Author Topic: DRSSTC I Fail  (Read 1018 times)

Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 693
  • Karma: +34/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC I Fail
« Reply #20 on: August 27, 2020, 12:34:52 AM »
Although it can work to some degree, powdered iron cores are not recommended for CTs.  Their low inductance requires more turns and lower burden resistance to avoid significant phase shift.  Powered iron core inductance/turn^2 covers a wide range, but all much below typical ferrite cores.  You could be OK, but it's hard to tell.  Something is making your scoped primary current look like IGBT switching roughly at zero-current, while the Vce waveforms clearly show switching 300+ns after zero-crossing.

If you have any scraps of coax cable around, run that from the scoping CTs out away from the coil before probing (or better, all the way to scope input).  The inductive loop of probe and ground clip picks up a lot of noise when in close to circuitry, especially compared to the relatively-low amplitude signal.

Hard to tell in the picture, but is there something else connected to the CT secondary burden resistor besides the scope probe?  Or is that black tape just insulating excess resistor lead length?

For primary CTs, do you have a single first core feeding three second-cores, one for feedback, one for OCD, and one for scoping?  The UD2.7 CT load resistance is relatively high, making powdered iron cores more likely to cause issues.  Your scope CT with 1-ohm is probably reasonable.

The GDT looks reasonable, although it can be improved.  Leakage inductance is lower if you wind with 4 twisted pairs, and use one side of each pair in parallel for the primary, rather than a single primary wire.  That may fix the issue of a residual ~8Vge spike at the end of enable.  That partial-turn-on will stress the IGBTs.

Your waveforms show sufficient dead-time, just not sufficient phase-lead.  Phase-lag effects look similar to shoot-through, but I'm fairly confident your situation is phase-lag, not shoot-through.

Thank you for snubber cap value.  That's lower than I'd guessed, so explains the higher-than-expected VBus ring frequency.
David Knierim

Offline AstRii

  • High Voltage Experimenter
  • **
  • Posts: 95
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Re: DRSSTC I Fail
« Reply #21 on: August 27, 2020, 01:08:06 AM »
I think you have transitioning noise with out using any dead time and the spikes between the turn on and off of each half bridge transition is cross conduction for 200-500ns.

Why do you think there is this rather large (200ns-500ns) cross conduction? From the waveforms i've attached i would say the most cross conduction there can be is 100ns.
I'm pretty confident that cross conduction is not that high here especially now when i see Dave having the same opinion :)

Hard to tell in the picture, but is there something else connected to the CT secondary burden resistor besides the scope probe?  Or is that black tape just insulating excess resistor lead length?

The measuring CT is grounded to mains earth, i hoped it could decrease the noise :) Otherwise nothing else is connected to it.
I only have OCD CT cascaded with 2 CTs approximately 1:500 ratio (if i remember correctly), feedback CT is single transformer with about 1:55 ratio and measuring CT is also separate ~1:50 ratio.

For now i want to fix the issue with phase-lead and afterwards i will try lower the GDT issue at the end of enable pulse.
I'm pretty sure that before the phase lead was okay (after all if it wasn't, the coil would probably died long before), i don't know why now there is this phase lag. I will try to fix it tomorrow (hopefully just adjusting the adjustable inductor on UD2.7)

I ran the coil at 80V DC bus just now, it was working well, tomorrow i will provide you with some waveforms :)

Thank you
-Mark
« Last Edit: August 27, 2020, 01:14:56 AM by AstRii »

Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 693
  • Karma: +34/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC I Fail
« Reply #22 on: August 27, 2020, 02:26:17 AM »
55:1 is a very low ratio for feedback CT.  The UD2.7 schematics I see show 51 ohm burden resistor.  250A / 55 = 4.5A, or 232V across 51 ohms, about 1kW peak.  Some resistors can handle that peak power, others (especially thick-film) will develop fine cracks and resistance rises eventually to open.  If I take a wild guess at 0.1uH/turn^2 for your iron core, that's about 300uH for 55 turns.  That's low for 51-ohms at 170kHz, but not disastrous.  It should produce a bit of extra phase-lead.  (OCD burden resistance is 2.55 ohms, so powered-iron is less likely to cause issues there.)

Overall coil performance depends very little on phase lead or lag until it gets past ~+-30 degrees or so.  Phase lag just stresses IGBTs due to forced rapid diode reverse recovery when opposing IGBT turns on as I've discussed previously (including resulting voltage spike).  Since your scoped current appears to lag behind actual current (reason still unknown to me), I'd guess you've always had phase-lag, even if it looks like lead based on the current trace.  (One possible reason is iron core losses.  Powdered iron cores intended for DC output chokes can be extremely lossy when used for AC waveforms.)

For cross-conduction, look at your gate waveforms.  There's about 100ns from when one gate drops below Vge-threshold (~5V) to when the other gate rises above Vge-threshold.  So, at least under those conditions, you have negative cross-conduction, a 100ns guard-band before just starting to get any cross-conduction.
David Knierim

Offline AstRii

  • High Voltage Experimenter
  • **
  • Posts: 95
  • Karma: +2/-0
  • Czech Technical University in Prague
    • View Profile
Re: DRSSTC I Fail
« Reply #23 on: August 28, 2020, 06:38:01 PM »
Hey guys :)
Sorry for my late response, i'm currently having a lot of studying for my Calculus II exam in tuesday :D

So i've tried to get the phase-lead to work better.
This is how it was before:



And this is the best i could get from it:



There is a very slight (almost non-existing) difference, but when you zoom up close, you can see the two cursors measuring the one half cycle to be 2.07us



So i would expect, if the phase lead is correct, you would measure peak current at the exact half of this time after the start of the cycle (1.035us)
And that seems to be (almost) the case.



I would love to check if the phase lead is correct by simply checking the zero crossing of the waveforms, but the noise on the primary current makes it impossible for me.

The phase lead acts weird, i tried few more measurements after some time (few minutes/hours) and it was off again for no apparent reason.
It's maybe because i'm using these iron powder CTs, maybe not. I have no idea why it acts this strange.

Anyway i know the phase-lead is very far from perfect, but anyway should i be even worried about it?
It's also not so horrible, it's maybe not switching at zero current, but i would say pretty close to zero current.
Since those are 1200V IGBTs and i'm planning to run maximum current of 300A (still withing datasheet pulse rating of those IGBTs), they should not fail or am i wrong?

55:1 is a very low ratio for feedback CT.

Sorry Dave, i was incorrect. The feedback CT ratio is more like 1:200. I don't really know, it's been more than a year since i put those CTs in there :D
Anyway if you look closely at the feedback CT, it seems there is about 200 turns.



Still too low ratio, i should really upgrade those CTs to ferrite cores (would probably fix most of my issues :D), but i really want this coil to be finally finished.



Offline davekni

  • High Voltage Senior
  • *****
  • Posts: 693
  • Karma: +34/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC I Fail
« Reply #24 on: August 28, 2020, 09:40:12 PM »
Engineers have a reputation of being perfectionists.  Often "good enough" is the right answer, rather than wasting more time and resources making something a little closer to perfect.

So, yes, my guess is that you'll be fine with 1200V IGBTs.  The phase-lag is not enough to effect coil output, and voltage spikes aren't likely to reach 1200V or even 880V of your TVS pairs.  The 8V Vge ring at the end of enable pulses looks short enough to not likely fry IGBTs either.

If you did manage to achieve the desired small phase-lead, the current measurement noise will go down, since diodes are no longer being hard-forced off.  (Does your scope have a 20MHz bandwidth-limit option as many do?  That will filter out much of the noise for better current viewing.)  For detailed phase checking, I find it easier to use Vge and Vce waveforms.  With phase-lag, Vce looks like your waveforms, dropping below ground before switching.  With phase-lead, Vce will go below ground just after the IGBT turns-on rather than just before it turns-off.  At 0 degrees (where zero-current occurs during the dead-time when no IGBTs are conducting), Vce will have a ring during transition.  Bottom IGBT turns off, current starts pulling Vce up, current reverses and starts pulling Vce back down, then high-side IGBT turns on pulling low-side Vce back up to VBus.

Noticed one other thing, that 1200V IGBTs should also handle:  Your VBus ring (ring on high parts of H-Bridge output waveforms) is close to 2x primary coil frequency.  2x primary is the dominant ripple-current component (full-wave-rectified primary current waveform).  A slight change in primary frequency might make VBus ring match 2x even more closely, which can result in extremely large VBus ringing, building up from cycle to cycle.  See my reply 36 of the thread linked below for a bit more on this 2x VBus ring issue, and previous posts in that thread showing the resulting large ripple:
https://highvoltageforum.net/index.php?topic=1058.msg8182;topicseen#new

Good luck with your test and with your coil!
David Knierim

High Voltage Forum

Re: DRSSTC I Fail
« Reply #24 on: August 28, 2020, 09:40:12 PM »

 


* Recent Topics and Posts

post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
panjaksli
December 02, 2020, 08:57:34 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 02, 2020, 08:17:09 PM
post Re: mini tesla (hopefully not a fail)
[Beginners]
davekni
December 02, 2020, 08:12:55 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
panjaksli
December 02, 2020, 07:34:53 PM
post Re: HF Litz wire as a primary coil
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 02, 2020, 07:14:18 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 02, 2020, 07:05:18 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
panjaksli
December 02, 2020, 06:58:40 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Max
December 02, 2020, 06:56:50 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
profdc9
December 02, 2020, 06:20:35 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
panjaksli
December 02, 2020, 05:24:23 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
buchtawill
December 02, 2020, 03:25:59 PM
post Re: Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
panjaksli
December 02, 2020, 03:02:01 PM
post Have you ever tripped a breaker during normal operation ?
[Dual Resonant Solid State Tesla coils (DRSSTC)]
panjaksli
December 02, 2020, 01:36:04 PM
post Re: mini tesla (hopefully not a fail)
[Beginners]
bogdan
December 02, 2020, 08:55:06 AM
post Re: HF Litz wire as a primary coil
[Dual Resonant Solid State Tesla coils (DRSSTC)]
johnf
December 02, 2020, 07:43:14 AM
post Re: How to design FPGA based on EDA technology?
[Electronic Circuits]
Genterman
December 02, 2020, 06:53:52 AM
post Re: DRSSTC trouble
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 02, 2020, 03:29:14 AM
post Re: DRSSTC trouble
[Dual Resonant Solid State Tesla coils (DRSSTC)]
buchtawill
December 02, 2020, 02:43:33 AM
post Re: SKP DRSSTC = aliexpress driver + medium coil + large bricks ;)
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Maju
December 01, 2020, 10:44:31 PM
post Re: mini tesla (hopefully not a fail)
[Beginners]
davekni
December 01, 2020, 07:44:14 PM
post Re: transistor markings question
[Beginners]
davekni
December 01, 2020, 07:39:48 PM
post Re: HF Litz wire as a primary coil
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 01, 2020, 07:38:53 PM
post Re: How to design FPGA based on EDA technology?
[Electronic Circuits]
petespaco
December 01, 2020, 06:30:33 PM
post Re: Question about arduino polyphonic MIDI interrupter
[Dual Resonant Solid State Tesla coils (DRSSTC)]
bozidar
December 01, 2020, 04:58:29 PM
post Re: Question about arduino polyphonic MIDI interrupter
[Dual Resonant Solid State Tesla coils (DRSSTC)]
TMaxElectronics
December 01, 2020, 04:22:10 PM
post Re: Question about arduino polyphonic MIDI interrupter
[Dual Resonant Solid State Tesla coils (DRSSTC)]
bozidar
December 01, 2020, 03:02:51 PM
post Re: Half-brigde with pot on mosfet gates question
[Solid State Tesla Coils (SSTC)]
TMaxElectronics
December 01, 2020, 12:08:56 PM
post transistor markings question
[Beginners]
NOOBPASTE51
December 01, 2020, 12:02:19 PM
post Re: Question about arduino polyphonic MIDI interrupter
[Dual Resonant Solid State Tesla coils (DRSSTC)]
TMaxElectronics
December 01, 2020, 11:55:01 AM
post Re: HF Litz wire as a primary coil
[Dual Resonant Solid State Tesla coils (DRSSTC)]
TMaxElectronics
December 01, 2020, 11:50:19 AM
post Re: Analog HFBR amplifier
[Electronic Circuits]
TMaxElectronics
December 01, 2020, 11:21:56 AM
post Half-brigde with pot on mosfet gates question
[Solid State Tesla Coils (SSTC)]
costas_p
December 01, 2020, 11:17:16 AM
post Re: Question about arduino polyphonic MIDI interrupter
[Dual Resonant Solid State Tesla coils (DRSSTC)]
bozidar
December 01, 2020, 11:11:50 AM
post Re: mini tesla (hopefully not a fail)
[Beginners]
bogdan
December 01, 2020, 09:29:24 AM
post Re: HF Litz wire as a primary coil
[Dual Resonant Solid State Tesla coils (DRSSTC)]
johnf
December 01, 2020, 07:03:56 AM
post Re: How to design FPGA based on EDA technology?
[Electronic Circuits]
johnf
December 01, 2020, 06:49:26 AM
post Re: SSTC or DRSSTC with GaN Transistors
[Solid State Tesla Coils (SSTC)]
Steve Ward
December 01, 2020, 04:59:28 AM
post How to design FPGA based on EDA technology?
[Electronic Circuits]
Genterman
December 01, 2020, 04:02:06 AM
post Re: DRSSTC trouble
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 01, 2020, 03:46:10 AM
post Re: DRSSTC trouble
[Dual Resonant Solid State Tesla coils (DRSSTC)]
buchtawill
December 01, 2020, 02:00:54 AM
post Re: HF Litz wire as a primary coil
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Steve Ward
December 01, 2020, 01:47:10 AM
post Re: DRSSTC trouble
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 01, 2020, 12:02:23 AM
post Re: Site Info desperately needed: http://www.stevehv.4hv.org/
[General Chat]
Mads Barnkob
November 30, 2020, 10:28:56 PM
post Re: I started making a drsstc
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Miki_407
November 30, 2020, 08:58:41 PM
post DRSSTC trouble
[Dual Resonant Solid State Tesla coils (DRSSTC)]
buchtawill
November 30, 2020, 08:53:00 PM
post Re: I started making a drsstc
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
November 30, 2020, 08:49:30 PM
post Re: Simple H-Bridge construction with low parasitic inductances (for SSTC or ...)
[Beginners]
davekni
November 30, 2020, 08:45:46 PM
post Re: mini tesla (hopefully not a fail)
[Beginners]
davekni
November 30, 2020, 06:56:13 PM
post Re: I started making a drsstc
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Miki_407
November 30, 2020, 01:16:04 PM
post Re: I started making a drsstc
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
November 30, 2020, 12:56:05 PM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 
SimplePortal 2.3.6 © 2008-2014, SimplePortal