marking or clean packaging.  Could be more moisture sensitive.  Could have higher failure rate if die-attach is reason for reject (causing hot-spots within die).  Since MagnaChip parts are not generally available from normal sources, they seem like unlikely parts to counterfeit.  Anyone bothering to counterfeit parts would presumably be interested in faking popular parts from well-known companies.

Have fun with your coil!

[Rafft]

David

initial reply(using only phone)
using my store-bought LC meter.
MBQ60T65PES G-C 6nF G-E 4.17nF G-C&E(shorted) 8nF

using my diy LC meter(proven to be more accurate at small values)
MBQ60T65PES G-C 3.479nF G-E 4.193nF G-C&E(shorted) 6.25nF

just for comparison: diy LC
IRG4PF50WD(1200v device) 4.89nF/4.847nF/6.983nF
IRG4PC50FD(slower 600v device) 5.05nF/5.629nF/7.35nF

so, is this MBQ fake?

my actual IGBTs


start ramp 454KHz


ramp top 431KHz. also I think I have a problem with that GDT core(material). output doesnt look good enough(?)


yellow= bridge output blue=Iprim(100A per div)


I have been also playing with the igbt gate resistor values. Im using just 1 core for gdt for the 4 fullbridge IGBT. above scopeshots was from recent 2R2(and 5819).

here is from 4R7


and 15R (what are those extra spikes again? body diode?)


thanks for the sim. 

I do have another topload. not a uniform 3.4"×9.6" because it was formed from a 4" diameter aluminum ducting. one thing I noticed using THAT torroid was, it was not that straight of a spark(compared to the smaller 2 stacked) even with oscillation above 400KHz.

Ralph

[Rafft]

David

UPDATE #1

as per your posted schematic. as I have mentioned before, MAX913. its all I have for a fast comparator. using 10R and 24uH for the filter. still have to play around with 22R 33R 47R in actual.




with ALL the parts connected, 446KHz, else 2.4MHz with just the comp and some passives. output taken from pin7. this should be at 470KHz, right? I will try smaller value for C33, 200pF - 330pF maybe. btw for clock circuits, MLCC not good? maybe get a regular ceramic cap in there?


-Ralph

[davekni]

Ralph,

so, is this MBQ fake?

Based on capacitance measurements, I'd guess fake, but can't be certain.  That's presuming the datasheet for MBQ60T65PES that you posted is accurate.  Figure 7 shows Cies (G to CE shorted) as being about 3.5nF at 0V.  Your measurement of 6.25nF is 78% high.  Other two parts you measured are about 18% and 22% above what their data sheet graphs show.  These data sheet graphs are typical values.  Thus this is not proof of counterfeit.  It is quite possible that MagnaChip couldn't meet their initial values and updated a datasheet.  Hard to tell since little information shows up on the web.  Their datasheet capacitance values and gate charge are atypically low, one of the reasons I was impressed with the part.  Don't see any obvious evidence of counterfeit in the images you posted.

ramp top 431KHz. also I think I have a problem with that GDT core(material). output doesnt look good enough(?)

My guess is scope ground lead inductance or wire inductance issues - emitter current across wiring inductance being superimposed on Vge.  I run into the same issue when scoping Vge at high currents.  Switching looks cleanest in this scope capture.

and 15R (what are those extra spikes again? body diode?)

This is exactly what happens when phase lead is not quite sufficient.  Higher gate resistance slows turn-on, so would require more phase lead to allow turn-on to occur by zero-current time.  Better to stay with lower gate R rather than increased phase lead, especially at your high frequency.  Good evidence that phase lead is about right for lower gate R.  The 4R7 traces show just the beginning hint of triple-transitions.  That is a fine level to have.  (I've posted here somewhere a couple times explanations about why marginal phase lead causes triple-transitions on bridge output.  Even less phase lead does not produce triple transitions, just hard switching spikes.)

with ALL the parts connected, 446KHz, else 2.4MHz with just the comp and some passives. output taken from pin7. this should be at 470KHz, right? I will try smaller value for C33, 200pF - 330pF maybe. btw for clock circuits, MLCC not good? maybe get a regular ceramic cap in there?

Is 446kHz with H-bridge powered, so entire feedback loop?  2.4MHz seems quite high.  What parts were connected for that.  BTW, the high duty cycle (longer positive) is likely due to comparitor's lower output voltage and higher input bias current.  Can be fixed with value changes.  If you want to experiment, it requires high local feedback amplitude (at comparitor) and low primary current feedback amplitude.  That is the change to silicon diodes and 5-7k (instead of 1k) on input, 10ohm burden resistor, and 500:1 CT ratio.

No inherent issue with MLCCs.  For timing, look for NP0 or C0G dielectric.  Those designations indicate nominally no temperature or voltage coefficient to capacitance (stable capacitance value).  Usually fairly high Q (low ESR) too.  Z5U and Y5V dielectric caps are worthless (exaggerating slightly).  X5R and X7R dielectrics have gotten worse over the years as part sizes shrink.  Some lose 90% of capacitance at rated DC voltage.  Others aren't quite that bad.  For precision, stay with NP0/C0G (or film capacitors).

Do you have a schematic for the little board you show in the last image?  Then we can discus specific part values by reference designator. 

[Rafft]

David

  I'd guess fake

kinda expecting that especially that they seem 'çheap', about 10USD for 5pcs. the other 2 IGBTs 4PF50WD(this is genuine) & 4PC50FD(bought locally YEARS ago, perhaps genuine as well(?)).

My guess is scope ground lead inductance or wire inductance issues - emitter current across wiring inductance being superimposed on Vge.  I run into the same issue when scoping Vge at high currents.  Switching looks cleanest in this scope capture.

This is exactly what happens when phase lead is not quite sufficient.  Higher gate resistance slows turn-on, so would require more phase lead to allow turn-on to occur by zero-current time.  Better to stay with lower gate R rather than increased phase lead, especially at your high frequency.  Good evidence that phase lead is about right for lower gate R.  The 4R7 traces show just the beginning hint of triple-transitions.  That is a fine level to have.  (I've posted here somewhere a couple times explanations about why marginal phase lead causes triple-transitions on bridge output.  Even less phase lead does not produce triple transitions, just hard switching spikes.)

Ooops forgot to mention, that trace was at bridge output, not on igbt gate/s. sorry.
I was also expecting a cleaner output since I have used a supposedly fast IGBT  so yeah, I will just keep gate resistors at 2R2.

Is 446kHz with H-bridge powered, so entire feedback loop?  2.4MHz seems quite high.  What parts were connected for that.  BTW, the high duty cycle (longer positive) is likely due to comparitor's lower output voltage and higher input bias current.  Can be fixed with value changes.  If you want to experiment, it requires high local feedback amplitude (at comparitor) and low primary current feedback amplitude.  That is the change to silicon diodes and 5-7k (instead of 1k) on input, 10ohm burden resistor, and 500:1 CT ratio.

446KHz is just the comparator board, not with whole system.  it oscillated at 2.4MHz with just MAX913/50K feedback/1k+1k mid on pin#2/470R+470p. once I connected the rest of the components(all in your schematic, oscillation lowered down to 446KHz(but still not installed in the coil). my feedback CT is 81:1 (9:9:1). indeed the output on pin#7 has higher on-time than on inverted output pin#8.

Do you have a schematic for the little board you show in the last image?  Then we can discus specific part values by reference designator.

lets use this for reference 



-Ralph

[Rafft]

UPDATE #2

done 50/50% duty ratio and adjustable(to some extent) frequency & set to 470KHz



edit:
with the circuit running(bench test), without CT present, I touched the 'ct terminals' with my bare hands & sure enough, output frequency is influenced(a bit)

[flyglas]

I love the idea to use the feedback comparator as self oscillation circuit. Nice idea!

[Rafft]

David

initial results:
MAX913 set to 50/50% ratio and 470KHz self-osc

yellow-bridge output blue-Iprim. wick or no-wick, doesnt matter. adding wick though lengthens the ramp-up time.





starting freq. im unsure why it never started at 470KHz. but it does start from the free-run freq because qcw doesnt ignite w/o wick.


and here is end of ramp-up. horrible waveforms.  I wonder though if the feedback did take-over. will try to lower more the MMC to 19.5nF(currently its on 16.3nF, from 14.2nF). and I need to go back to 24uH+50R for the lowpass.


Ralph

[davekni]

done 50/50% duty ratio and adjustable(to some extent) frequency & set to 470KHz

I'm guessing that you reduced R4 in your schematic to get 50% duty cycle.  That is a good tweak.  To make it less sensitive to input bias current, you could reduce values of R1 and R5 proportionately, even down to 600 and 5k.  That lowers impedance, so reduces voltage induced by bias current.  (Might then need to tweak R4, since it's value is compensating for both bias current and output voltage.)

my feedback CT is 81:1 (9:9:1).

My simulated circuit was for 500:1 CT ratio.  81:1 will swamp the self-oscillation too soon.  Reducing R6 to about 1.6 ohms would get feedback amplitude about the same with your 81:1 CT as in my simulation using 500:1 and 10 ohms.

I wonder though if the feedback did take-over. will try to lower more the MMC to 19.5nF(currently its on 16.3nF, from 14.2nF). and I need to go back to 24uH+50R for the lowpass.

I recommend staying with 16.3nF until everything is working there.  Further reduction makes upper-pole starting more difficult.
Your thought about 24uH+50R is along the right lines.  My low 0.1uF inductance was to match zero-delay H-bridge simulation.  You need phase lead still, the same amount as before, since you are using a real H-bridge.  The lack of phase lead is why waveforms look ugly.  Glad IGBTs haven't fried due to that insufficient phase lead.  However, to get the same phase lead with 1.6 ohms requires 24uH*1R6/50R=0.77uH.  If inductor resistance is significant compared to 1.6 ohms, then the 1.6 ohm resistor (1R6) needs to be reduced further so that total resistance of L1+R6 is 1.6 ohms.

starting freq. im unsure why it never started at 470KHz. but it does start from the free-run freq because qcw doesnt ignite w/o wick.

Insufficient phase lead may explain this too.  I'm not certain.

Ooops forgot to mention, that trace was at bridge output, not on igbt gate/s. sorry.

I forgot to look at scope scale too, so didn't guess correctly.  Likely cause of the curved/sloped tops and bottoms is the same - parasitic wiring and/or scope ground lead inductance.

kinda expecting that especially that they seem 'çheap', about 10USD for 5pcs. the other 2 IGBTs 4PF50WD(this is genuine) & 4PC50FD(bought locally YEARS ago, perhaps genuine as well(?)).

Noticed one other think that points to fake:  Spec thermal resistance values suggest a diode die about 25% of the area of the IGBT die.  Picture appears to have a diode over 50% of IGBT area.  That die size ratio would fit the slower MagnaChip part MBQ50T65.  Capacitance measurements also fit MBQ50T65.  Since there are no worst-case AC parameters specified, there is no way to prove counterfeit to get your money back.  (I was surprised how many IGBT and FET specifications have worst-case values for only DC parameters, and just typical values for AC parameters such as switching times and losses, gate capacitance and charge.  Saves cost in production test.  DC parameters are easier to measure.  Makes counterfeiting easy, however.  Designing fast IGBTs requires trade-offs with DC parameters.  It is easy to make a fake part that meets guaranteed DC values but is much slower.)

I love the idea to use the feedback comparator as self oscillation circuit. Nice idea!

Thank you for the compliment.  All my coils are self-oscillating.  Two use this circuit idea, as do many of my other (non-Tesla coil) resonant H-bridge circuits.  When I found this forum 3 years ago, was surprised to find that none of the standard UD1/2 drivers were self-oscillating.

Good luck experimenting!  I'm thrilled to see upper-pole starting tested this way.  Satisfies my curiosity without needing to change my future QCW design plans.

[Rafft]

0.78uH & 1R8 2watt latest mod only



ramp start.sometimes it starts at lower freq like 438KHz.


end of ramp-up.concered of those 100v spikes though


did another zoomed-in capture of ramp-up end(to check Iprim=blue)


spark length are short. MMC value perhaps? AFAIR  10.7uH/14nF had the best spark length(feedback CT, no self-osc). compared to 16.5nF

I was thinking maybe use 24uH/50R for lowpass and a resistor voltage divider after D1&D2.. and/or change C2 to something smaller like 1nF. phase lead still terrible

AND

I think I will change R1 to 600R(as you suggested) for a stronger free-running start-up.


edit#2
I have been playing with the sim just to confirm parts value interaction. and sure enough THEY DO by quiet a LOT! and is a bit finicky to tune... my main problem for today is HOW I would like my feedback to not be strong enough during start-up. here on the waveform can be seen the taking of feedback(430KHz) w/c I have set it on a 50uS delay(on sim sine generator) before going to the comparator(w/c Ive set the freq HIGHER, so it can bee seen clearly on plot). here for sim schematic, Ive taken from UD2.1b. btw , this is a realy elegant solution for having a free-running + feedback switchover.

[davekni]

0.78uH & 1R8 2watt latest mod only

Not sure why, but this image is not showing up.  Just a blank rectangle where the image belongs.  Just tried opening image in a new tab and then it shows up.  Perhaps some issue with the place you are hosting images.  You can post JPG format images directly to this forum.  That is what I do.  Your method is generally working, so no pressure to change.

end of ramp-up.concered of those 100v spikes though

Yes, those are concerning.  Looks like too much phase lead.  However, hard to tell given noise in current trace.  Did something change in the way you are scoping current?  Or in H-bridge output scoping (ie. how scope ground is connected to bridge)?

did another zoomed-in capture of ramp-up end(to check Iprim=blue)

Lower voltage, current, and spikes in this capture.  Still guessing a bit too much phase-lead.  (Was this a slower ramp, or a capture earlier in the ramp when voltage was lower?)

I was thinking maybe use 24uH/50R for lowpass and a resistor voltage divider after D1&D2.. and/or change C2 to something smaller like 1nF. phase lead still terrible

Voltage divider is a fine idea, but needs to be before D1&D2.  Otherwise diodes will clamp feedback signal too low to ever overcome self-oscillation signal.

I have been playing with the sim just to confirm parts value interaction.

Great!  Simulation is very helpful.  However, I'd suggest staying with the topology where AC coupling (C2 in your sketch, C3 in latest simulation) is after clamp diodes.  With this latest simulation schematic, if 1k and/or 470R values need to change to hit 50% duty cycle, both dividers need to change together, adding complexity.

I need to update my simulation too.  Not enough time today.  Realized that my shortcut of simulating a zero-delay H-bridge and almost-no lead compensation misses some details.  For one, phase-lead inductance increases feedback voltage (increases impedance of R+L burden on CT), so causes feedback to kick in sooner.  Separately, I didn't look at phase lead during startup and transition.  To get phase lead at startup, self-oscillation frequency needs to be set slightly higher than upper pole frequency.

In case it is of interest, here's my zipped ltspice source for my previously-posted simulation.  Not updated yet with simulated H-bridge delay and corresponding phase lead:
sstc_ac1.zip

[Rafft]

David

this is just a quick reply. will update later.

did you say 'phase lead too much'? hmmmn havent thought of that. indeed the Iprim trace has 'shifted' more to the right(of the bridge output-yellow trace).

refering to my sketch, I will make a guess its R5(6k) that adds MORE  phase delay to this. how about we make lower like 1k?

I know for sure this will again alter the main comparator self-osc freq and therefore need readjustment.

can you recalculate that part with 0.77uH/1R9 and 81:1 ratio? with using 1k for R5.


================================

Yes, those are concerning.  Looks like too much phase lead.  However, hard to tell given noise in current trace.  Did something change in the way you are scoping current?  Or in H-bridge output scoping (ie. how scope ground is connected to bridge)?

nothing is changed, same all connections..same probe, same scope settings etc. the only thing connected to mains is the scope. my coil is battery-powered. and is it too much? or lack of phase lead though.. maybe its time you include a phase lead in the sims  me im not sure how Id do that in sim.

Lower voltage, current, and spikes in this capture.  Still guessing a bit too much phase-lead.  (Was this a slower ramp, or a capture earlier in the ramp when voltage was lower?)

yes lower bridge output. different capture. I just wanted to see another capture of it, Primary Current.

 

Voltage divider is a fine idea, but needs to be before D1&D2.  Otherwise diodes will clamp feedback signal too low to ever overcome self-oscillation signal.

based on your schematic, changing R5 value(6k) also influences the free-running comparator oscillation. as I have mentioned before, MAX913 R1 R2 C1 R3 R4 alone oscillates on 2.4MHz. adding the rest of the components C2 D1 D2 R5 L1 R6 =  lowers the comparator freq (little board only, not connected to coil) now I will re=check little board AND ADD the CT, and check again the free-running freq(still not connected to coil, but output#7 #8 only)

 

I need to update my simulation too.  Not enough time today.  Realized that my shortcut of simulating a zero-delay H-bridge and almost-no lead compensation misses some details.

yes, if at all possible, good to include phase delay(lead)

To get phase lead at startup, self-oscillation frequency needs to be set slightly higher than upper pole frequency.

by how much higher? my 1st test was on 470KHz, now I removed little board again and did some more mod, and adjusted freq to 485KHz.. would that be enough for higher-than-upper-pole? or maybe go 500KHz?

[davekni]

With delay and phase-lead added, simulation is starting at upper pole even with original 24uH+50R CT burden and 81:1 CT ratio.  Simulation attached at end, and image here:



This is with 270ns simulated gate-driver and H-bridge total delay.  That's what looked about right for 24uH+50R.  Not sure how well that matches reality for your circuit.  I haven't had time to play with values much.  Increasing self-oscillation frequency didn't seem to matter.  (I slightly reduced self-oscillation strength by dropping from 6k to 5k.  That didn't seem to matter either.)

nothing is changed, same all connections..same probe, same scope settings etc. the only thing connected to mains is the scope. my coil is battery-powered. and is it too much?

Hmm.  Increased current noise still seems unexpected.  Could you try scoping the other H-bridge output to see if it matches (opposite phase of course)?  I'm wondering if something partially-fried or whatever, some explanation for the noise increase.  Could you run a coax cable from scope-measurement CT back to scope to minimize noise coupling?

did you say 'phase lead too much'? hmmmn havent thought of that. indeed the Iprim trace has 'shifted' more to the right(of the bridge output-yellow trace).

Yes, that is the shift I was seeing.  Also, the spike makes sense for switch-off at higher current (farther before zero-current).  Parasitic inductance of H-bridge (even just what is inside IGBT leads) will cause a voltage spike as current suddenly transitions from lower IGBT to upper IGBT (and visa versa).

udx2.zip

[Rafft]

David

Thanks again for the simulation

This is with 270ns simulated gate-driver and H-bridge total delay.  That's what looked about right for 24uH+50R.  Not sure how well that matches reality for your circuit.  I haven't had time to play with values much.  Increasing self-oscillation frequency didn't seem to matter.  (I slightly reduced self-oscillation strength by dropping from 6k to 5k.  That didn't seem to matter either.)

Im using 510+510 for R5, so had to retweak R2 C1, for 480KHz & R3 R4 for 50% duty

Hmm.  Increased current noise still seems unexpected.  Could you try scoping the other H-bridge output to see if it matches (opposite phase of course)?  I'm wondering if something partially-fried or whatever, some explanation for the noise increase.  Could you run a coax cable from scope-measurement CT back to scope to minimize noise coupling?

kindly check attached. Im not sure if there is a PC app that can open this. before 'going back to stock circuit', I checked my IGBTs(diode check), all are ok. I dont have a coax cable for I prim, but will make some. need to buy sheilded wire and plug/jack.

Yes, that is the shift I was seeing.  Also, the spike makes sense for switch-off at higher current (farther before zero-current).  Parasitic inductance of H-bridge (even just what is inside IGBT leads) will cause a voltage spike as current suddenly transitions from lower IGBT to upper IGBT (and visa versa).

again, pls check waveforms. pls tell me if it is indeed too much phase lead(how does your 10R/0.1uH compare? sgould I use this values instead?).

going back to my stock is using 22-24uH/51R. our mini-board is 0.77uH/1R9 (and R5=1k). what to do next? I think I see a small difference with using R5=1k(current ckt config) versus R5=6k . R5-1k looks like Iprim cleaned up a bit.

meanwhile here are scope waveforms:
NewFile1 = with the self-oscillationg, R5=1k 1R9/0.77uH and re-tweaked value R2 C1, changing R5 value affects oscillating freq of MAX913
NewFile2 = went back to feedback only , still same 16.5nF on PRIMARY
power supply used for testing both , is 50Vdc
still the same wiring, nothing changed



edit:
oh sh.. I made a mistake on my little board! R5 was connected between L1 and R6! fixxing error

[Rafft]

yet another (corrected)UPDATE 

R5 - 6k
L1&R6 - 0.77uH 1R9(recalculated from 24uH 50R)
free-running 485KHz
MMC 16.5nF

it finally cleaned up(Iprim). my initial question, did it work? (freq takeover)

does L1 R6 value need re-adjusting?  final tweaks?

kindly check attachments. shown are start of ramp, ramp-up peak, and overall ramp

again, same probe hookup

[davekni]

ZIP files didn't open on my computer.  Images are coming through fine.  Yes, scope traces look good now.  (Of course, sometimes things change at full voltage, especially the optimum amount of phase lead.  IGBT delay depends on voltage and current to some degree.)

Appears to be working well.  If I recall correctly, lower pole was ~330kHz.  So I think you are starting at upper pole as intended.

[Rafft]

hi David

they are RAR files. just renamed to ZIP so I could attach it here. inside is .wfm file(about 6Mb). from scope. I think tektronix has a pc app to open that kind of file. there you could zoom-in on the scope waveform. and way better than just a screen grab.
-dont mind downloading PC apps for viewer, they do not work!


I guess the 0.7uH/1R9 is sufficient enough phase lead? will check later with higher bus voltage

-Ralph

edit#1
scope shots with using its power supply, only about 150vdc bus(for testing). spikes on output are manageable?

AND sparks are MORE straighter than before. im not maxxing out my ramp. just wanted a clean ramp-up(and down). cranking it more results in more longer sparks as well. ramp starts from around 470KHz(from a set 485KHz) down to 416KHz. thats a big freq difference, right? I will try changing MMC values higher. lets see if I can bring it to lower pole. but in my previous attempts (w/o the phase lead working) I could not bring it to oscillate in the lower pole.. it seems its getting 'pulled' by the free-running osc to be at upper pole 430KHz..

[davekni]

they are RAR files. just renamed to ZIP so I could attach it here. inside is .wfm file(about 6Mb). from scope. I think tektronix has a pc app to open that kind of file. there you could zoom-in on the scope waveform. and way better than just a screen grab.
-dont mind downloading PC apps for viewer, they do not work!

Now that I know what they are, I likely could decode them.  There are two or three different versions of Tek's .wfm file format over the years.  The PC app for the oldest version doesn't appear to be available any more.  However, I've written a few PERL scripts to decode at least a few specific cases.  That's at my work, not here at home, so would be a bit of nuisance to test on your files.  Don't see any need at this point.  Your images are fine for communicating here.

I guess the 0.7uH/1R9 is sufficient enough phase lead? will check later with higher bus voltage

Looks good on your scope plots.  I think spikes are reasonable.  Might be a bit lower spikes with a bit less phase lead.  However, I suggest not changing.  Much more likely to fry IGBTs from too little phase lead than too much.  It is better to stay with slightly too much phase lead to handle increased IGBT delay when they get hot.

AND sparks are MORE straighter than before. im not maxxing out my ramp. just wanted a clean ramp-up(and down). cranking it more results in more longer sparks as well. ramp starts from around 470KHz(from a set 485KHz) down to 416KHz. thats a big freq difference, right? I will try changing MMC values higher. lets see if I can bring it to lower pole. but in my previous attempts (w/o the phase lead working) I could not bring it to oscillate in the lower pole.. it seems its getting 'pulled' by the free-running osc to be at upper pole 430KHz..

Your start-of-ramp scope capture at 470kHz may be before feedback has taken over fully.  I suspect it gets to your measured 430kHz upper-pole frequency at a bit higher Vbus level, then drops to 416kHz due to arc loading.  I think this is all fine and as expected.  Yes, try larger MMC values to see how they perform and if this driver circuit still manages to start at upper pole.  Of course, upper pole frequency will drop some as MMC increases.  So it is possible that you would need to drop self-oscillation frequency a bit to match.  Also possible that no tweak is needed.

Have fun experimenting!  Great to see that this self-oscillation concept is working in practice, at least so far.

[Rafft]

David

Might be a bit lower spikes with a bit less phase lead.

less phase lead,  meaning a slightly higher value for 0.77uH? or lesser? maybe something like an adjustable 0.1uH - 2uH?

currently the inductor is a basic(no plastic former) coil with a ferrite slug inside. had to adjust it close to 0.77uH(and glue it) prior to installing on mini board. Im planning to make another board for this, with a better adjustable inductor(plastic former). Dont worry about frying IGBTs. Ive even ran it without the phase lead.  I always test with lower Vbus.

Your start-of-ramp scope capture at 470kHz may be before feedback has taken over fully.  I suspect it gets to your measured 430kHz upper-pole frequency at a bit higher Vbus level, then drops to 416kHz due to arc loading.  I think this is all fine and as expected.

I think its already working as intended. I do have to experiment with the tuning (MMC) more. maybe get back to using higher-coupling primary and the bigger topload.

Have fun experimenting!  Great to see that this self-oscillation concept is working in practice, at least so far.

Couldnt have done it without you. once again, Thanks David.

I think this UD2.1b should be named UD2.1R (with free-running osc) 

cheers
-Ralph

[davekni]

less phase lead,  meaning a slightly higher value for 0.77uH? or lesser? maybe something like an adjustable 0.1uH - 2uH?

Less phase lead is lower inductance, perhaps 0.6 to 0.7uH.  However, be careful.  Less may look great in initial testing.  Once you run your coil harder (rapid firing at max voltage), IGBTs get warmer and slow down some.  Slower IGBTs need more phase lead to compensate.  You could test by heating your H-bridge heat sink with a heat-gun (or whatever) to 100C or a bit more, then run short bursts to measure phase lead.

Couldnt have done it without you. once again, Thanks David.

I think this UD2.1b should be named UD2.1R (with free-running osc) 

You are certainly welcome!  Please keep us all posted with your progress.  And, thank you for trying this experiment.

[Rafft]

Hi again

I have been playing with higher coupling+slightly bigger topload versus lower coupling+2stacked smaller toploads.

the goal really for the topload was the 2 stacked ones give me straighter sparks compared to the single bigger topload(slighly bendy and branchy even at 430kHz). Im not going back using the bigger one, many times Ive tried swapping that in but no straight sparks.

here are the results

lower coupling


higher coupling. and I hit my OCD limit of around 120A peak.


I already know that with higher coupling will farther the lower and upper poles. will also raise the resonant freq.

going back to the shots, it is seen that with higher coupling result in faster primary current rise.

Im not at all so much concerned about tuning(maybe?) because If I wanted to get longer arks, I just had to add more  ramp length (and Vbus as well) BY lowering C=14nF and L around 11uH or so(basically making it Low inductance)..

but I believe Im still missing something else...

primary coil impedance?

detuning?

can someone pls explain this in a more laymans term way 


edit:
I just ordered yet another cheapER IGBT. 40T60(ANF)
cancelled it. im unsure if this IGBT has the version WITHOUT body diode(?) no wonder they where so cheap(?) Im also unsure if the ANF should be separated from the part# upper part(?)


and if ever I go for full bridge 8 IGBTs(double each switch). will a single GDT be enough to drive them all? I was thinking of using a ferrite E-I or EE core (no gap of coarse).

I also have this circuit built a few years ago. dick smith lopt/fbt tester. would this suffice in testing gdt for inductance check?


thanks

[davekni]

higher coupling. and I hit my OCD limit of around 120A peak.

going back to the shots, it is seen that with higher coupling result in faster primary current rise.

Have you zoomed in to see if you are still running at upper pole frequency?  Any QCW coil owners have insight on expected effect of coupling on primary current?  I could imagine a small current increase since upper-pole frequency increases and capacitances remain the same.  However, arc voltages tend to drop as frequency increases, so the net effect might be minimal.

I'm not clear on your remaining questions.  Tuning is important.  Larger difference between primary and secondary frequencies will result in higher primary current.  (More current is required to get power to secondary.)

and if ever I go for full bridge 8 IGBTs(double each switch). will a single GDT be enough to drive them all? I was thinking of using a ferrite E-I or EE core (no gap of coarse).

Yes, a single GDT is fine.  Construction technique (minimizing parasitic inductance in GDT and leads) is important, especially at your 430kHz frequency.  E-I or EE cores work fine for GDTs too, with no gap as you know.

[Rafft]

David

yes coil starts at upper pole around 475KHz then goes lower.
I have set back my primary to be 11uH/14nF(just like before w/o the 'upper pole initial freq') w/c has given me  longer spark output.

Im not sure why but when I go higher coupling, it likes more primary current( even at a lower Vbus) compared with the lower coupling. it should have been the opposite(higher coupling lower Pri current), or so ive read on other posts.

I was hoping to understand more about 'detuning' and if would help in any way with my coil setup

anyways ill UP my ocd to maybe something like 150Apk and see if nothing explodes.

regarding impedance, this is the topic im talking about. sorry if I havent posted link before. and Im talking the sweet spot of 10uH/15nF for MY COIL. if it makes sense.

https://highvoltageforum.net/index.php?topic=113.0

[davekni]

yes coil starts at upper pole around 475KHz then goes lower.

Have you checked how much lower the frequency goes?  In light of Uspring's latest post to the other thread:
 https://highvoltageforum.net/index.php?topic=113.msg15012#msg15012
I wonder if operation is jumping to lower pole as upper pole gets too damped for ZCS.  However, higher coupling should (I think) help maintain ZCS at upper pole for any given arc load.

Other than that, I'm out of ideas here.  Higher current at higher coupling is not what I'd expect.  I don't have enough relevant experience to guess at any other possible reasons for your results.

Have fun experimenting!  Hope you end up with performance you are satisfied with.

[davekni]

In case it is of interest, I purchased some "MBQ60T65PES" parts from two different EBay suppliers.  Both were counterfeit, but not exactly the same.  They were roughly half-capacitance.  Vce reached worst-case voltage at ~30A rather than specified 60A.  Even lower current for spec'ed typical Vce.  Got refunds from both suppliers, one immediately and the other more reluctantly.

[Rafft]

too bad they are fake too. glad that it was refunded

thats why I dont overdrive much my 60T65PES 

anyways I ordered(yet another) 10pcs because they where sooo cheap.
used my yellow LCR meter to check my un-used 60T65PES(for comparison) with these cheaper igbt. checked datasheet and gate capacitance seems close to the LCR reading. its only 80A@25degC so I will 'double switch' this for a full bridge and drive this to 150Apk. ran out of igbt heatpads, ordered them. will arrive 2 weeks tops.

will see if it holds up!

just curious, how did you test it to ~30A only? id imagine you also 'heat up' the igbt prior to testing

[davekni]

just curious, how did you test it to ~30A only? id imagine you also 'heat up' the igbt prior to testing

Test setup was nothing like normal IGBT use.  I biased Vge at 15Vdc (permanently on).  Discharged a 10.0uF capacitor through an inductor across collector-emitter.  Measured forward and reverse (diode) Vce during this L/C ring-down waveform.  That way Vce is never large, so doesn't saturate scope input amplifier when set to 1V/division.  Current measured by dV/dt across 10.0uF capacitor (actually two 5uF MKP capacitors in parallel that measure 10.00uF together).  Used four-wire connection to IGBTs to avoid connection resistance.  Testing is with single ring-down waveforms spaced out in time, so very little heating.  Current is adjusted by adjusting initial 10.0uF capacitor voltage.

(I already have circuitry around, both FET and TRIAC based, for discharging L+C circuits.  Mostly use these for testing inductors up to saturation current.  Can be done by manually touching two wires together.  That often requires several tries for each test to get a clean switch event without contact-bounce.)

For gate (input) capacitance, I connected collector and emitter to a 25Vdc supply and measured gate-to-emitter capacitance.  Includes gate-to-collector capacitance too since Vce is fixed.  Spec values are at 25Vce for this part.  Graph shows input capacitance as a function of Vce, so you can see what it is supposed to be (roughly) at 0Vce from the graph.

Your parts are measuring high input capacitance, opposite what my counterfeit parts measured.  There is a slower Magnachip part with higher capacitance, so that may be what your parts are.  Mine are likely some other non-Magnachip smaller die inside.  No idea how fast they switch.  That is harder to test.

[Rafft]

David

thanks for the test setup explanation.

Ive measured BT40T60 with Vce 21vdc(current limited) and with gate discharged.

attached also is screenshot of graph. is this the one to look up to? then 5nF is too high @21Vdc?

edit:
tested spare magna igbt. its about 4nF @21Vdc. so I guess the magna was still a bit better then? is 4nF way off with a genuine 60T65PES Cies=2.3nF?

[davekni]

Ive measured BT40T60 with Vce 21vdc(current limited) and with gate discharged.

Not sure I'm understanding the bit about "current limited".  Usually, with 0Vge DC and a small AC voltage for capacitance measurement, there should be almost no collector current.  Usually <1ua.  Perhaps your meter is applying a larger voltage (and perhaps a voltage with a DC component) for capacitance measurement.  You could try testing with a larger (~0.1 to 1uF) capacitor in series with the meter lead, and a high-value resistor (1meg to 10meg) gate-to-emitter to force average Vge to be 0V.  Otherwise, I'm inclined to believe only the gate to emitter/collector shorted readings.

[Rafft]

current regulated. its my CV CC power supply. I set it to 2mA only.

I have re-checked both igbts 60T65PES and 40T60 with 100nF(94nF actual) and 10meg

[davekni]

Both still well higher than the 2327pF typical specification.  (Mine were ~1200pF, so lower rather than higher.)  That would roughly fit for the similar but slower Magnachip part.

Even 2mA shouldn't limit when feeding Vce at 0Vge. Either these IGBTs have high leakage current or gate swing from cap measurement is large or 2mA current limit is not precise.

[Rafft]

yes I agree, that gate capacitance is high as per datasheet.

the 2mA setting in PS was to somehow protect the LC meter(previous test w/o 100nF&10meg).

anyways, I had a full refund with that 40T60. 1 of the 10 igbt has zero capacitance reading gate/emitter(Vce @21vdc). im guessing factory defect? or because it was sent w/o an anti-static packaging

also, Ive used a bnc cable connection for checking Iprim. its somewhat clean on scope(ch2), but gets spikes when I  make connection with fullbridge output(ch1)

[davekni]

the 2mA setting in PS was to somehow protect the LC meter(previous test w/o 100nF&10meg).

Meter may have internal protection for applied voltage, to handle charged capacitors.  Doesn't hurt to be safe.
Have you verified that your PS is accurate at 2mA?  That is very low for most supplies.  Could easily be the supply not going above 21V by itself at such low current setting.

1 of the 10 igbt has zero capacitance reading gate/emitter(Vce @21vdc). im guessing factory defect? or because it was sent w/o an anti-static packaging

Likely a factory defect - bad bond wire.  Those are really cheap counterfeit parts to have such a gross defect.  Would not be caused by ESD.  Unless your meter reads 0 capacitance into a short circuit or low resistance.  (Most meters read over-range for short circuit capacitors.)  ESD damage is generally gate oxide punch-through causing gate-emitter conductivity.  (For SiC FETs, I've seen ~1ohm G-S after gate over-voltage event.  Still works as a FET if short Vgs pulses are applied with sufficient current to drive 1ohm load.)

also, Ive used a bnc cable connection for checking Iprim. its somewhat clean on scope(ch2), but gets spikes when I  make connection with fullbridge output(ch1)

Sounds like an issue with either internal scope crosstalk due to ground current or effects of imperfect shielding on coax.  Either cause can be improved by ferrites around the coax cable and/or scope probe cable (cable from scope to probe body).  Large ferrites with multiple turns are best.  I often use several turns around a ferrite E55 core set on my scope probes and/or coax cables even when using my fancy (but old) Tektronix TDS754.

[Rafft]

the (kit)power supply I am using:
https://www.qsl.net/z33t/dc_0-30v_0-3A_eng.html

so I just set the 'current' pot to minimum, and checked if there is output voltage.yes there is.

ok my igbts are counterfiet, except the IRG im using in the buck switch.

my lcr meter shows. 0.00 reading with open leads. ill try using ferrites for the coax then. thanks

[Rafft]

Sounds like an issue with either internal scope crosstalk due to ground current or effects of imperfect shielding on coax.   

David, I tried 

[davekni]

David, I tried 

Looks like that is working well.

so I just set the 'current' pot to minimum, and checked if there is output voltage.yes there is.

Current limit accuracy (including temperature drift and change with output voltage) is going to be poor down at 2mA.  True for almost any supply capable of 1amp or more.  That power supply circuit powers U1 and other circuitry after current sense resistor R7 (0.47 ohms).  Also, 2mA is only 1mV across R7, well under sensing opamp U3's worst-case input offset voltage specification (+-6mV).

To test leakage current, I'd suggest feeding collector through 10k resistor.  Measure voltage before and after resistor to see how much current the IGBT is drawing.  If more than 40uA, that is a clear violation of a worst-case specified limit, so proof of counterfeit (or damaged) parts.  (Spec. is 40uA at 650V at 25degreesC.)

[Rafft]

Hi David

kindly check my scope shots. should I be worried about those negative going spikes? and why only on the negative side? and also, they only show when I change to shorter scope timebase. scope error?

Im now using 8pc IGBTs. BT40T60.  is it still ok to use 2R2 gate res for each IGBT? or should I raise it something like 5R since they are now connected as parallel.

[davekni]

kindly check my scope shots. should I be worried about those negative going spikes? and why only on the negative side? and also, they only show when I change to shorter scope timebase. scope error?

Negative spikes are within reason.  Only negative because duty cycle is under 50%, longer low time than high time.  Bit hard to tell without zooming in even farther, but the rising edges appear to have a bit of the triple-transition typical when phase-lead is slightly less than ideal.  Falling edges are earlier (farther before current zero-crossing) due to duty-cycle (shorter high time), so have more phase lead.  I think the spikes are due to slightly too much phase-lead on falling edges.  Too much is better than too little.  Spikes aren't large enough to be an issue.
Is the scope acquisition in hi-res mode?  With long timebase, sample rate is low enough to miss spikes or average them away if in hi-res mode.  If you change to peak-detect mode, the spikes should show up.

Im now using 8pc IGBTs. BT40T60.  is it still ok to use 2R2 gate res for each IGBT? or should I raise it something like 5R since they are now connected as parallel.

2R2 is lower than typical for TO247 IGBTs, whether single or paralleled.  However, if I am seeing triple-transitions correctly, that suggests dead-time is sufficient.  (BTW, are there a diode across each of the eight 2R2 resistors?  It is hard to tell in the picture.)

[Rafft]

David

yes each IGBT has its own (sb) diode. and about the ringing on fullbridge output, ive tried from the current 2R2(each igbt) to using 10R(and SB260 diode) 30v tvs as I ran out of 20v ones.bridge ourput ringing seems  to have gone down. also added more cap decoupling on the fullbridge bus rail(around 13uF).

primary LC is 17uH/11nF. im using a newer secondary 4" x 6.2" #30(about 18mH). I also noticed that using a bigger top load makes a great sounding thump than using smaller topload(harsh sounding).

I have a couple scope shots run on 230Vdc bus and Iprim=100A. I see ringing. im not sure where this comes from(already readjusted duty to 50% on self-osc). Im guessing its the IGBT(?)

spark length around 26" or so 230Vdc@100A. still needs some more tuning

[Rafft]

to be more complete, I redid the scoping on 'peak' acquisition and did some more zooming
------
and as compared to a 'high-res' mode

[davekni]

These latest scope plots show mostly lagging phase (voltage switching after current zero-crossing).  Not sure what changed to cause this.  Previous plots show sufficient phase lead.

The glitches you zoomed into are clearly a problem.  Coil control circuit (modified UD2.1) is getting confused.  Did some ground connection come loose?  Some switching noise from buck converter coupling into coil driver circuit?  IGBTs appear to be surviving these occasional instances of confused gate drive control even though they are causing hard switching.  Don't know if these issues are related to whatever is causing lagging phase instead of desired slight-leading phase.

If nothing else, you are getting good practice in using different scope acquisition modes:)

[Rafft]

Hi David

testing from 4igbt to 8igbt inverter , I did modify the (Gao)buck to use discrete high-side driver, for more repairability than using UCC chip(from free samples few years ago). its all working as expected.

next was the self-osc mod. I added IC socket for adding parallel cap for freq adjustment. and used cheap ceramic ones. they have thin component legs(wiggled quiet easily), Im sure this is the reason why I see those hard-switched outputs. so I just soldered it (parallel cap) direct to board. one thing I also noticed was 'self-osc' should be far from the running freq, else I get dirty Iprim. Ive set my self-osc to 440KHz and my running freq is 370KHz.  my older setup with  self-osc=390KHz and running freq=370KHz resulted in ugly Iprim. anyways, Ive set my ocd limit to around 1.55V(155Apk). Ive tested 250Vdc bus and got good spark length. and there is always that triple transition I have also been trying a few coupling K.

spark length around 26-27" 250Vbus  around 130Apk 370KHz

btw, I have already killed 2 scope probes(open ckt). RIGOL PVP2150 on scoping the bridge output. is there an easy hack to converting this to something like x100 (probe in x10)?

addtl (last 2 attachments):
modified the ramp gen to do this(branching at tip), but only in 115Vdc bus:

[davekni]

and there is always that triple transition

Those look fine.  A slight reversal at the end of the main transition is not problematic.  Indicates you have sufficient dead-time and barely enough phase-lead.

spark length around 26-27" 250Vbus  around 130Apk 370KHz

Looks great!  Nice progress.

btw, I have already killed 2 scope probes(open ckt). RIGOL PVP2150 on scoping the bridge output. is there an easy hack to converting this to something like x100 (probe in x10)?

All scope probes have reduced maximum voltage capability as frequency increases.  Some (especially cheap probes) drop dramatically at TC frequencies.  Good probes (and some cheap ones) include max-voltage vs frequency curves in specifications or user manual.

For many purposes you can get away with a small capacitor (of sufficient voltage rating) in series with the probe tip, value ~1pF.  I often home-make such capacitors with just two twisted insulated wires, or a bit of heat-shrink or tape over the probe tip with a bit of copper foil over that.  This foil version is less capable for high voltages (>1kV), but much better for fidelity.  Avoids most stray signal pickup.  Any random capacitance from other nodes to probe end of ~1pF capacitor will add unwanted signal.

[Rafft]

Those look fine.  A slight reversal at the end of the main transition is not problematic.

Thats great!

Indicates you have sufficient dead-time and barely enough phase-lead.

with "barely enough", does this mean I need a slight component adjustment on the lowpass filter? ~1uH perhaps since my coil works better at lower frequency (370KHz). and perhaps why this is so, because of said IGBT being THAT slow? based on datasheet for 40T60, its calculated speed is said to be ~270KHz. when changing primary MMC from 13nF(Fres=370KHz) to 8.2nF(Fres=430KHz), the triple transitions have higher swing and Iprim more noise. so I take it as a hint that it could either be lacking on the lowpass filter, or slow IGBT (or maybe im wrong?)

Looks great!  Nice progress.

Thanks. goal is 1Meter for this battery-powered

All scope probes have reduced maximum voltage capability as frequency increases.  Some (especially cheap probes) drop dramatically at TC frequencies.

good info! seem to have forgotten about the freq response. but is ~400KHz too much for a 100MHz rated probe?

For many purposes you can get away with a small capacitor (of sufficient voltage rating) in series with the probe tip, value ~1pF.  I often home-make such capacitors with just two twisted insulated wires

I would go this route. gimmick capacitor. but probe switch set to x1 only? how do I set this up on scope?  and could I get an accurate voltage reading (for TC use) of these?

[davekni]

with "barely enough", does this mean I need a slight component adjustment on the lowpass filter? ~1uH perhaps since my coil works better at lower frequency (370KHz). and perhaps why this is so, because of said IGBT being THAT slow? based on datasheet for 40T60, its calculated speed is said to be ~270KHz. when changing primary MMC from 13nF(Fres=370KHz) to 8.2nF(Fres=430KHz), the triple transitions have higher swing and Iprim more noise. so I take it as a hint that it could either be lacking on the lowpass filter, or slow IGBT (or maybe im wrong?)

"Barely enough" means it is OK as is, but not much margin for IGBT temperature to increase or any other change that increases delay.
The standard UD2.7 phase-lead circuit is high-pass, not low-pass.  Is that what you are referring to?  What values do you have for inductance and feedback CT resistance?  If I recall correctly, you were close to the practical limit for amount of phase lead.  May be hard to get much more.  If impedance of inductor at operating frequency is already 2x or more of resistor value, it may not make sense to go farther (higher inductance).  Perhaps a smaller resistance would be better.  I don't know if anyone has explored the practical limit to phase-lead.  Theoretical limit is 0 resistance (90 degrees of phase lead).  Real inductor will have some non-zero resistance.  Might work with only inductance (and it's bit of internal resistance).  Haven't heard of anyone going to that extreme.

good info! seem to have forgotten about the freq response. but is ~400KHz too much for a 100MHz rated probe?

Just looked at specifications for my cheap 100x 2kV 100MHz probe.  It is rated for 300Vpeak at 400kHz, only 15% of it's 2kV low-frequency capability.

I would go this route. gimmick capacitor. but probe switch set to x1 only? how do I set this up on scope?  and could I get an accurate voltage reading (for TC use) of these?

I've always used 10x probes with series capacitors.  Most probes have much better frequency response at 10x than at 1x (if they have a 1x option).  It is difficult to hit exactly 100x, but not too hard to measure actual attenuation.  Measure a somewhat-lower voltage signal directly with another probe and simultaneously with the series-capacitor probe.  Calculate the ratio.

[Rafft]

Thank you for the replies David

regarding UD, I will just leave it then. and Im using counterfeit IGBTs. who knows what the real specs of these are. I expect the worst.

cheaper scope probes voltage rating dimish fast even at 400KHz only . wow.  so I will just have to use x10 and ~2pF for monitoring bridge output. and x10 as well for Iprim scoping

[davekni]

Thank you for the replies David

You are certainly welcome.  Great to see your progress.

[Rafft]

David

before, Ive had smps power for UD and buck control circuitry(Gao), power input from LiPo battery.

I have tried using Linear PS for the UD. IGBT driver ps is from 7824 regulator.

on the ramp buck side, Ive got the IGBT high-side driver in un-regulated 25vdc(17Vac traffo).

can you please check attached. where are those hashes/spikes coming from? previously, same problem with using smps power for UD and buck control.

edit: I was thinking maybe I would get 'cleaner' system with Linear PS for the logics and drivers.


the only smps power left, is the boost (16v8 to 300Vdc for buck power)

[davekni]

can you please check attached. where are those hashes/spikes coming from? previously, same problem with using smps power for UD and buck control. though what Ive noticed is that the Iprim current waveform is cleaner when there are no spikes present(with using linear PS)

the only smps power left, is the boost (16v8 to 300Vdc for buck power)

It appears that the UD2.x driver is getting momentarily confused at times (as before), switching away from zero-current point (hard switching).  Would be a bit easier to tell for sure if either a GDT input or H-bridge output were scoped too.

Why UD2.x is getting confused is harder to determine.  Could be interference from 16-300V converter or from ramp generator buck converter or from some other source.  General techniques to reduce interference include ground planes within circuitry, shielding around circuitry, and common-mode chokes (ferrites) around cables between modules.  Sometimes it is easier and more effective to make such general interference reduction changes than to determine the exact cause.

It still may be useful to find the cause.  Perhaps the 16-300V converter or ramp generator is also misbehaving momentarily, generating large noise spikes that couple into UD2.x driver.  In that case it is helpful to locate and fix the initiating noise source.  You may want to scope a switching node in each converter simultaneously with primary coil current to look for any correlation between current noise spikes and other switching events.

Good luck!

[Rafft]

David

as per your reply, here are some images and scopeshots.
blue=primary current , 1v=100amp
yellow=Full-bridge output, scope x1 , probe x10 with 2pF in series
note=I made a new GDT here. taken from VGA cables, making sure they both look the same. they have the same dimensions as per your posts. 3T and measures 40uH.

judging from my previous Iprim scopeshot, I think these now have better(?) results... same setup, only GDT changed. there are still those noise but are far lesser

**I have opto on the interrupt input (RCA)
***I have direct connections (GND) from arduino (pwm-ramp) to buck, but controller and buck are separately powered though they share the same GND. I wonder if I should opto-isolate this as well(?)
reason Im using opto is that fiber optic parts are hard to come by.

**point0000.png the initial ramping on the bridge output is 450KHz. and when Iprim starts ramping up, its 370KHz. and because of the 'fast rise' of bridge output, spark output branches early on! 

this -stuffs/early branching of arcs- just happened when I increased UD mosfet drivers supply from previous 20v supply(smps) to 24v Linear-regulated. and from the buck gate driver 15v(smps) to  25v Linear-UNregulated.

-or

just my controller settings!
edit: redid test, adjusted the 'wick' control, even to the minimum setting(virtually removing wick voltage) but there is still the 'quick rampup' on bridge output..

edit2 on phone photo. I have adjusted scope brightness to minimum. no its not 'quick rising'.

[davekni]

**point0000.png the initial ramping on the bridge output is 450KHz. and when Iprim starts ramping up, its 370KHz. and because of the 'fast rise' of bridge output, spark output branches early on! 

I think the "fast rise" is when feedback takes over from self-oscillation.  Perhaps self-oscillation frequency needs to be reduced, and/or feedback signal made stronger relative to self-oscillation.  Not sure how far you can go with either change before frequency locks to lower pole instead of upper as desired.

Glitches appear to occur even when bridge output is not switching.  Any chance the other bridge output is switching there?  Shouldn't be given single GDT.  Unless there is some difference between half-bridges, such as a damaged IGBT or damaged diodes/resistors between GDT and gates.  At least on this side of the bridge, it appears that glitches occur first, then the controller gets confused and causes more glitches.

***I have direct connections (GND) from arduino (pwm-ramp) to buck, but controller and buck are separately powered though they share the same GND. I wonder if I should opto-isolate this as well(?)
reason Im using opto is that fiber optic parts are hard to come by.

Not sure if isolating buck control will help or not.  You could try to see.
Hand-wired breadboard circuitry tends to be noise sensitive.  Ground planes are difficult to add.  Best to have as much of a ground grid as possible:  ground nodes connected to each other both horizontally and vertically within the breadboard wiring.  Also, you could add a copper ground plane to the top of the plastic sheet mounting your driver board.  Cover copper where necessary with tape to avoid shorting to board.  Wire board ground nodes to this ground plane where possible, ideally at least one connection on each edge or each corner of the board.  Wire all connector grounds to this plane too.

[Rafft]

thanks for the inputs David.

whole system got 'different' with my 1st mod attempt, mainly switching my 12v6 boost to 20vdc(powering whole UD) --->> Linear PS 24vdc with 7824. so making the igbt drive to 24v.

do remember these are counterfiet(!) IGBTs.. maybe they cant handle higher than 24v gate drive 

my previous hardware UD ps was 20vdc(and 7809,7805). for the buck, 15vdc only for IGBT high- side driver. maybe its better to go back my previous PS voltage. and do the added GND planes for UD. and maybe add an opto for the buck input as well.

btw, about your phase lead input section on the other thread(20R 10uH 6R8), these values good also for 300KHz-450KHz operation?

edit1
I dont think some of the IGBTs are damaged. I always test them in low-power mode(45vdc & in DRSSTC mode only) and check bridge output. this I do especially when changing new GDT(phasing). I get about 100vpk2pk with a freq of around 410KHz. if all works, then proceed with the usual qcw

below images.bridge output.checked from usb 'wave' saves
**1 burst in DR mode, scope x10 probe x10(no cap in series)
** 410KHz on start (self osc IIRC was set to 480KHz)
** then at end of switching, settles to 360KHz (coil resonance)

[davekni]

do remember these are counterfiet(!) IGBTs.. maybe they cant handle higher than 24v gate drive 

I doubt the IGBTs have any specific issue with 24V.  However, switching times will change a little.  Might change dead time slightly.  Also, at 24V any hard-switching may cause larger noise spikes.  If driver never makes such mistakes, then there should be no hard switching and therefore no such issues.

btw, about your phase lead input section on the other thread(20R 10uH 6R8), these values good also for 300KHz-450KHz operation?

Probably best to use OCD version from later posts, with 5R1 followed by 1k and BAT54 clamp diodes.  Older versions are roughly same equivalent bottom resistance, just made with 6.8 ohms paralleled with 20 ohms (10 ohm POT and 10 ohm resistor).
10uH is probably high for this frequency.  I'd go with the final 4.7uH.
20R on top is fine.

[Rafft]

David

regarding OCD, I have my 3rd CT(isolated) with 100Amp/per volt.

What do I have to modify so the existing LM311 have the same output?

I would be using my existing 1:32:31(currently ocd) to be the new fb_ct input.

thanks

[davekni]

regarding OCD, I have my 3rd CT(isolated) with 100Amp/per volt.

What do I have to modify so the existing LM311 have the same output?

I would be using my existing 1:32:31(currently ocd) to be the new fb_ct input.

Since you already have three CTs, do you have any reason to change to my new OCD version?  Only reason I designed it was to avoid an additional CT.

If you do want to use the new OCD version, your circuit looks good.

[Rafft]

 ok no reason at all except for future reference.

I would just have to rewind my  existing 1:9:9 fb to 1:31:31 for the new self-osc circuit (only for fb, will use existing ocd)

pretty excited about the phase lead adjustment. will also house the entire UD board on a aluminum case.

thanks again 

[Rafft]

guys

sanity check please 

qcw is battery powered, nothing is connected to mains. using UD2.1b with David self-osc mod(set to 470KHz)

*SEC + topload standalone on cardboard, same height as when mounted = 315KHz
                                                                                                              266.22KHz lower pole
                                                                                                              406.66KHz upper pole
                                                                                                              k = 0.4

                                                                                                              Equip used:
                                                                                                              Rigol scope , mains powered
                                                                                                              555 Fres finder, 9v powered
                                                                                                              *LEDs light up brighter (resonance) when scope GND is connected to 555ckt

*PRI alone (MMC parallel) on mount, no SEC 8Turns(12.6uH/14n4) = 367KHz

                                                                                                              Equip used:
                                                                                                              Rigol scope , mains powered
                                                                                                              Sig Gen, 4s(12.6v) batt powered + 10k
                                                                                                              *only one strong peak(resonance) reading

*PRI + SEC (grounded-counterpoise)

8Turns (12.6uH 14.4nF)
279KHz
433KHz

9.5Turns (17uH 14.4nF)
259KHz
407KHz

Equip used:
Rigol mains powered / Sig Gen + 10k, shows 2 peaks (lower and upper poles right?) actually seeing this on scope made me HAPPY 

Actual QCW Operation @225VdcBus:
8Turns = 446KHz~415KHz~ 128Apk & triple-transition can be seen
9.5Turns = 463KHz(self-osc) then 403KHz~381KHz 72Apk only
Rigol scope mains powered. bridge output (yellow) is capacitively coupled with 2pF

Am I nearing tuned state?
why is 8T have more streamer length compared to 9.5T(w/c has close upper/lower poles to SEC)?
what to do?

also posted a shot with phase lead tuning. (yellow) Vce,lower sw (cyan)Vge, lower sw (magenta) separate CT for Iprim 100A/div.
 close enough perhaps(?). done on DR mode only and at a low 40Vdc

Im using 8 Igbts for these. 10R/IGBT. who knows what the specs are, but these are sold(cheaply) as replacments for inverter/welders. IIRC these had high gate capacitance.

thank you for the time reading
-Ralph

[davekni]

*SEC + topload standalone on cardboard, same height as when mounted = 315KHz
                                                                                                              266.22KHz lower pole
                                                                                                              406.66KHz upper pole
                                                                                                              k = 0.4

Don't think I understand this.  What does "standalone" mean?  I presume no primary.  But that means no coupling factor (k) and only one resonant frequency.  Unless you are getting the second resonant mode with voltage peaks at top and middle of secondary.  Are you driving the bottom of the secondary relative to earth ground (signal source grounded on one side and fed to secondary bottom on other side)?  I think that is the normal way to measure secondary resonant frequency.  Should be measured with counterpoise in place so that all sources of topload-to-ground capacitance are present.  Might as well leave primary in place too, just not connected to MMC to avoid primary resonance.  Do you have a ground strike rail protecting primary?  If so, that should be in place too.

also posted a shot with phase lead tuning. (yellow) Vce,lower sw (cyan)Vge, lower sw (magenta) separate CT for Iprim 100A/div.
 close enough perhaps(?). done on DR mode only and at a low 40Vdc

Low Vdc scope capture looks to have more phase lead than the higher voltage captures.  Probably all OK.

why is 8T have more streamer length compared to 9.5T(w/c has close upper/lower poles to SEC)?
what to do?

My guess is that your primary impedance is a bit too high.  Detuning increases primary current and therefore energy transfer.  May be a fine solution.  Optimum might be to reduce primary frequency with a higher capacitance MMC rather than with more primary coil turns.

[Rafft]

David

I measure sec+top Fres the 'usual' way. lower sec winding is signal input.using 555ckt for 'peaks' or using signal gen for 'dips'.

I get my coupling coeff by inputing pri(&top) and sec dimensions on  javatc. thats how I get upper/lower pole calculations.

regarding PRI lower/upper, they are measured by sig gen/scope, again seeing 2 'peaks' , lower & upper. measured when tesla coil is assembled(with mmc parallel to pri)

will again TRY tuning. goal is tuning PRI 5% lower than sec. or maybe id make lesser K by halfing PRI height(and remeasure again upper/lower pole).

[davekni]

or maybe id make lesser K by halfing PRI height(and remeasure again upper/lower pole).

Lower coupling is another way to get higher primary current (instead of detuning).  However, for QCW, I think high coupling is generally better.  Allows energy transfer to continue better as arc grows.  My guess is that lower primary impedance would be the ideal solution.  However, I don't have enough experience to be certain.

I get my coupling coeff by inputing pri(&top) and sec dimensions on  javatc. thats how I get upper/lower pole calculations.

OK, so 315kHz is a measured value and 266kHz and 407kHz are calculated values.

[Rafft]

 do you think it would be better measuring Fres of SEC mounted in place with everything & leaving MMC out?

from what ive seen on some qcw, 13~17uH PRI & 12~15nF MMC. maybe there is some *magic* in there(?). qcw operation 300~500KHz.

have some more question

Im using MKP X2 1uF 225VAC rated caps. this is for bridge snubber. 2s4p. 225VAC would equal 500VDC? I series two, just to be on the safe side. max bus sees only 225Vdc(for now)

for my final phase lead, Im using 1R+R47(but total resistance is only around 1R2) and made a variable inductor 0.6uH~1.5uH. 1:10:10 fb . would this be enough for 380KHz~440KHz operation? was it lowering inductance 'adding' phase lead? or the opposite? when I adjust the inductor around 0.6uH, I see triple transitions. when inductance higher, triple transitions go away. BUT when I test it in DR mode only(40vdc) and low bps, the one with triple transitions have a bigger 'bang' sound than when there is no triple transitions. just wondering w/c of the 2 is best at phase lead tuning. is 'have more phase lead' better? 1R2/1.5uH need another value adjustment perhaps?

[davekni]

do you think it would be better measuring Fres of SEC mounted in place with everything & leaving MMC out?

Yes, though it may not make a significant difference.

from what ive seen on some qcw, 13~17uH PRI & 12~15nF MMC. maybe there is some *magic* in there(?). qcw operation 300~500KHz.

I think most QCW coils use higher DC bus voltage than you do.  For your lower DC voltage ramp, a lower primary impedance may be appropriate.

m using MKP X2 1uF 225VAC rated caps. this is for bridge snubber. 2s4p. 225VAC would equal 500VDC? I series two, just to be on the safe side. max bus sees only 225Vdc(for now)

1s8p would reduce parasitic inductance and almost certainly be fine.  225VAC is 318V peak.  X2 caps are designed to handle much higher pulse voltage for line surges.  Very unlikely to have any issue with voltage on your X2 caps even with 1s8p.

for my final phase lead, Im using 1R+R47(but total resistance is only around 1R2) and made a variable inductor 0.6uH~1.5uH. 1:10:10 fb . would this be enough for 380KHz~440KHz operation? was it lowering inductance 'adding' phase lead? or the opposite? when I adjust the inductor around 0.6uH, I see triple transitions. when inductance higher, triple transitions go away. BUT when I test it in DR mode only(40vdc) and low bps, the one with triple transitions have a bigger 'bang' sound than when there is no triple transitions. just wondering w/c of the 2 is best at phase lead tuning. is 'have more phase lead' better? 1R2/1.5uH need another value adjustment perhaps?

Your adjustment range looks good.  Higher inductance is more phase lead.  Yes, the triple-transition point is optimum.  However, any adjustment mistakes are more problematic if too-little phase lead than too-much phase lead.

[Rafft]

I think most QCW coils use higher DC bus voltage than you do.  For your lower DC voltage ramp, a lower primary impedance may be appropriate.

Yes I agree with this. more PRI inductance for higher bus/ramp voltages.

I have been meaning to up my Vbus supply lately, but my IGBTs are just 600v devices. I have also been testing with PRI with K=0.166(L=13uH) 6" diameter but not pancake. even with measured 420KHz upper pole, I get branching streamers. I have also tried really high k=0.57, problem is the upper pole goes too high (~520KHz), they do work, but I don't want to stress the unknown spec IGBT... so I just used k=0.4 (somewhere between too low and too high) and it works quiet well. Iprim somewhere around 150~170Apk(sometimes OCD). maybe  I need to upper this as well. maybe 200Apk.

and actual measurement of PRI upper/lower pole , and comparing it with SEC calculated upper/lower pole gave me a better chance of tuning my qcw. but I do still need 'calculator' for getting the coupling coeff. next best thing to do is fine tune. PRI taps -or- mmc taps -or- PRI height. still wondering how  wide(or narrow) the tuning range is with QCW, though Im pretty sure its -almost- tuned. just need to squeeze a few more 

225VAC is 318V peak.  X2 caps are designed to handle much higher pulse voltage for line surges.

Good to know. actually it was rated 275vac, so will be around 389v peak. still ok with my low voltage qcw  btw, is it ok to mix 100nF 220nF 330nF 1uF for bridge snubbers? maybe as a multi-freq snubber(?)

Your adjustment range looks good.  Higher inductance is more phase lead.  Yes, the triple-transition point is optimum.  However, any adjustment mistakes are more problematic if too-little phase lead than too-much phase lead.

got it. Thank you! I opted for the higher inductance adjustment on mine. as what you have said before, a cleaner Iprim means good phase lead. I have also noticed a triple-transition when Fres/tune is off (like: tuned 410KHz , off-tune: 350KHz or 480KHz, then I see triple-transitions on bridg output.)  so this means(?) phase lead adjustment is freq dependent... if so, does this mean "tune coil first/look for optimal Fres" before doing phase-lead adjustment? or is (my) phase-lead has wide freq range of when it can work optimally?

[davekni]

btw, is it ok to mix 100nF 220nF 330nF 1uF for bridge snubbers? maybe as a multi-freq snubber(?)

Yes, mixing snubber caps is beneficial.  Smaller value caps tend to do better with the high frequency currents.  (For low-voltage bypass capacitors on logic chips etc., many data sheets suggest using two or three different value caps.)

phase lead adjustment is freq dependent...

Yes, the standard phase-lead adjustment is frequency-dependent.  Ideal phase-lead circuit would lead by a fixed time rather than a fixed phase.  The required time is slightly more than the delay through driver and H-bridge.  Delay time is roughly constant over frequency, so ideal phase-lead would also be constant.  In other words, time-lead would be more ideal than phase-lead.  For normal UD2.7 style phase lead, as frequency increases, the lead increases in phase, but decreases in time.  (For example, 60 degrees at low frequency is more time than 60 degrees at high frequency.)
Yes, proper phase (time) lead is most important when primary current is highest.  That is the best frequency to use for adjusting.  However, make sure that phase lead isn't too little at the higher starting frequency.  Some compromise is needed.
BTW, I designed a roughly-constant-time lead circuit for my DRSSTC covering 50-80kHz.  That circuit has down-sides too, so wouldn't recommend copying.

[Rafft]

finally!

99% tuned

 after 9 months of working(from scratch) with my QCW DRSSTC, I have finally acheived  +1 meter(50" or x10 sec length) of arc from a  battery-operated coil. using 4s(16.8v) 2000mAh Lipo boosted to around 295Vdc. bulk cap is 6600uF 350vdc.
and fed to TL494 buck (asynchro) & single switch IRG4PC50FD . modified UD2.x for self-osc(450KHz). using single full-bridge MBQ60T65PES(counterfeit) for inverter. OCD set to 200A limit. 9n8 MMC wima mkp10. secondary coil 3"×4.75".
big-ass topload 4"x12.5".coil resonates at 459~439KHz 124Apk@295Vdc bus.

included are couple shots from buck output. single trace(cyan) is with low voltage test & 1R load. next 2 traces(yellow) is from ramp output +200vdc input. blue(Iprim). done only on low voltage. dont have x100 probe.

tuning IS the key 😉

[davekni]

Congratulations!!!  Your perseverance is impressive, as is the result.

[Rafft]

Thanks David

it has been a fun long journey .

anyways my current topload is 4" minorDia x 12.5" majorDia.

I would like to make this look smaller. I have been running numbers on topload capacitance calculator.

4" × 12.5" = 13.851pF
vs
2.26" × 6.27" = 6.9899pF , 2 stacked = 14pF

would they 'be the same' ?

meanwhile ill go check javatc

edit: short answer = NOT

[davekni]

would they 'be the same' ?

meanwhile ill go check javatc

When paralleling capacitors in electronic circuits, capacitances do add.  Electric fields are primarily between capacitor plates contained within each part.  When electric fields overlap as in the two toploads, capacitances no longer add.

anyways my current topload is 4" minorDia x 12.5" majorDia.

I would like to make this look smaller. I have been running numbers on topload capacitance calculator.

This might start another long journey.   As you've found, smaller topload has smaller capacitance.  Either frequency increases or secondary coil needs more turns.  Larger toploads can also help direct the initial arc away from the topload.  Finally, smaller topload capacitance means frequency will change more as arc length grows.  (Arc capacitance is a higher fraction of total capacitance.)  Might require higher coupling and/or other changes to handle that increased frequency range.

[Rafft]

http://amasci.com/tesla/toroid1.txt 

 
Do not worry about a perfect and solid connection between all
sections of a homemade toroid. Overlapping foil with an adhesive
layer between may show a poor or non-existent connection when
measured with the VOM, but in practice the skin effect makes
this a moot point. The toriod will be function perfectly even
if all sections are not perfectly electrically bonded. 

just about to finish when I read this.

indeed there is no electrical connection between *almost all* parts of the toroid.  i have calculated this to be almost *the same* topload capacitance to the single 4×12 one.

[davekni]

Do not worry about a perfect and solid connection between all
sections of a homemade toroid. Overlapping foil with an adhesive
layer between may show a poor or non-existent connection when
measured with the VOM, but in practice the skin effect makes
this a moot point. The toriod will be function perfectly even
if all sections are not perfectly electrically bonded.

Conclusion is generally accurate presuming sufficient overlap.  However, it is not skin effect causing this.  Rather it is capacitance between the overlapping sections.  As long as capacitance between sections is high compared to total top-load capacitance, voltage drop between sections will be small.  If capacitance isn't high enough, arcing between sections will still limit voltage drop, though with less stability and predictability.

[Rafft]

I had an aluminum tape toroid before. I had a feeling it was underperforming. forgot to measure electrical conductivity as well. it doesnt. breakout point had no electrical connection to topload. same goes to this build. not unless I poke the taped area with wire connection from breakout.

I just tested it last night, but could not reach ×10 length. only x8 or so. tuning is not so far from my previous single topload.

maybe (?) alu taped covered toroid is too lossy

[davekni]

breakout point had no electrical connection to topload. same goes to this build. not unless I poke the taped area with wire connection from breakout.

If you have a meter that can measure capacitance, check capacitance from breakout to topload.  Depending on how breakout point is mounted, it may not have enough surface area against aluminum to have sufficient capacitance.

maybe (?) alu taped covered toroid is too lossy

If aluminum is all electrically connected there would be no loss issue.  With sufficient overlay, I'd think it would be fine too.  For better contact, try folding over a tiny bit (1-2mm) of tape at the ends and perhaps along edges too.  Those folded over edges will contact directly to aluminum underneath, avoiding non-conductive adhesive.  Not guaranteed given aluminum's natural oxide layer.  Enough contact points mean statistically some will conduct.

[Rafft]

If you have a meter that can measure capacitance, check capacitance from breakout to topload.  Depending on how breakout point is mounted, it may not have enough surface area against aluminum to have sufficient capacitance.

sorry for late response. here it is . my second SEC 4"x 6" . testing .its about 12.6~20pF . red lead is where the breakout point connects. its directly connects to sec winding hv end

another angle, where the secondary wire ends. metal tab and bolt/nut going to other side


NOT currently using it. tested/quicky tuned a few time only, bad performance.. still using my smaller one


btw, here is what Im busy at. this circuit mod will be at the user side. my current controller outputs +5 INTERRUPT & RAMP PWM over RCA sheilded wires. one for UD, and other for buck, w/c also holds the RC ramp filter at input. with audio mod, RAMP PWM path will be replaced with analog line(going buck) 0-5V analog ramp . I would prefer something robust of a low impedance output unity buffer. input doesnt need to be high impedance(but is a plus).
->> will good ol 741 do its job?
->> TL071?
->> or maybe something discrete? Emitter follower?

single supply would be simpler (+7v)

 output goes over 2 meters of RCA sheilded cable. should I worry of plasma spewing around? I felt safer with PWM signal/s on shielded cables

I saw this video on yt(audio capable qcw controller). and this gave me an idea. so I tested it in sim.. it works,  testing with 1KHz ""tone"". hope it sound good in real world




cheers
-Ralph

[davekni]

testing .its about 12.6~20pF

I'd guess that is low enough to cause arcing and resulting instability.  With breakout point wired directly to secondary top, all top-load current (not including breakout current) passes through that capacitance.  It is on the same order as top-load capacitance, making a ~2:1 capacitive voltage divider.  Half output voltage will certainly arc across a tiny gap between foil pieces.

->> will good ol 741 do its job?
->> TL071?
->> or maybe something discrete? Emitter follower?

Probably any of above will work.  Recommend a resistor between buffer and cable, in the 10 to 100 ohm range.  Buffer circuits tend to be unstable with capacitive loads such as cables, both opamps and emitter followers.  Also helps with any stray RF picked up by the cable, limiting RF current feed back to buffer output.

[Rafft]

David

 all noted! thank you.

meanwhile im almost finish with the qcw ramp gen. still lack the output buffer part.

 

 [/url]

[Rafft]

hello again

 what do you guys use for sequencing tones?
tempo control? interrupt(pulse) output as well?

better if standalone. just need two outputs. the note and int signals

sorry but no idea how this is done

thanks

[Rafft]

and it actually works!  wow 

[davekni]

and it actually works!  wow 

Great!  How do the arcs look with audio modulation?  Does the modulation increase branching?

[Rafft]

and it actually works!  wow 

Great!  How do the arcs look with audio modulation?  Does the modulation increase branching?

Thanks. havent done much investigation yet on the arc side. just tested in 135vbus with 1~2 inch sparks. on tesla coil side, not all frequency(audio) have same sound amplitude(audibly) EVEN with signal generator having a constant output amplitude 5Vpk2pk. 1.4KHz~3.9KHz on my test.

currently looking  up a better square to triangle(or sine) converter circuit. 5v input square wave. I dont want to go the arduino+AD9833 route(too much coding for tones/melody maybe). not much of a coder anyway.

btw ive done couple ckt. for sine output melody. lowpass(RC) then amp -> output amplitude not constant, varies with freq. I just need something between 900Hz-5KHz. if only there was a converter with almost constant output amplitude 1KHz~5KHz(then Id be done).

 also with opamp integrator, but output just dies out. needs constant sq wave input.

anyways, that proves the concept. audio on QCW 

update: ok. using the AD9833 dds, was not hard to code  here it is shown a '1KHz note'. output from dds is 0.6vpk. just need to add an amplifier at the end. at least with this one, im guaranted with constant amplitude output.

[Rafft]

 welp

 here is a short run/test of my -musical- qcw of sorts. originally I had a melody that played notes of less than 1KHz. they dont sound well. so I just coded another test melody spanning between 1 and 5KHz just to check how it sounds. 150v bus only. all it lacks is a proper qcw melody(w/c I dont have  )

/>
edit:
I noticed too much branching in the arcs. kinda looks like a regular drsstc ,Im guessing Im over-modulating this. will have to adjust the audio gain lower, and probably achieve more straighter arcs. back to hw mod

edit2:
now it plays. this resonator setup is not fit for 'musical' use. has high coupling resulting in racing arcs. only for low voltage use

/>

edit3:
and just a proof that my junky coil has reached more than x10 its secondary (4.88") length. even the arc tip is out of frame  . its tuned in the upper pole and Ive just tested a CBB81 for MMC. not sure if its even ideal for such purpose, but it works.

https://youtube.com/shorts/aY-SdQSxIV8?feature=share

Ive used a pixel counting app for android (imagemeter)

[Rafft]

David

On a half(or full-bridge), is what Im seeing correct? Lower IGBTs can use lower gate resistors to 'speed it up' , while upper IGBTs cant/wont like it if they have same gate resistors as lower counterparts. Upper IGBTs are better if maybe 2  or 3 times the resistance of lower gate resistors.

In short, upper IGBTs are more 'sensitive' to drive or does not need as much driving(to turn on) as compared to lower switches. Is there some truth in this? Or some explanation perhaps. Still using counterfeit iGbTs so they have high gate capacitance.

[davekni]

On a half(or full-bridge), is what Im seeing correct? Lower IGBTs can use lower gate resistors to 'speed it up' , while upper IGBTs cant/wont like it if they have same gate resistors as lower counterparts. Upper IGBTs are better if maybe 2  or 3 times the resistance of lower gate resistors.

No, AFAIK it is best to have all gate resistors identical.  Diodes across some or all of gate resistance speeds up falling edges of Vge compared to rising edges, increasing dead time.  Again, diodes generally should shunt same fraction of each gate resistor.  (Often diode is across the entire single gate resistor of each IGBT.  Some designs add another resistor per gate, in series with diode or in series with diode // resistor.  That's what I mean by diode being across a fraction of total gate resistance.)

I could imagine some situation where GDT winding-to-winding capacitance causes significant difference between high-side and low-side Vge waveforms.  Perhaps such a difference could be mitigated with changes in gate resistance.  Haven't seen any such cases, however.  I prefer to add common-mode chokes to high-side leads of GDT (a few turns of high-side GDT output twisted pair around a ferrite bead, one bead per high-side IGBT).  Minimizing GDT turns also helps by minimizing capacitance between windings.

[Rafft]

Mine does have low R in series with diode, to minimize undershoot. Also have bead as per your suggestion months ago.

But does. 3-5T vs 1T(wire passing bead, no loop) have big impact?

[davekni]

But does. 3-5T vs 1T(wire passing bead, no loop) have big impact?

Depends on the frequencies that need attenuating.  Inductance goes as turns^2, so 3T has 9x inductance compared to 1T.  For lower frequencies, that is 9x impedance.  5T is 25x impedance.  However, more turns isn't "free".  Turns add capacitance between turns, which couples high frequencies between turns, reducing the advantage.  Also, more turns increases GDT lead length, which increases leakage inductance, which increases Vge ring.  I typically use 3-4T.  If I'm being a perfectionist, I'll wind 3-4T around one core, then pass through (1T) another core.  That's what is on my DRSSTC.  Likely not necessary.

[Rafft]

Currently have this on mine.

Left is lower sw. Right is high side. Im guesstimating maybe 2-3T would do enough for 450KHz? Bead is big

[davekni]

Left is lower sw. Right is high side. Im guesstimating maybe 2-3T would do enough for 450KHz? Bead is big

Even your existing 1 turn is probably fine.  Most coils don't use any common-mode ferrites.  Do you have a particular issue to solve?  If you have enough spare GDT lead length, 2 turns wouldn't hurt.

BTW, for multiple turns, NiZn ferrites are ideal.  I use 4S2 material cores.  NiZn ferrites are typically used for high-frequency EMI filtering.  Material is insulating (non-conducting) at room temperature.  More common MnZn ferrites are conductive.  That core conductivity increases capacitance between turns, so makes high turn counts less helpful.  Presuming your core is MnZn, 2T might be the ideal compromise.

[Rafft]

On my current setup, hi side lead length is already short, so id probably add a few, to have the additional turns on bead. Whats the reason though for adding a 1T series after the mentioned 3-4T?

Issue to solve: still a bit of overshoot(bridge output). Undershoot was solved with the R+D on discharge.

Im unsure what material I used on the hi side bead. IIRC these came from inside a printer. From paper feeder motor leads(?) .

[davekni]

On my current setup, hi side lead length is already short, so id probably add a few, to have the additional turns on bead.

Then I'd probably leave as is.  No reason to add lead length.

Whats the reason though for adding a 1T series after the mentioned 3-4T?

The 1T bead increases common-mode impedance at very high frequencies where capacitance may shunt the 3-4T.  As I said, this is the perfectionist version.  I probably didn't need either of the common-mode chokes, and especially not the extra bead.

Issue to solve: still a bit of overshoot(bridge output). Undershoot was solved with the R+D on discharge.

Common mode chokes on gate drive are not likely to fix output overshoot.  Is it bad enough to need fixing?  If so, it is likely related to one of the following:
1) Too little (or too much) phase lead.
2) Too little (or perhaps too much) dead time.
3) High parasitic inductance in H-Bridge construction.  Probably not this issue.  Image shows nice short IGBT leads.  Presuming ECB behind IGBTs has overlapped parallel planes as I've discussed elsewhere, that's about as good as is practical.

Im unsure what material I used on the hi side bead. IIRC these came from inside a printer. From paper feeder motor leads(?) .

Most likely an MnZn material.  Those are more common.  If you have more of these cores, you can test by holding one (with an insulating handle) in any form of spark (from spark coil or flyback or etc.).  I you can make arc jump to one corner of bead and exit other corner, it is MnZn.  If arc goes around bead (along bead surface), then it is NiZn.

[Rafft]

My most likely answer to what might be my problem is on your 3rd qoute.

Im using a gapped EE core (for L= around 14uH) and 33R, from the stock UD phase lead(variable L, 51R). And BACK to using stock UD comparator circuit.

Things ive noticed:
*I can start my coil even w/o the initial 'wick'
*triple-transition is absent no matter how I tweak inductor value

I think too little phase lead I presume(?)

Regarding layout, its 2layer board. Low inductance layout as per your suggestion before.

Really cant seem to get rid of that distortion (Ipim) 1v=100Apk. Have also upgraded my bulk cap from 6600uF to 13200uF. My boost also to a max 327Vdc. Scope shots shown with only ~290vdc bus

[davekni]

*triple-transition is absent no matter how I tweak inductor value

LoneOceans has a good DRSSTC design project page showing triple transitions and phase lead adjustment:
https://www.loneoceans.com/labs/drsstc3/
He initially has excess dead time, making large long triple transitions.  After reducing dead time, triple transition shows up as just a small bump in H-bridge output, not full transitions.  As he says, this is ideal dead time.  If dead time is too short, there will be no hint of triple transition at any phase lead adjustment.  Or, if dead time is sufficient, then phase lead isn't being adjusted across the ideal point.  Either all too little or all too much phase lead (less likely).

[Rafft]

Yes Ive been reading that exact artice as well

But what is the reason why lowering the resistor value from 51R to 33R made the coil more 'sensitive' and easier to start(even w/o wick)? Even with 45Vdc on bus, it can start

Going back to using 51R, it has erratic operation(needing wick) but even at 150vdc bus, it still has erratic operation. Turning OFF my boost ckt(discharging the bulk caps) and it will work *just fine* until bus voltage gets too low..

[Rafft]

David

I did the 3T+1T bead for the hi-sw. All gate res are equal(hi lo)

It must have done something because now coil again needs 'wick' to operate.

Done some adjustments as well on the phase lead. Removed secondary, using only primary & mmc. This was done in regular DR mode. No ramp supply. Test with 45vdc bus. DR interruptor. Yellow(fullbr output) blue(Iprim)

Adjustments where done to get the cleanest Iprim waveform
While also noting the bridge output 'peaks'.

Here are the results. W/c of the 2 shots is better?

Edit
Added 3 more images. Iprim cleaned up a bit 

[davekni]

It must have done something because now coil again needs 'wick' to operate.

UD2.7 startup at low Vbus is inherently marginal.  Depends on first half-cycle generating enough feedback to trigger second half-cycle.  Required feedback voltage depends on diode Vf of the back-to-back clamp diodes, which is temperature dependent.  You can improve startup by adding resistors across the two IGBTs that are off during the first half cycle of each burst.  That way MMC is charged to opposite state, making larger first-half-cycle signal.

Here are the results. W/c of the 2 shots is better?

Hard to say.  Both are likely OK.  Duty cycle looks close to 50%, but there must be enough difference to cause the significant difference between rising and falling edges of H-Bridge output.  Would be a bit easier to tell looking at each H-Bridge output separately rather than the difference between the two.  No way in these images to determine if the two outputs contribute equally to shape of transitions.  In an ideal world, H-Bridge outputs are identical (except for 180 degree phase shift) and rising edges look exactly like inverted falling edges.  Real world examples can come close, but often don't.

[Rafft]

1st photo was using 56R(LR on input) for phase lead & triple transition present. 2nd photo was using 33R & just a hint of the triple transition. Both adjust to have minimum spikes.

I do recall I had a trimmer to adjust duty ratio for MAX913. Will go check on that

Thanks

Edit
Here are scope shots of Vce. 45vbus.DR mode.no secondary
Phase Lead adjusted to have minimum spike(fullbr output)

1 low
2 low
2 high
1 high
Output from comparator. Max913 pin#7(out) 50% duty. Inv_out is on pin#8(edit-sorry for that)

[davekni]

Edit
Here are scope shots of Vce. 45vbus.DR mode.no secondary
Phase Lead adjusted to have minimum spike(fullbr output)

1 low
2 low

The two low Vce captures are all that is needed.  Avoids scope ground capacitance loading H-Bridge output.  Falling edge of low Vce should be identical to rising edge of high Vce.  Differences in your captures prove that scope ground loading is affecting measurements at least some.

There is still a significant difference between rising and falling edges.  Duty cycle seems close to 50%.  Slightly above 50% on Max913 pin#7(inv out) capture you show.  Not sure that is enough to explain differences in H-Bridge outputs.  I wonder if your four IGBTs aren't all the same inside.  Perhaps some have more gate capacitance or have slower turn-off times.  Looks like more dead time on 1 rising edges and a bit more on 2 falling edges, with less on 1 falling and 2 rising.  (I think you said that all gate resistors are identical, as is recommended.)

These all look good enough even though not quite symmetric.  I'd perhaps aim for slightly more phase lead to compensate for IGBTs slowing down when hot, even though spikes will be a bit higher initially.

[Rafft]

Thanks for checking

Yes all IGBT are of same batch(on order). Yes all gate res to be the same.

Btw how should lo Vce look like? Almost/no spike?

Should I just remove 3T +1T? And just make it 1T?

'Aim for more phase lead' ,is it literally like more R or more L value at feedback input?

[davekni]

es all IGBT are of same batch(on order).

I was just wondering.  If the batch are counterfeit, perhaps they aren't all identical inside even if from same "batch".  I'm just struggling to explain why there's differences between the two H-Bridge outputs.  Differences between rising and falling could be the slight duty cycle deviation from 50%, but even that seems questionable.  Or, perhaps there is enough difference in GDT + leads leakage inductance between the four outputs to make a difference.

Btw how should lo Vce look like? Almost/no spike?

Ideally, as you adjust phase lead, there will be a point where there is a small triple-transition.  As phase lead is adjusted slightly higher, that triple transition will fade to just a bump.  There is likely to be a small spike after each edge as opposite IGBT diode turns on.  No large spikes.

Should I just remove 3T +1T? And just make it 1T?

If 3T is adding significantly to GDT lead length, then it might be an issue.  However, if leads are twisted pair through 3T ferrite, length isn't likely to add enough inductance to be significant.

'Aim for more phase lead' ,is it literally like more R or more L value at feedback input?

Phase lead is determined by L/R.  More lead is either more L or less R.

[Rafft]

Thanks again David

My gdt leads are as short as can be possible.

I was thinking maybe add more gate res... Because D+R certainly fixxed the undershoot(bridge output)

Regarding phase lead, I will user a higher inductor value and keep the 33R(or I might try 22R if my junkparts have)

[Rafft]

David

/>
1 lo Vce
2 lo Vce (why so high spike?)
3 bridge output(yellow) and blue(Iprim)

Is this ok now? Or should I add more phase lead? Iprim looks clean enough

[davekni]

2 lo Vce (why so high spike?)

Might be just IGBT package inductance.  Even with perfect H-bridge construction, TO247 packages have about 13nH lead inductance.  Internal IGBT anti-parallel diode forward recovery time may be contributing.  This ~30V spike will likely grow some with higher current.  Even if it grows to 100V, you have sufficient margin between Vbus and 600V IGBT rating.

Is this ok now? Or should I add more phase lead? Iprim looks clean enough

Yes, I think it is good enough.  No need to obsess with elusive perfection.

Video nicely shows the normal Vbus ring as interrupter turns off.  Also shows how residual H-bridge output voltages changes, which is one factor in why startup sometimes works and sometimes doesn't at low Vbus voltage.

[Rafft]

Few component tweaks & phase lead re-adjust did give me a small performance increase. 12mF bus cap. 150Apk@300vdc bus. That bigass topload really makes the spark look tiny .... And those tiny twigs ruin the sword. Wonder if high k has something to do with it

Module next to coil (& inverter) is my boost-buck.

Thanks again David 

 


 

[davekni]

Great performance!  Always nice to see success.

Wonder if high k has something to do with it

I don't think so.  I think most QCW coils have high coupling, certainly much higher than conventional DRSSTCs.  In general, I think higher is better for QCW.  However, my only real QCW experience is with my failed low-frequency experiment.  I believe its issues were due to 100kHz (and eventually due to insulation failure), not due to its 0.9 coupling factor.

Edit:  Bit more thought leads me to the possibility that lower coupling could have one advantage.  More energy stored in primary could smooth over imperfections in buck converter ramp smoothness.  However, lower coupling will make energy transfer from primary to secondary change more as arc grows and changes secondary frequency.

[Rafft]

Hi David

Sorry to be back again.. But.. Im here to learn  ;. More
Ive read on others posts, ZVS tuning..so I will try. Gifted with a proper scope & I should need to use it properly!

vbus 45vdc
Yellow(Vce lo sw)
Blue(Vge lo sw)

This is a problem, right?  Thats why the spike. Looks like igbt needs to switch off earlier? Or is this something else? Gate drive looks non-symetrical(?) Or I should probably accept the fact im using counterfk igbts 

Last image; higher rep rate drsstc int
Yellow(Vce) blue(Vge) cyan(Iprim)

Last last 2 images. I removed the big bead 3T. And removed the extra wires going to HI sw.but installed a teeny tiny bead instead. 1T. Your right, almost no difference w beads. 2nd photo is w bridge output and Iprim

Regarding gate resistors, with UD2.7(output) , is 1/4watt enough?

EDIT:

Bit more thought leads me to the possibility that lower coupling could have one advantage.  More energy stored in primary could smooth over imperfections in buck converter ramp smoothness.  However, lower coupling will make energy transfer from primary to secondary change more as arc grows and changes secondary frequency.

by how much lower? like .1 or .2? last time I checked mine was around .6. lower coupling WILL increase Iprim(maybe >200Apk) if trying to get same output. everything is in a balance.

everything from coupling k/lower-upper poles/size of torroid/ratio (of primary L and MMC C, big or small values). regarding the buck converter ramp smoothness, I havent touched anything on the resonator part, just the gate res/L in hi side phase lead/phase lead adjustment. thats why my previous was I could get it to ramp-up from zero, to the recent ramp w 'blip' on the start.

Ive made mine to achieve long sparks with minimum primary current I could give it. my OCD though is setup for 200Apk

btw hows your HF(?) QCW coming along?

[Rafft]

I will just continue here 

Now I understood more

**Initial scoping
**Replaced gate resistor values(lower R), faster turn-on perhaps(?)
**Replaced phaselead resistor(lesser phaselead)worst ringing Vce(yellow) Vge(cyan) Iprim(magenta). Also noting the earlier ON of Vge
**Replaced again higher gate resistor. Adding more phaselead. Now Vge can be seen turning on just in time before Vce switches off
**Iprim(magenta) bridge output(yellow)

I hope Im correct on what I see on scope

[davekni]

Nice detail in your scope captures!  I appreciate the zoomed-in shots being added (100ns/div), and the ones with current centered and scaled up to make zero-crossing points easier to see.  Takes me less time to analyze your scope images now.

This is a problem, right?  Thats why the spike. Looks like igbt needs to switch off earlier? Or is this something else? Gate drive looks non-symetrical(?) Or I should probably accept the fact im using counterfk igbts 

Actually, I think these initial scope traces look quite good.  The positive spike on Vce is due to some combination of H-bridge wiring inductance (including or perhaps mostly IGBT package lead inductance) and internal IGBT antiparallel diode forward recovery time.  (Counterfeit IGBT parts may have slow diodes.)  Low-side IGBT is turning off before current reaches zero, exactly as it should.  Remaining current causes rising edge of Vce.  When Vce reaches Vbus (45V), diode of high-side IGBT clamps it to Vbus.  Delay in diode turn-on (forward recovery time) and/or inductance allows Vce to spike above Vbus for ~50ns.  I don't think that will cause any problems.  Following the spike, notice that there is another 60-70ns plateau then a small drop in Vce.  This is high-side IGBT diode forward drop changing to IGBT on-state Vce as current passes through zero and changes polarity.  This brief plateau and small drop demonstrates the desired ZVS turn-on condition.

Regarding gate resistors, with UD2.7(output) , is 1/4watt enough?

Probably not enough power.  However, many small details can affect power.  Not easy to calculate with any simple formula.  One way to estimate is to measure UD2.7 driver FET power (current times voltage) consumption (or perhaps total +24V power if hard to separate).  Worst-case half of that power will end up in gate resistors, so 1/8th per resistor.  More likely less than that, perhaps 1/16th of UD2.7 output FET power per gate resistor.

by how much lower? like .1 or .2? last time I checked mine was around .6. lower coupling WILL increase Iprim(maybe >200Apk) if trying to get same output. everything is in a balance.

Wow, is your coil coupling really 0.6?  That high is difficult to achieve without secondary-to-primary arcing.  Other QCW coils I've seen on the forum range from 0.35 to 0.55.  0.55 coil required ferrite core to get that high.  I wouldn't go below 0.35 minimum.

btw hows your HF(?) QCW coming along?

Slowly as always.  Some work last October and November, then a break for Christmas etc.  Hard to host family when my dining room table is filled with project material.  Hope to have it operational by summer.

**Replaced gate resistor values(lower R), faster turn-on perhaps(?)

Yes, faster turn-on, which reduces dead-time given same turn-off time (which isn't quite accurate due to higher loading on UD2.7 and GDT).  Rising edge still has sufficient phase lead for clean transition.  Falling edge doesn't complete Vce transition before low-side IGBT turns on.  That causes the noise on falling edges.

**Replaced phaselead resistor(lesser phaselead)worst ringing Vce(yellow) Vge(cyan) Iprim(magenta). Also noting the earlier ON of Vge

Yes, made falling edge even worse.  Vce barely starts to fall before low-side IGBT turns on.  Current is so close to 0 by the time high-side IGBT turns off that it can't effectively cause low-side Vce to drop (high-side Vce to increase).

**Replaced again higher gate resistor. Adding more phaselead. Now Vge can be seen turning on just in time before Vce switches off

This is best adjustment yet.  At least best under these conditions.  Conditions in your first post have a bit more spike due to a bit more phase lead, so current is a bit farther before zero at turn-off.  That margin might be helpful if IGBTs slow down significantly as they get hot (normal behavior for IGBTs and most silicon devices).  On the other hand, higher bus voltage reduces IGBT capacitance.  Lower capacitance might add just enough margin for slowdown due to heat.  Difficult to know exactly where perfection is.  Either this state or the one of your first post of this pair is probably just fine.  Seems that you are learning a lot with this drive for perfection.  That part is good.  Don't get obsessed with perfection as I too-often do.

[Rafft]

David

I know this is probably fine by now but.. When phase lead tuning, whats 'best' to watch out for? Get minimum 'spike' (yellow,Vce) or the ringing(of Vce) when Vge(cyan) turns ON?

Or just disregard that ringing of Vce(Vge turn-on) b/c its insignificant already(?)

[davekni]

I know this is probably fine by now but.. When phase lead tuning, whats 'best' to watch out for? Get minimum 'spike' (yellow,Vce) or the ringing(of Vce) when Vge(cyan) turns ON?

Or just disregard that ringing of Vce(Vge turn-on) b/c its insignificant already(?)

Yes, I'd disregard the Vce turn-on ringing, at least within the range you show in the video.  It's low amplitude and does not change that significantly over the range shown.  Important feature of turn-on is that Vce is already close to 0V before Vge rises above threshold voltage (above ~5V).  That appears to remain true over range shown.

Turn-off Vce spike is larger when phase lead is larger.  Turn-off is farther before zero current.  The larger remaining current causes a larger voltage spike when that current transfers to the high-side IGBT's diode.  Within shown range, the minimum Vce turn-off spike (minimum phase lead) is still sufficient.

Even though the Vce turn-off spike isn't large enough to be a likely problem over any of the shown range, you could for learning figure out what is causing it.  Leave scope probe ground as connected for low-side Vce scoping.  Trigger on Vge or current or any other signal in order to have fixed trigger timing.  Then move scope probe tip from low-side collector to high-side emitter.  See how much of the spike is still present.  Then move probe tip to high-side collector.  See how much spike occurs there.  If a spike on high-side collector (Vbus), that is a result of inductance in snubber cap leads or bridge construction.  If the spike is not on high-side collector but is on high-side emitter, then spike is due to high-side IGBT lead inductance or diode forward recovery time.  This testing will be easiest with a large spike to measure, so set phase lead high (large Vce spike setting) just for this experiment.

[Rafft]

David



**LO Vce, Vge as trigger

**HI emitter leg , LO emitter gnd reference
I notice a shorter spike

**collector leg(Vbus) LO emitter gnd reference
Are these spikes on Vbus too much? Im using 1uF 275v~ mkp x2(4pc in parallel) parallel-wired using 2 layer pcb(planes) and wired as short as possible

**fullbridge

[davekni]

Thank you for the scope traces.  Makes it clear that the majority of spike voltage is due to snubber capacitor inductance, with a little contribution from IGBT lead inductance.  Don't see much evidence of diode forward-recovery time issue.

Are these spikes on Vbus too much? Im using 1uF 275v~ mkp x2(4pc in parallel) parallel-wired using 2 layer pcb(planes) and wired as short as possible

No, these spikes should be fine.  Only issue would be if they went far enough above Vbus to exceed IGBT Vce rating.  Seems quite unlikely.
Your construction is better than most.  No need to worry.  If you want to be excessively perfectionistic:  Snubber inductance is likely dominated by the short wires down to IGBT board.  Pairs of Vbus+ and Vbus- wires closely-spaced will reduce inductance.  Multiple pairs and close wire spacing both help.  Or, snubber caps on the same board as IGTBs.  That can be better (lower inductance) even if snubber caps need to be physically farther from IGBTs.  Overlapping parallel planes have lower inductance than paired (twisted or adjacent) wires unless many pairs are paralleled.
BTW, I'm presuming overlapping parallel planes already exist separately on the two boards, IGBT board and snubber capacitor board.

[Rafft]

Makes it clear that the majority of spike voltage is due to snubber capacitor inductance

Im relieved. I have no proper snubber caps at the moment, had to use whats on the bin. as noted from your previous reply --> mkp x2 as excellent spike catchers

Snubber inductance is likely dominated by the short wires down to IGBT board.

Ive mine constructed on 2 layer ECB(overlapping planes) and "wired" to IGBT board(also 2 layer planes) as short as possible. around 10mm or so leads.

these spikes should be fine.

glad to hear this as well. 

Pairs of Vbus+ and Vbus- wires closely-spaced will reduce inductance.

mine is twisted and short , to 5mm bullet connectors(used on RC toys), w/c connects to twisted wire from buck output

BTW, I'm presuming overlapping parallel planes already exist separately on the two boards, IGBT board and snubber capacitor board.

yes correct! overlapping planes for IGBT board and another for the paralleled snubber caps

Thank you 

EDIT: forgot to mention,  scope shots where done with the -worst- spikes possible, so when I adjust back again the phase lead, spikes will all go down 

EDIT 2: changing gate resistor from 1/4watt to 1watt made a small difference. Now I see tripple transition. I will adjust this to have the smallest Vce spike

/>

[davekni]

Im relieved. I have no proper snubber caps at the moment, had to use whats on the bin. as noted from your previous reply --> mkp x2 as excellent spike catchers

Yes, X2 caps aren't ideal for snubbing.  I still suspect at least half of the inductance is in 10mm wires rather than in the cap leads, given there are four parallel caps.

mine is twisted and short , to 5mm bullet connectors(used on RC toys), w/c connects to twisted wire from buck output

I was referring to the 10mm wires from snubber board to IGBT board, not your buck-to-H-bridge connection.  Yes, very short.  But pairing will still reduce inductance even for those short wires.  Best I could tell from image, the two 10mm wires were not adjacent.

EDIT: forgot to mention,  scope shots where done with the -worst- spikes possible, so when I adjust back again the phase lead, spikes will all go down 

Yes, that is exactly what I'd suggested you do for this experiment.  Could see that in scope traces too.

EDIT 2: changing gate resistor from 1/4watt to 1watt made a small difference. Now I see tripple transition. I will adjust this to have the smallest Vce spike

Minimum phase lead in this video is a bit too little.  Shows small triple-transition on rising edge.  More concerning is that falling edge isn't quite complete at the point where lower IGBT starts turning on.  One evidence of this is the longer plateau on Vge at turn-on, as well as observing that Vce isn't to 0V at the start of that plateau.  (Would still be OK, except that IGBTs will slow down as they get hot, so need a bit more phase lead to compensate.)

[Rafft]

I can see a small plateau of Vge at 0v. Vce is also below 0v now.

Edit (2 shots 1 diagram)
Same phase lead setting
**1 lo Vce - where I can see worst rising spike

**2 lo Vce - best Vce

**fb /ocd/ 3rd ct connections

[davekni]

Edit (2 shots 1 diagram)

I see three scope captures.  First two look like plenty of phase lead.  Third looks like barely enough phase lead.  That does minimize current at switching time, so minimizes overshoot spike.  Good as long as IGBTs do not slow down too much when hot.  The ones with more phase lead are safer in that respect - margin for IGBT heating.

**fb /ocd/ 3rd ct connections

As expected.  Thank you for confirming.

[Rafft]

Hi David

First scope shot was with different phase lead L value

Next 2 shots, are from another L value. Are taken from 'side 1 Lo Vge' & another 'side 2 Lo Vge' w/c is reffered in the diagram.

I read in another web post, fb/ocd should be connected on the other side, like bridge output before MMC(?)

[davekni]

First scope shot was with different phase lead L value

Next 2 shots, are from another L value. Are taken from 'side 1 Lo Vge' & another 'side 2 Lo Vge' w/c is reffered in the diagram.

Are you certain?  First two look quite similar.  Third looks like less phase lead (lower L or higher R).  If the second and third shots are with same phase lead, then there is a significant unexpected difference between side 1 and side 2 (different gate R, or perhaps quite different internal IGBT die).

I read in another web post, fb/ocd should be connected on the other side, like bridge output before MMC(?)

Doesn't matter which side.  Just not in the middle between primary and MMC.  Voltage is much higher there, risking insulation breakdown within CTs.

[Rafft]

Yes very certain    1st and 2nd scopeshots are similar because they are from same '1 lo side igbt' aka the worst Vce turn-on spike. I can also see '2 lo side' Vce/Vge to have better phase lead. Yes I know they look different(Vce Vge side1 & side2). different IGBTs as you say. Also, all those Vce Vge shots where taken from 'side 1' w/c has worst Vce spike.

Anyways I will be testing cbb21 470nF 630v caps for snubber.. Spikes do look smaller than with mkp x2. Wonder if this is because of lower value capacitance(1.4uF vs 4uF) or because leads are very short.

EDIT:
Since I have changed -snubber- caps to lower value, why has the bridge output become wavy?

I have also been too preoccupied having a good looking wavefirm. Am I correct in saying that these IGBTs can switch on-off in less than 100nS? If so, then its not that bad of an IGBT? These are sold as 'stick welder' IGBTs(about 2$ each).
Btw what is right term for this(turn-on time? Rise time?)

And could it possibly why im getting higher Vce spike because im scoping on the PRImary coil side and not on MMC side? (Just a thought)

[davekni]

Anyways I will be testing cbb21 470nF 630v caps for snubber.. Spikes do look smaller than with mkp x2. Wonder if this is because of lower value capacitance(1.4uF vs 4uF) or because leads are very short.

Because of short leads, including that caps are directly on IGBT board, so no additional lead/wire length to separate board.  Cap impedance is very small at the spike frequency whether 1.4uF or 4uF.

Since I have changed -snubber- caps to lower value, why has the bridge output become wavy?

That is due to the reduction from 4uF to 1.4uF.  It is the resonance of 1.4uF with wire inductance from H-Bridge to buck converter.  Happens to now hit 2x coil frequency, which is exactly the dominant frequency of H-Bridge supply current.  One easy fix for this is to add back in 4uF too, even with somewhat longer leads.  Local 1.4uF will still reduce spike.

I have also been too preoccupied having a good looking wavefirm. Am I correct in saying that these IGBTs can switch on-off in less than 100nS? If so, then its not that bad of an IGBT?

Yes, under your operating conditions, IGBTs are turning off in 100ns or a bit less.  However, IGBT data sheet specifications are usually at hard-switching conditions.  Switching is faster at ZCS conditions.  Thus it is hard to say how these parts compare with genuine ones.  May be close.

Btw what is right term for this(turn-on time? Rise time?)

IGBTs typically have four time specifications.  Turn-on delay, rise time, turn-off delay, fall time.  Rise and fall here refer to IGBT current.  Generally measured in hard-switching conditions, half-bridge feeding an inductor.

And could it possibly why im getting higher Vce spike because im scoping on the PRImary coil side and not on MMC side? (Just a thought)

No, that does not directly matter.  However, routing of scope probe and ground lead can make quite a difference.  Loop formed by probe and ground lead will have induced voltage from adjacent fields.  Routing scope ground lead along probe tip as much as possible will help.  Some people wrap (spiral) ground lead around probe, then connect to circuit as close to probe point as possible.

[Rafft]

Wow! Those 'snubber' capacitor lead inductances DO matter, how they should be mounted/connected. Ive tried adding the 4uF mkp x2 above the 220nF(4p) but still, resonance can be seen & it looked like the 4uF isnt doing anything. So I just removed all caps, removed the 1uf 4p caps from its 'board', and just connected them to the igbt board. But I only manage to fit 3uF(no more space on the board) . result is almost no Vbus resonance and very little spike.

Ive also taken photos of #1 lo Vce Vge & #2 lo Vce Vge. Same phase lead settings.

Also tried short and long breakouts. Longer breakout point makes longer sworksparks. Almost same primary peak currents.

[davekni]

result is almost no Vbus resonance and very little spike.

First scope capture shows the normal Vbus resonance at end of enable pulse, about 80kHz in this capture.  (Presume this is normal DRSSTC mode.)  80kHz is so far from twice 310kHz operating frequency.  Makes me think my original explanation for 620kHz resonance was incorrect.  This 80kHz resonance is probably due to snubber capacitance and wire inductance to buck converter.  Not sure where the 620kHz resonance came from.  Are there any other caps across VBus on the H-bridge board?

Ive also taken photos of #1 lo Vce Vge & #2 lo Vce Vge. Same phase lead settings.

These two captures look similar enough, as they should be.  They do show that current measurement is about 90ns ahead of actual current.  That can be caused by CT inductance and/or by inductance of burden resistor.  Is burden resistor wire-wound?

Final picture shows one scope probe with no ground clip (I think).  Which signal is on that probe?  High frequency details will not be accurate with an ungrounded probe.  The loop from one probe ground to other probe is quite long with lots of inductance and lots of loop area to pick up signals.

[Rafft]

Vce scope shots is my final hardware. 1uF mkp x2 3parallel. 3uF total. As snubber.Those are the only caps left on the igbt board. Yes DRSSTC mode. No secondary coil.

CT burden resistor is 1watt carbon

Final picture. Vce has the GND clip. The unGNDed probe is Vge. I leave the GND clip for Vge open(when checking the bridge output/s

[davekni]

CT burden resistor is 1watt carbon

If I recall correctly, burden resistor is 1 ohm.  So only 90nH is needed to make 90ns of phase shift.  Could easily get that much with lead wires even though not a wire-wound resistor.  Ideal to have both CT secondary and scope connections close to resistor body (short resistor leads), with scope leads heading opposite direction from resistor as CT secondary leads.  Of course, now that a bit of phase shift is clear, can be just mentally compensated for in viewing scope traces.  I'd guessed way back that there was some phase shift (lead) in CT scoping.  It was clearer in recent captures.  (CT inductance is almost certainly not an issue with 1 ohm burden resistor.  CT inductance can become an issue with high-value burden resistors.)

[Rafft]

Just another -fine tuning- update:
 
As per filename shows:  primary current difference between high and low coupling. My coil is telling me 'something' (on that faster rise on ramp peak) but I dont quiet get it yet 

- does it want more ramp voltage? More ON-time perhaps? Is it indicating coil is resonating better? Is it at its prime of resonance? Any thoughts and insights welcome.

Hardware: same secondary and topload. Same PRI inductance(to maintain k to sec winding). I adjust MMC only.
Sec Fres vs Pri Fres, Fres is always higher(upper pole)

Just for reference, highK=400kHz lowK=355kHz

Thank you

[davekni]

As per filename shows:  primary current difference between high and low coupling. My coil is telling me 'something' (on that faster rise on ramp peak) but I dont quiet get it yet 

My guess:  High coupling case looks good.  Low coupling case shows current rising quickly at the end.  That is likely arc loading pulling secondary frequency too low.  With lower coupling, primary-to-secondary power coupling drops off more rapidly as frequencies move apart.  Without much power transferring to secondary, primary current ramps up.  I think that is why QCW coil designs typically have high coupling.  Normal DRSSTCs would also be better with high coupling, except that high coupling causes racing sparks in normal DRSSTC mode.

[Rafft]

I have a question regarding the UD2.1b

During no oscillation, what is the state of comparator output?

Im instead using MAX913 as my comparator. Pin#2 is on a voltage divider @1.599v(as per diagram). When I used this, pin#7(non inv out) is at active high. During test on DR mode(no sec), I get erratic output. So I adjusted that voltage to about 1.542v(pin#7 is now active low). Is my hardware corrent? Or it doesnt matter if active low or high on pin#7 output?

Reason I asked is,  a single big glitch(Iprim) near ramp top. Was wondering if this has something to do with comparator output/s state.

Ive tuned phase lead to the best output(clean Iprim) I could get.

Blue(single channel) is comparator pin7(non-inv out) with pin2# on 1.599v. On 1.54v it is inverted

Phase lead capture. Yellow(Vce) cyan(Vge) magenta(Iprim)

No idea where/what this glitch is comming from. Maybe cause of slow/fake IGBTs?

A small video. Just random ramp area to check Iprim. Ignore the 'glitch' on ramp-down. Scope memory error perhaps. qcw test done with 300vbus.

/>
And a short phase lead tuning

/>


Edit:
Issue Fixxed !
On higher ramp ON-times, 'ripples' on the ramp edges hinted me of something. Looks like inverter bridge resonance(?) Or buck output 'problems'. Tried increasing buck output LC Filter capacitance. Seems to have -fixxed- it.

[davekni]

Im instead using MAX913 as my comparator. Pin#2 is on a voltage divider @1.599v(as per diagram). When I used this, pin#7(non inv out) is at active high. During test on DR mode(no sec), I get erratic output. So I adjusted that voltage to about 1.542v(pin#7 is now active low). Is my hardware corrent? Or it doesnt matter if active low or high on pin#7 output?

The default state should be with pin 7 high (1.599V on pin 2).  Pin 3 is one diode drop lower when quiescent due to current through 100k resistor R27.  I'm a bit surprised that such a small change of pin 2 voltage switches state.  If D1 and D2 diodes are 1N5819 or similar, then perhaps that is reasonable given the small current through R27.

Do you have any bleed resistor across H-bridge outputs?  If not, initial state is indeterminate.  Perhaps your H-bridge initial state tends towards one direction or the other due to imbalances in IGBT leakage current.  If so, changing comparitor initial state may be providing a better initial half-cycle signal to get oscillation started reliably.  Startup is particularly problematic at low bus voltage as in your QCW use.

And a short phase lead tuning

Looks like good setting at the end.  Video is a great way to show the range.  Behaves roughly as expected.

Issue Fixxed !
On higher ramp ON-times, 'ripples' on the ramp edges hinted me of something. Looks like inverter bridge resonance(?) Or buck output 'problems'. Tried increasing buck output LC Filter capacitance. Seems to have -fixxed- it.

Glad it's fixed.  My first guess is that some noise coupling from buck to UD2.1 is causing an H-bridge transition at the wrong time when current is high, causing that glitch.  Just a guess.

[Rafft]

Yes Im using 5819 there. On the 1k/470 pin#2, ive used multiturn in place of 470. I have bleeder resistor on bridge output(33k 3watts). 

Coil is running OK now (electronically).

Thanks

[Rafft]

David

I have placed all my electronics 'in a box' except for the dc-dc boost, 16v8 battery -> 330vdc. I have tried placing the C=23uF(4u7 5parallel mkp x2) on the inverter full-bridge side. It serves as the C from buck LC. Inductor is around 4inches(end to end) wire connection from inductor L -> full-bridge. And also serves as full-bridge snubber.

In my head, its 'calculated' as a snubber of single 4u7 only? Or as a whole ~23uF? Is it ok though this kind of arrangement? Though ive also soldered a single 4u7(same make/type) of the 5p placed underneath the board. Just making sure(?) I still have a proper 'snubber'  on the bridge.. So in total, C on bridge is 4u7 6p. With the single 4u7 wired (top of board)to be 'as close as possible to ecb(as snubber low lead inductance).

Added are some scope shots during phase lead tuning.
Yellow = bridge output
Magenta = Iprim

Single shot of Iprim, with ch1 probe physically detached from scope.it affects the other channel/s reading. Bottom shots show as 'tuned' , by getting a cleaner Iprim and not minding of 'spike' at bridge output

[davekni]

I have tried placing the C=23uF(4u7 5parallel mkp x2) on the inverter full-bridge side. It serves as the C from buck LC. Inductor is around 4inches(end to end) wire connection from inductor L -> full-bridge. And also serves as full-bridge snubber.

In my head, its 'calculated' as a snubber of single 4u7 only? Or as a whole ~23uF? Is it ok though this kind of arrangement? Though ive also soldered a single 4u7(same make/type) of the 5p placed underneath the board. Just making sure(?) I still have a proper 'snubber'  on the bridge.. So in total, C on bridge is 4u7 6p. With the single 4u7 wired (top of board)to be 'as close as possible to ecb(as snubber low lead inductance).

Presuming I understand correctly, this is ideal.  Wire inductance between inductor and caps is no issue at all.  Just adds a tiny bit to buck inductance.  BTW, I have ~20inches of wire from buck inductor to H-bridge on my new QCW experiment platform, with caps on H-bridge board.

I have tried placing the C=23uF(4u7 5parallel mkp x2) on the inverter full-bridge side. It serves as the C from buck LC. Inductor is around 4inches(end to end) wire connection from inductor L -> full-bridge. And also serves as full-bridge snubber.

In my head, its 'calculated' as a snubber of single 4u7 only? Or as a whole ~23uF? Is it ok though this kind of arrangement? Though ive also soldered a single 4u7(same make/type) of the 5p placed underneath the board. Just making sure(?) I still have a proper 'snubber'  on the bridge.. So in total, C on bridge is 4u7 6p. With the single 4u7 wired (top of board)to be 'as close as possible to ecb(as snubber low lead inductance).

All looks good.

[Rafft]

David

For reference I will use MBQ60T65PES

If I use IGBT internal diode(across CE), how do I determine its forward current? Can I safely assume x10 current to be its peak(qcw use)? Same as whats done/calculated on real fast diodes.

What do I look for in the datasheet?  Should I ask for more questions im not aware of?

Is this 'diode' affected by switching speed?

Pls take note IGBT gate=gnd.

[davekni]

For reference I will use MBQ60T65PES

Spec. sheet for this part looks great.  However, I don't see any in normal electronics distribution.  Haven't since I first heard of this part.  No evidence of this, but I'd guess that any parts you purchase are counterfeit.  So no way to know actual spec's.

If I use IGBT internal diode(across CE), how do I determine its forward current? Can I safely assume x10 current to be its peak(qcw use)? Same as whats done/calculated on real fast diodes.

IGBT diodes generally can't handle 10x current.  Look at spec peak current, but even that is usually for only 1ms, shorter than QCW.

Is this 'diode' affected by switching speed?

Yes, at least if you run buck inductor current in continuous mode, where current doesn't ramp all the way to 0 between pulses.  Most IGBT diodes will be fast enough.

[Rafft]

hi again

since I have switched into using (TOSLINK)fiber optic for INT(inverter) and pwm ramp(buck), I have finally acheived real isolation between int/ramp controller and inverter&buck.

my initial hardware had ramp gen GND connected to buck GND.using audio shielded wire. INTerrupt was isolated by PC817 opto only.

but I have also noticed something!

it has totally eliminated my previous "glitches" with Iprim.

BUT it has also become "less forgiving" at tuning part. like changing the MMC from 12.57nF to 9.8nF immediately showed something, whereas my -before GNDed- hardware config, those to MMC values showed little effect only(like shorter spark).

using 12.57nF result in early branching. going back 9.8nF resulted in straight sparks. it seems to "visualy" show that its more in-tune with later mmc value. scope shots below. with the "hump" (12.57nF), Vbus around 150v only. the cleaner Iprim is 220vdc(9.8nF).

any idea what that hump is?

im also planning to change my gapped-ferrite (LC buck) to powdered iron torroid. waiting for part to arrive.

edit:
those above scopeshots, was using MKP10 series-parallel connection. I also remember using CBB81 10s1p 9.6nF for mmc , and also got similar hump at around ramp-up portion. but 9.6nF(cbb81) and 9.8nF(mkp10) are close-enough values, but why no hump on mkp10? capacitor materials maybe?