Author Topic: SG3525 push pull snubber calculation  (Read 629 times)

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
SG3525 push pull snubber calculation
« on: November 13, 2021, 07:03:10 PM »
First hi all, and sorry for my english  :)

I built test SG3525 push pull inverter from chinese SG3525 LM358 Inverter Driver Board module and pair of IRFP260N, drain-source is protected by transil 1.5KE200CA, now waiting parcel with R-C snubber parts.
HV transformer based on UF86A ferrite core, homemade primary & secondary - 5T+5T/700T.
Driver has separate 15V power supply.
Main power supply is 600W transformer, under load(450W) after rectifying -35V.
Transistor heating is moderate, after 1min case temperatue is 45*C.(one large heatsink for rectifier and transistors)
In future i add EMI filter to prevent high frequency going in home mains line.



Connected oscilloscope and get ugly waveforms  :-\
After looking on drain-source waveform i think that 1.5KE200CA is not working- voltage spike more than 200V, or it is normal for a short period ?

Drain-Source


Gate, driver only working


Gate, 450W power consumption, load - hv arc


Main DC bus, on load


220/32V transformer, secondary, on load


For RC snubber calculation i used method from google -" For a RC snubber, you want the C to be about three times the parasitic Cs and R to be equal to sqrt(Ls/Cs)", where Cs - transistor output capacitance, Ls - internal source inductance, this is correct ?
Any things to remove spikes ?
« Last Edit: November 13, 2021, 07:04:59 PM by PowerTech »

Offline Weston

  • High Voltage Technician
  • ***
  • Posts: 112
  • Karma: +8/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #1 on: November 13, 2021, 09:10:07 PM »
Ls here is going to be predominately the leakage inductance of the transformer, which you can measure by shorting the secondary and measuring the inductance of the primary.

A RC snubber or the TVS is both going to dissipate energy. Unless it's dissipating too much energy the TVS snubber should work by itself. The RC snubber will dampen the ringing and likely have lower EMI. The TVS snubber will just clamp the peak voltage and likely have a bit lower total power dissipation than the RC snubber.

The issue with the current TVS snubber is that the 1.5KE200CA has a clamping voltage above 200V. If you look at the datasheet the breakdown voltage, where 1mA is conducted, is 190-210V, while the clamping voltage, when the device sinks 5.5A is 274V. Based on the current it's going to be somewhere between those two values.

Your DC bus is also spiking up with the switching transitions, you could also benefit from better layout between the DC filter capacitor to the transformer and the return from the FET source to the capacitor / added film snubber capacitors. This would reduce the stray inductance.
« Last Edit: November 13, 2021, 09:37:14 PM by Weston »

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #2 on: November 13, 2021, 10:13:03 PM »
Thanks for the reply, need revise dc bus connections for optimal way, also I leave TVS and RC snubber together.

Your DC bus is also spiking up with the switching transitions, you could also benefit from better layout between the DC filter capacitor to the transformer and the return from the FET source to the capacitor / added film snubber capacitors. This would reduce the stray inductance.
Not shown in schematics, 3uF film capacitor is connected to dc bus electrolyte.

Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1386
  • Karma: +67/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #3 on: November 14, 2021, 12:49:13 AM »
As Weston mentioned, transformer leakage inductance is the source of energy that must be clamped.  For this topology, leakage inductance between one primary 5T winding and the other 5T winding is most important.  (To measure, short one primary and measure inductance of the other primary winding.)  Energy stored in leakage inductance between primary and secondary is mostly returned to the DC supply through the conducting FET, so not of as much concern.

To minimize primary leakage inductance, wind the 5 turns bifilar.  Wind two parallel wires for 5 turns.  Even better, use many pairs of smaller wire, wind 5 turns of the bundle, then split out wires of each pair, one for each primary winding.  Also, lead length from transformer to inverter board should be as short as is reasonable, with leads of each winding paired together (twisted or taped together or ...) to minimize lead inductance.  Lead inductance can be a significant portion of leakage inductance.  (Even though GDTs are low-power, looking at GDT designs may help, as low leakage inductance is important there too.)

Also as Weston said, 270V is normal for 1.5KE200CA clamping high current.  Chances are that the IRFP260 FETs are handling half or more of the current during clamping.  IRFP260 is rated for 28mJ repetitive clamping energy, and 46A clamping current, so should be fine clamping even without TVS diodes.  However, coming from China, the IRFP260 parts are very likely counterfeit.  IRFP260 counterfeit parts are widely available at much lower cost than genuine parts.  Probably a range of qualities of counterfeit parts, from junk to somewhat-close to real IRFP260 capabilities.  Obviously yours are working at least reasonably.
« Last Edit: November 14, 2021, 01:14:38 AM by davekni »
David Knierim

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #4 on: November 14, 2021, 06:35:06 PM »
Quote
For this topology, leakage inductance between one primary 5T winding and the other 5T winding is most important.  (To measure, short one primary and measure inductance of the other primary winding.)
With this method Ls=0.014mH, so calculate RC snubber -
Cs = 603pF = 0.603nF
Csnubber = 603*3.14=1809pf=1.893nF
Ls = 0.014mH = 14000nH
Rsnubber = sqrt(Ls/Cs) = sqrt(14000/0.603) = 152R
My calculations are absolutely incorrect, so when parcel arrives, I try with 10R+1nF, 10R+2nF and chose best by waveform.

Transformer don’t have any gaps between core halves, removed them after using in "ZVS driver".
Primary winding is bifilar, made from thick Cu wire, approx. 6mm*2, diy litz wire is almost ready - difficulties to qualitative remove enamel isolation from multiple wires.
IRFP260 is purchased long time ago from trusted local shop, IOR original, they worked perfectly in "ZVS driver" at ~1kW with many unwanted inductances - long wires in assembly.

Test inverter assembly




Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1386
  • Karma: +67/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #5 on: November 14, 2021, 09:30:35 PM »
Thank you for pictures.  That helps.  Nice job with bifilar primary winding and short leads!

From reply #2:
Quote
Not shown in schematics, 3uF film capacitor is connected to dc bus electrolyte.
It will help greatly if you move that 3uF film capacitor to adjacent the FETs and terminals connecting to transformer.  Wiring inductance of 35Vbus supply from coil/FETs to capacitors is likely adding more parasitic inductance than the transformer leakage inductance.  Any inductance in the transformer center-tap lead (including Vbus wiring) counts 4x compared to inductance in individual transformer wires.  Impedance is transformed 4:1 by the 2:1 turns-ratio.  BTW, you can see the effect of Vbus inductance in your "Main DC bus, on load" scope trace.  Moving the scope probe and ground points out to the transformer CT terminal and FET source will show the full effect of Vbus inductance.

Quote
With this method Ls=0.014mH
Looking at your pictures, I estimate primary leakage inductance to be between 700nH and 1uH, nowhere close to 14uH.  14uH must be mostly meter leads or some other measurement issue.  Even if 1uH, moving the film capacitor (or even adding multiple capacitors) at the FETs will be more important than any further transformer improvements.

450W/35V = 13A average DC supply current.  Peak FET current will be higher, perhaps 25-40A.  Even if you get leakage inductance + 4X Vbus inductance down to 1uH total, that is still 40A*40A*1uH/2=800uJ energy to clamp.  FETs should be OK with that, adding 800uJ * 30kHz = 24W additional heat to each FET.  A 2nF capacitor would need to charge to 900V in order to absorb 800uJ.  R/C snubbers will help with high-frequencies (ie. for EMI compliance).  TVS and/or FET avalanche clamping is more efficient for dissipating the primary leakage inductance energy.
« Last Edit: November 14, 2021, 09:32:49 PM by davekni »
David Knierim

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #6 on: November 14, 2021, 10:05:10 PM »
Thank U very much, great explanation, lot of useful information.

Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1386
  • Karma: +67/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #7 on: November 15, 2021, 03:57:17 AM »
Quote
It will help greatly if you move that 3uF film capacitor to adjacent the FETs and terminals connecting to transformer.
There will be one side-effect of moving the 3uF film capacitor.  That creates a new resonance of 3uF with wiring inductance to your 10000uF bulk capacitor(s).  You will see some ring on Vbus (measured at the FETs) at that resonant frequency.  It will become a problem if that resonance happens to be close to twice 30kHz operating frequency.  The ripple current drawn from VBus is dominated by twice switching frequency, so 60kHz in your case.

BTW, DRSSTCs can have the same issue of local-to-IGBT film capacitors on Vbus resonating with wire/bus-bar inductance to bulk Vbus capacitors.  If you are interested, here's a thread showing the issue with getting close to twice operating frequency:
https://highvoltageforum.net/index.php?topic=1820.msg13736#msg13736
Click on scope links in first post to see resulting waveforms.

My preferred solution is to drastically reduce parasitic inductance using overlapping copper planes to connect bulk capacitors to FETs and transformer center-tap.  This can be done with a scrap of double-sided copper-clad FR4, or copper foil tape on either side of a plastic sheet.  Here's a link for low parasitic inductance thread that might be useful, even though the examples there are half and full H-bridges:
https://highvoltageforum.net/index.php?topic=1324.msg9795#msg9795
If you want to go to the extreme in reducing switching spikes (reducing parasitic inductance), you could also wind the primary with copper foil.  Probably not worth the effort unless you are motivated to experiment with transformer magnetic design.

Another option if the 3uF resonance is problematic:  Add a snubber R+C across the 3uF cap at the FETs.  Your previous snubber calculations would be appropriate to use here, where Cs is now 3uF.  Ls can be calculated from the scope-measured Vbus ring frequency and 3uF.  Here Ls is the wiring inductance from 3uF to bulk 10000uF.  Snubber power dissipation can be reduced by using larger snubber capacitor and lower snubber resistor value.
« Last Edit: November 15, 2021, 04:17:07 AM by davekni »
David Knierim

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #8 on: November 18, 2021, 09:10:14 PM »
Quote
It will help greatly if you move that 3uF film capacitor to adjacent the FETs and terminals connecting to transformer.
I leaved 3uF cap on electrolytes + added 3.3uF 400V between primary center tap and FETs negative(ground), after powering up both capacitor legs turn red color and destructed in 10sec, resonance ?
This is Chinese CBB22 film capacitor, so quality issues may occur.



Comparison of voltage spikes on "right" FET, without and with film cap near FETs and transformer.


Also found significant difference between FET1 & FET2 drain-source voltage spikes.[without cap]
Drain- Source FET1, Drain-Source FET2, maybe influenced by left FET gate resistor crossing power GND, and/or different parasitic inductance.


Gate FET1, Gate FET2 signal


Quote
My preferred solution is to drastically reduce parasitic inductance using overlapping copper planes to connect bulk capacitors to FETs and transformer center-tap.
This is next to be done in my inverter, at least something near your method.

For now HV arc is not so good as in my ~1kW ZVS, later, when spikes will be less i try 60V DC in.
« Last Edit: November 18, 2021, 10:45:03 PM by PowerTech »

Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1386
  • Karma: +67/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #9 on: November 19, 2021, 12:25:48 AM »
Quote
I leaved 3uF cap on electrolytes + added 3.3uF 400V between primary center tap and FETs negative(ground), after powering up both capacitor legs turn red color and destructed in 10sec, resonance ?
This is Chinese CBB22 film capacitor, so quality issues may occur.
Wow.  Capacitor failure isn't overly surprising, but failure by leads burning out is very unusual.  Can you tell if the leads are made of copper as is spec'ed for CBB22 - scrape away whatever tin plating may be on the outer layer of the leads?  Low-power through-hole parts often use tin-plated steel leads, so you could check with a magnet too.  Looking at your waveforms, I'd make a very-crude estimate of 10-15Arms through that 3uF capacitor.  Couldn't find a rating for that capacitor, but wouldn't expect 10 second life.

Quote
Also found significant difference between FET1 & FET2 drain-source voltage spikes.[without cap]
Drain- Source FET1, Drain-Source FET2, maybe influenced by left FET gate resistor crossing power GND, and/or different parasitic inductance.
Yes, wiring inductances seem like most likely explanation.  Hard to tell what is cause and effect:  Gate differences could cause drain differences or visa-versa.  One other possibility is one of the TVS diodes is fried open.

Quote
This is next to be done in my inverter, at least something near your method.
Besides improving performance, low parasitic inductance will help with debug, making scope measurements easier to interpret.

Quote
or now HV arc is not so good as in my ~1kW ZVS, later, when spikes will be less i try 60V DC in.
What frequency and voltage and number of primary turns were you using for ZVS?  While drawing an arc, output voltage is relatively low.  Performance is more related to current, which is controlled by input voltage / impedance of transformer primary-to-secondary leakage inductance (I ~= turns_ratio * V / (L * 2 * PI * F)).
David Knierim

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #10 on: November 19, 2021, 01:02:50 PM »
Quote
Can you tell if the leads are made of copper as is spec'ed for CBB22
Non copper, iron like, magnetic.
Found Epcos 1uF 1000V film cap, screw mount, Cu connectors.

Quote
What frequency and voltage and number of primary turns were you using for ZVS?  While drawing an arc, output voltage is relatively low.  Performance is more related to current, which is controlled by input voltage / impedance of transformer primary-to-secondary leakage inductance (I ~= turns_ratio * V / (L * 2 * PI * F)).
As I remember f=30-35kHz, 45Vdc at 20A, 3T+3T primary, frequency is calculated from measured contour cap capacitance and primary inductance, I never connected oscilloscope to ZVS, used 2x IRFP250N parallel on each side, with common gate resistor (this is wrong).
With these 3+3 turns I hit saturation - primary heating very fast, later calculated primary via ExcellentIT software - 5+5 turns for 35kHz 50-60Vdc in.
When replaced primary to 5+5 and decreased contour cap,  also increased filtering cap to approx. 28mF, at this time with no measurements, but there is no more 50hz sound when arcing -
DC filtering is enough.




« Last Edit: November 19, 2021, 01:16:46 PM by PowerTech »

Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1386
  • Karma: +67/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #11 on: November 20, 2021, 12:43:50 AM »
Quote
Non copper, iron like, magnetic.
Found Epcos 1uF 1000V film cap, screw mount, Cu connectors.
Red hot makes sense now.  Magnetic materials have extremely-thin skin-depth.  Most current will be flowing in the outer tin plating layer.

Quote
s I remember f=30-35kHz, 45Vdc at 20A, 3T+3T primary, frequency is calculated from measured contour cap capacitance and primary inductance, I never connected oscilloscope to ZVS, used 2x IRFP250N parallel on each side, with common gate resistor (this is wrong).
With these 3+3 turns I hit saturation - primary heating very fast, later calculated primary via ExcellentIT software - 5+5 turns for 35kHz 50-60Vdc in.
When replaced primary to 5+5 and decreased contour cap,  also increased filtering cap to approx. 28mF, at this time with no measurements, but there is no more 50hz sound when arcing -
DC filtering is enough.
If I understand correctly, you are aiming for roughly the same parameters with this push-pull driver.  Seems reasonable.  At the same DC bus voltage and same frequency, peak core magnetic field should be the same between push-pull and ZVS circuits.  Square-wave vs. sine-wave voltage will have a small effect.  One possibly more-significant difference:  The ZVS was likely oscillating at a higher frequency when the arc was short (low impedance secondary load).  Dominant inductance will be primary-to-secondary leakage inductance rather than overall parallel (open-circuit) inductance.  With fixed-frequency push-pull drive, you may find the driver gets overloaded when the arc is short.  Will be interesting to see once you resolve parasitic inductance.
David Knierim

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #12 on: November 20, 2021, 11:34:49 PM »
Today fully reassembled inverter, most important changes:
- decreased parasitic inductance
- now 2xIRFP260N in each side
- filtering capacitor increased to ~ 18800uF
- TVS diodes replaced by new ones

Done fast test I’m noticed significantly less FETs heating at 500W.
Thank U davekni for great help.



Old assembly vs new, "left" FET drain-source.


Drain-source waveforms become almost identical on both sides.


Drain-source, holding by hand 1uF snubber film capacitor between primary center tap and FETs negative.
Very similar to this situation -
Quote
BTW, DRSSTCs can have the same issue of local-to-IGBT film capacitors on Vbus resonating with wire/bus-bar inductance to bulk Vbus capacitors.  If you are interested, here's a thread showing the issue with getting close to twice operating frequency:
https://highvoltageforum.net/index.php?topic=1820.msg13736#msg13736
Click on scope links in first post to see resulting waveforms.
« Last Edit: November 20, 2021, 11:46:37 PM by PowerTech »

Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1386
  • Karma: +67/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #13 on: November 21, 2021, 05:00:10 AM »
Quote
Today fully reassembled inverter
Definitely looks more symmetric.
If gate waveforms are good, there is no need to change this:  The gate shielding is likely not of value unless the left ends are connected to FET source pins.  Shielding connected at one end only helps with capacitive-coupled noise, but not inductive-coupled.  Impedances are low enough enough that inductive-coupling is the only significant concern here.

Quote
Drain-source, holding by hand 1uF snubber film capacitor between primary center tap and FETs negative.
This clear ring signal and known 1uF capacitance allows calculation of Vbus inductance and a rough estimate of current at the switching time.  Scoping bus voltage at the added cap would make it even more clear.  Ring should be half the voltage amplitude at Vbus relative to a FET drain.  Best I can estimate from the scope trace, ring frequency is about 270KHz.  L = 1 / (4 * PI * PI * 270kHz * 270kHz * 1uF) = 0.35uH.  This is inductance of wiring from 1uF snubber to bulk caps.  Most is where wires are farther apart, across the bridge rectifier and then out of the image to caps.

Extrapolating ring amplitude back to the switching point, looks to be about 70V peak, or 35V peak at Vbus.  Impedance of the 1uF cap at 270kHz is 1uF / (2 * PI * 270kHz) = 0.59 ohms.  35V / 0.59 ohms = 59A.  This peak ripple is caused by peak-to-peak current.  Implies Vbus current is switching from 29.5A to -29.5A (29.5A FET current at turn-off point).  Actual current is likely slightly higher.  Ring hasn't damped out before switching and appears to be opposite phase.  So the actual change in Vbus ring is likely from about -10V to +35V, for a 45V peak rather than 35V, implying ~38A FET current at turn-off point.  These are all rough estimates.

Vbus construction at the FETs is great.  Just need parallel planes from there to bulk capacitors.  Or large capacitor(s) at the FETs.
« Last Edit: November 21, 2021, 05:01:42 AM by davekni »
David Knierim

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #14 on: November 27, 2021, 06:00:42 PM »
Quote
Just need parallel planes from there to bulk capacitors.  Or large capacitor(s) at the FETs.
Done, +3uF snubber, also replaced driver.
https://aliexpress.ru/item/4000448800664.html?sku_id=10000001835722634&spm=a2g2w.productlist.0.0.73ff1ae3Em2Z5c



Quote
L = 1 / (4 * PI * PI * 270kHz * 270kHz * 1uF) = 0.35uH.
For this run used 1uF film capacitor.
New L = 1 / (4 * PI * PI * 500kHz * 500kHz * 1) = 0.10uH

For this run used 1uF film capacitor.


Old vs new drain-source, no snubber capacitor.


Drain-source 3uF snubber.


DC bus near FETs.


Gate 1 & 2.


Gate & Drain-Source, same FET
« Last Edit: November 27, 2021, 06:51:48 PM by PowerTech »

Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1386
  • Karma: +67/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #15 on: November 27, 2021, 08:51:05 PM »
Looks good!

Quote
For this run used 1uF film capacitor.
New L = 1 / (4 * PI * PI * 500kHz * 500kHz * 1) = 0.10uH
This 0.10uH includes inductance of the 1uF capacitor and its leads.  Inductance to bulk caps will be slightly under 0.10uH.  Perhaps 0.08uH to bulk caps and 0.02uH for 1uF cap and leads as a guess.

Quote
DC bus near FETs.
Seeing this plot allows one more calculation.  Switching pulse goes from ~36V up to ~63V, for a 27V delta.  Since this is the transformer primary center tap, 27V pulse here corresponds to 54V pulse on FET drain.  The measured FET drain pulse is ~72V baseline (with FET off) to ~235V peak, for a ~163V delta.  Of this 163V delta, 54V is from DC bus, leaving 109V due to transformer primary leakage inductance.  So, roughly, your switching losses are now 1/3rd due to remaining DC bus inductance and 2/3rds due to transformer leakage inductance.

Transformer leakage inductance can also be estimated from above.  DC bus inductance is transformed by 4x due to 2x turns ratio, so about 0.08uH * 4 = 0.32uH.  Leakage inductance is (109V / 54V) * 0.32uH = 0.65uH.  Of course, this is somewhat rough estimate, as I'm presuming 0.02uH for inductance of 1uF cap/leads.

BTW, the FET DS pulse voltage is lower now (235V rather than previous 270V) due to FETs running cooler.  The parasitic inductance reductions reduce pulse width, but not amplitude directly.  Amplitude is based on FET avalanche breakdown voltage, which increases with temperature.  Pulse voltage is one way to monitor FET internal die temperature.

How is performance with your clean upgrades?
David Knierim

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #16 on: November 27, 2021, 09:40:34 PM »
Quote
How is performance with your clean upgrades?
A lot better - FETs are much cooler, max power consumption is 550W, now is time switch to 1.2kW 60V DC power supply  :)
Thank U again davekni !!!

Offline PowerTech

  • High Voltage Enthusiast
  • *
  • Posts: 10
  • Karma: +0/-0
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #17 on: November 28, 2021, 03:16:11 PM »
Finally run it on 60V DC in, FETs temperature after 30s - 30*C, snubber capacitors same, device working nice.

ZVS power achieved, on same power supply, at peak 993W load Vdc drops to ~50V.(measured when run ZVS at same power)



Drain-source @ 660W in.


Gate


Trying to google FET gate resistor calculations, found that most people trying different values to find right one.
Found some formulas, but it's useless - no needed values in datasheet.
So It's worth to replace 10R with 4.7R, turn on time will be little faster, less switch losses ?
« Last Edit: November 28, 2021, 11:08:56 PM by PowerTech »

Offline davekni

  • High Voltage Expert
  • ******
  • Posts: 1386
  • Karma: +67/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: SG3525 push pull snubber calculation
« Reply #18 on: November 28, 2021, 11:58:24 PM »
Quote
Trying to google FET gate resistor calculations, found that most people trying different values to find right one.
Found some formulas, but it's useless - no needed values in datasheet.
So It's worth to replace 10R with 4.7R, turn on time will be little faster, less switch losses ?
For your particular case, gate waveforms look fine, so I'd leave it as is.  Your switching losses are dominated by power transformer leakage inductance and the resulting avalanche breakdown clamping of turn-off voltage spikes.  Switching speed related losses will be comparatively insignificant.

In general, optimal gate resistance depends on many tradeoffs, so is difficult to calculate in advance.  For GDT-driven gates, damping the ring (overshoot and undershoot) caused by GDT leakage inductance is usually the key consideration.  A good low-leakage-inductance GDT and wiring allows lower gate resistance and therefore faster switching.  Even without a GDT, long gate wires can have enough inductance to need damping with gate resistance.

The second common consideration is intentional slowing of switching times.  For bridge designs, parasitic inductance causes voltage spikes that otherwise are theoretically clamped by the opposing FET/IGBT's internal diode.  Slowing switch-off speed reduces such voltage spikes.  Slowing switch-on speed reduces diode reverse-recovery current spikes when switching after zero-current.

Third, in typical hobby bridge designs, dead-time (cross-conduction prevention) time is provided by gate resistors paralleled with diodes to speed-up turn-off relative to turn-on.  Here the value is determined by the needed dead-time (how slow the IGBT's turn-off is relative to turn-on).
David Knierim

High Voltage Forum

Re: SG3525 push pull snubber calculation
« Reply #18 on: November 28, 2021, 11:58:24 PM »

 


* Recent Topics and Posts

post Re: help with finding resonant frequency
[Beginners]
HVuser
Today at 04:43:39 AM
post Re: Looking to build by first DRSSTC as a high voltage beginner
[Dual Resonant Solid State Tesla coils (DRSSTC)]
futurist
Today at 01:37:46 AM
post Re: Practical limitations in size for Marx or CW
[Voltage Multipliers]
abstruse1
December 02, 2021, 11:35:39 PM
post Re: Practical limitations in size for Marx or CW
[Voltage Multipliers]
MRMILSTAR
December 02, 2021, 09:53:13 PM
post Re: gate drive transformer method
[Dual Resonant Solid State Tesla coils (DRSSTC)]
thedoc298
December 02, 2021, 05:10:30 AM
post Re: gate drive transformer method
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 02, 2021, 04:58:53 AM
post Re: Syntherrupter - A Feature-rich, Polyphonic Interrupter
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Max
December 02, 2021, 12:47:10 AM
post Re: gate drive transformer method
[Dual Resonant Solid State Tesla coils (DRSSTC)]
thedoc298
December 01, 2021, 10:30:50 PM
post Re: gate drive transformer method
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
December 01, 2021, 09:10:11 PM
post Re: gate drive transformer method
[Dual Resonant Solid State Tesla coils (DRSSTC)]
thedoc298
December 01, 2021, 06:36:14 AM
post Re: DRSSTC build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
November 30, 2021, 08:10:17 PM
post DRSSTC build
[Dual Resonant Solid State Tesla coils (DRSSTC)]
balazs
November 30, 2021, 06:55:58 PM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
November 30, 2021, 06:23:03 AM
post Re: Help on my "first" SSTC (LabCoatz SSTC 2.0)- Popping transistors
[Solid State Tesla Coils (SSTC)]
davekni
November 30, 2021, 05:58:42 AM
post Re: gate drive transformer method
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
November 30, 2021, 05:35:58 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ItsChloeUwU
November 30, 2021, 01:50:18 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
futurist
November 30, 2021, 01:19:18 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ItsChloeUwU
November 30, 2021, 12:07:59 AM
post gate drive transformer method
[Dual Resonant Solid State Tesla coils (DRSSTC)]
thedoc298
November 29, 2021, 11:45:44 PM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ItsChloeUwU
November 29, 2021, 10:52:43 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Intra
November 29, 2021, 03:15:23 PM
post Re: Help on my "first" SSTC (LabCoatz SSTC 2.0)- Popping transistors
[Solid State Tesla Coils (SSTC)]
TiagoBS
November 29, 2021, 03:08:58 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Netzpfuscher
November 29, 2021, 12:15:12 PM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
November 29, 2021, 12:26:28 AM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
davekni
November 29, 2021, 12:07:30 AM
post Re: SG3525 push pull snubber calculation
[Transformer (Ferrite Core)]
davekni
November 28, 2021, 11:58:24 PM
post GDT (Gate Drive Transformer) tutorial
[Beginners]
davekni
November 28, 2021, 11:40:56 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
MRMILSTAR
November 28, 2021, 10:21:11 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
johnnyzoo
November 28, 2021, 10:05:40 PM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Mads Barnkob
November 28, 2021, 08:13:08 PM
post Re: Infrastructure and Architecture
[General Chat]
futurist
November 28, 2021, 06:12:44 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
MRMILSTAR
November 28, 2021, 04:48:04 PM
post Re: SG3525 push pull snubber calculation
[Transformer (Ferrite Core)]
PowerTech
November 28, 2021, 03:16:11 PM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ItsChloeUwU
November 28, 2021, 09:36:10 AM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
plasma
November 28, 2021, 05:32:16 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
November 28, 2021, 04:28:33 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ItsChloeUwU
November 28, 2021, 02:19:00 AM
post Re: Where to buy DIY or home made capacitor book(s)?
[Capacitor Banks]
huntergroundmind
November 28, 2021, 01:58:12 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
November 28, 2021, 12:53:27 AM
post Re: Driverless SSTC
[Solid State Tesla Coils (SSTC)]
davekni
November 28, 2021, 12:32:09 AM
post Re: Help on my "first" SSTC (LabCoatz SSTC 2.0)- Popping transistors
[Solid State Tesla Coils (SSTC)]
davekni
November 28, 2021, 12:16:04 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ItsChloeUwU
November 27, 2021, 11:42:08 PM
post Re: Hi! 811 A tesa coil
[Vacuum Tube Tesla Coils (VTTC)]
304er
November 27, 2021, 11:32:12 PM
post Re: Hi! 811 A tesa coil
[Vacuum Tube Tesla Coils (VTTC)]
jpvvv123
November 27, 2021, 11:09:43 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
paulj
November 27, 2021, 10:06:41 PM
post Re: SG3525 push pull snubber calculation
[Transformer (Ferrite Core)]
PowerTech
November 27, 2021, 09:40:34 PM
post Re: SG3525 push pull snubber calculation
[Transformer (Ferrite Core)]
davekni
November 27, 2021, 08:51:05 PM
post Re: Infrastructure and Architecture
[General Chat]
Da_Stier
November 27, 2021, 08:09:50 PM
post Re: SG3525 push pull snubber calculation
[Transformer (Ferrite Core)]
PowerTech
November 27, 2021, 06:00:42 PM
post Re: Someone built Big TC
[Spark Gap Tesla Coils (SGTC)]
MRMILSTAR
November 27, 2021, 05:04:46 PM
post Re: Someone built Big TC
[Spark Gap Tesla Coils (SGTC)]
304er
November 27, 2021, 04:45:14 PM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General Chat]
Mads Barnkob
November 27, 2021, 07:59:37 AM
post Re: Contest of motor slowness, or lowness of voltage?
[Science, Research and News In Other Fields Than Electronics]
Mads Barnkob
November 27, 2021, 07:52:59 AM
post Re: FPS1000HD unboxing and first impression
[High Speed Filming]
Mads Barnkob
November 27, 2021, 07:51:01 AM
post Re: Help on my "first" SSTC (LabCoatz SSTC 2.0)- Popping transistors
[Solid State Tesla Coils (SSTC)]
Mads Barnkob
November 27, 2021, 07:47:23 AM
post Re: Infrastructure and Architecture
[General Chat]
Mads Barnkob
November 27, 2021, 07:37:57 AM
post Re: Someone built Big TC
[Spark Gap Tesla Coils (SGTC)]
Mads Barnkob
November 27, 2021, 07:21:43 AM
post Re: BrOdin coil - High power Big sparks!
[Dual Resonant Solid State Tesla coils (DRSSTC)]
patracy
November 27, 2021, 06:06:59 AM
post Re: Someone built Big TC
[Spark Gap Tesla Coils (SGTC)]
304er
November 27, 2021, 05:27:27 AM
post Re: Kind of old single piece of glass plate capacitor
[Spark Gap Tesla Coils (SGTC)]
304er
November 27, 2021, 03:19:55 AM
post Re: Kind of old single piece of glass plate capacitor
[Spark Gap Tesla Coils (SGTC)]
304er
November 27, 2021, 02:31:46 AM
post Re: Kind of old single piece of glass plate capacitor
[Spark Gap Tesla Coils (SGTC)]
304er
November 27, 2021, 02:22:20 AM
post Driverless SSTC
[Solid State Tesla Coils (SSTC)]
bismallah
November 26, 2021, 11:59:27 PM
post Infrastructure and Architecture
[General Chat]
Da_Stier
November 26, 2021, 11:27:47 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
klugesmith
November 26, 2021, 11:17:55 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
MRMILSTAR
November 26, 2021, 08:24:39 PM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
November 26, 2021, 07:48:07 PM
post Help on my "first" SSTC (LabCoatz SSTC 2.0)- Popping transistors
[Solid State Tesla Coils (SSTC)]
TiagoBS
November 26, 2021, 07:44:25 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
davekni
November 26, 2021, 06:58:16 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
plasma
November 26, 2021, 05:40:37 PM
post Re: "Exploding" water
[Capacitor Banks]
MRMILSTAR
November 26, 2021, 04:39:07 PM
post Re: Kind of old single piece of glass plate capacitor
[Spark Gap Tesla Coils (SGTC)]
MRMILSTAR
November 26, 2021, 04:31:20 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Intra
November 26, 2021, 03:52:08 PM
post Re: "Exploding" water
[Capacitor Banks]
johnf
November 26, 2021, 09:17:07 AM
post Re: Kind of old single piece of glass plate capacitor
[Spark Gap Tesla Coils (SGTC)]
paulj
November 26, 2021, 08:45:34 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ItsChloeUwU
November 26, 2021, 06:31:56 AM
post Kind of old single piece of glass plate capacitor
[Spark Gap Tesla Coils (SGTC)]
304er
November 26, 2021, 05:59:58 AM
post Re: FPS1000HD unboxing and first impression
[High Speed Filming]
Kamui
November 26, 2021, 05:12:14 AM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
304er
November 26, 2021, 05:07:10 AM
post Please Help Me Find Datasheed Of IGBT
[Dual Resonant Solid State Tesla coils (DRSSTC)]
ItsChloeUwU
November 26, 2021, 04:08:22 AM
post Re: Contest of motor slowness, or lowness of voltage?
[Science, Research and News In Other Fields Than Electronics]
davekni
November 25, 2021, 07:35:57 PM
post Re: 13.56MHz ISM frequency HFSSTC (or perhaps HFDRSSTC)
[Solid State Tesla Coils (SSTC)]
davekni
November 25, 2021, 07:28:52 PM
post Re: Practical limitations in size for Marx or CW
[Voltage Multipliers]
davekni
November 25, 2021, 07:22:32 PM
post Re: "Exploding" water
[Capacitor Banks]
304er
November 25, 2021, 07:01:03 PM
post Practical limitations in size for Marx or CW
[Voltage Multipliers]
abstruse1
November 25, 2021, 05:58:34 PM
post Gas Discharge Tubes for Marx Generator?
[Voltage Multipliers]
abstruse1
November 25, 2021, 05:55:53 PM
post Contest of motor slowness, or lowness of voltage?
[Science, Research and News In Other Fields Than Electronics]
klugesmith
November 25, 2021, 05:50:35 PM
post "Exploding" water
[Capacitor Banks]
MRMILSTAR
November 25, 2021, 04:48:10 PM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
paulj
November 25, 2021, 10:52:44 AM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Netzpfuscher
November 25, 2021, 07:09:29 AM
post Re: 13.56MHz ISM frequency HFSSTC (or perhaps HFDRSSTC)
[Solid State Tesla Coils (SSTC)]
Steve Ward
November 25, 2021, 06:58:56 AM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
davekni
November 25, 2021, 05:51:03 AM
post Re: Some Starting Questions About My New DRSSTC ???
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
November 25, 2021, 05:13:17 AM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
MRMILSTAR
November 25, 2021, 04:50:11 AM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
davekni
November 25, 2021, 02:25:10 AM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
304er
November 25, 2021, 12:42:59 AM
post Re: 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
klugesmith
November 24, 2021, 11:03:52 PM
post Re: My completed 14-stage Cockroft-Walton voltage multiplier
[Voltage Multipliers]
MRMILSTAR
November 24, 2021, 10:00:47 PM
post Re: Next Gen DRSSTC
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Intra
November 24, 2021, 02:32:28 PM
post 60hz synchronous motor on 50hz?
[Spark Gap Tesla Coils (SGTC)]
paulj
November 24, 2021, 08:28:50 AM

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