Author Topic: DRSSTC low power testing killing IGBTs  (Read 10305 times)

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
DRSSTC low power testing killing IGBTs
« on: November 15, 2023, 12:45:37 AM »
So I recently built the microbrute drsstc from EVR and was having pretty good results. I was up to 23in arcs at 350A and 80VAC in on my variac.

I decided to make the breakout point a little longer and pointed upwards a bit more so as to avoid ground strikes. That’s when everything went wrong. Blew an IGBT (60n60 minibricks).

I replaced the bad IGBT, went back to the original breakout point, and then tried again. Blew both IGBTs this time. Thinking that the 60n60s were simply fakes or whatever (got them on eBay for cheap) I then swapped the IGBTs to the “Thunderbolt” IGBTs that EVR sells. I decided to remove the secondary and place a tin can in its place.

Tested at very low power (30vac in from variac), heard the buzzing from the Interruptor, and then blew one of these new and $$$ IGBTs.

Completely at a loss for what on earth could be happening here? I don’t even know where to begin with testing seeing as I only changed the breakout point before all of this started.

Before going into great detail of the system, what could cause these IGBTs to just suddenly begin failing (shorted) at low voltage?

-Matt

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #1 on: November 15, 2023, 03:01:35 AM »
Here are the gate drive signals with a function generator connected. (no IGBTs connected).  They look to be operating as expected, but I do see significant ringing. Gate drive resistors are 5ohm, maybe I should go to 10ohm? Might this be enough to be eating up IGBTs?

-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #2 on: November 15, 2023, 04:51:16 AM »
Quote
Here are the gate drive signals with a function generator connected. (no IGBTs connected).  They look to be operating as expected, but I do see significant ringing. Gate drive resistors are 5ohm, maybe I should go to 10ohm? Might this be enough to be eating up IGBTs?
Vge ringing measurements without IGBT load are not useful.  Damping comes from series resistor into IGBT Cge.

For testing, run very low duty cycle, perhaps one enable pulse per second and <100us pulse width.  Even at low voltage, current can ramp up to a high level.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #3 on: November 15, 2023, 01:57:50 PM »
I’ve got one IGBT left, so I can connect it and measure again later today. Assuming that fixes the ringing, what would be the next thing to suspect? Bad TVS that’s failing to catch spikes on the bus? None of them are shorted, but that doesn’t mean that they aren’t open.

When I was testing previously, current got up to around 100A. I could see bus voltage switching close to zero crossing with current.

-Matt

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #4 on: November 16, 2023, 03:36:16 AM »
Looks like the gate signal with the one good IGBT connected is clean. Not sure where to go with this?

-Matt


Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #5 on: November 16, 2023, 04:08:09 AM »
Quote
Assuming that fixes the ringing, what would be the next thing to suspect? Bad TVS that’s failing to catch spikes on the bus? None of them are shorted, but that doesn’t mean that they aren’t open.
Yes, testing bidirectional TVS diodes requires enough voltage to check breakdown.

General thought is to run at very low duty cycle and post scope waveforms.  Perhaps I or someone else can detect an anomaly.

One possible issue, though perhaps not too likely, is occasional slow or failing oscillation startup at beginning of enable pulse.  Startup was a problem with my friends EVR Microbrute, though it did not result in damaged IGBTs.  As Mads has suggested for other builds, a resistor across bridge output helps with startup.  That fixed my friends Microbrute.  A resistor between Out+ and Out- nodes.  Makes output center of voltage-doubler supply.  Without resistor, slight leakage current differences between IGBTs can cause initial half-bridge output to be at one supply rail or the other.

Quote
Looks like the gate signal with the one good IGBT connected is clean. Not sure where to go with this?
Duty cycle is above 50% and positive voltage is above negative voltage.  Perhaps a capture zoomed out to include the initial edge at beginning of enable pulse would help show what is happening.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #6 on: November 16, 2023, 04:51:00 AM »
I see that now. Not sure what’s causing that. I’ll add the resistor and try again when some more IGBTs arrive.

Here is the zoomed out waveform. Still shows the irregularities.

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #7 on: November 16, 2023, 05:29:40 AM »
Quote
I see that now. Not sure what’s causing that. I’ll add the resistor and try again when some more IGBTs arrive.
That bridge output resistor will not fix Vge issue.  Probably worth figuring out Vge problem before running with any bridge voltage.  Perhaps scope GDT secondary along with Vge and capture an even longer scope trace.  GDT secondary can't support an average DC voltage.  Longer capture of those two signals might hint at what is wrong.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #8 on: November 17, 2023, 12:32:28 AM »
Here is the measurement from the GDT secondary (in purple). Yellow is VGE on the good IGBT.

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #9 on: November 17, 2023, 01:08:41 AM »
I decided to use my diff probe to see how VGE looks. Looks perfect to me?




Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #10 on: November 17, 2023, 01:09:58 AM »
Quote
Here is the measurement from the GDT secondary (in purple). Yellow is VGE on the good IGBT.
I'm guessing one of the driver chips is bad.  I'd scope those two outputs next.  And perhaps captures of both sides of coupling capacitors C10 and C11 between driver U4 and GDT primary to make certain those are not shorted.

BTW, the instructions and schematics my friend received from EVR are somewhat inconsistent.  Is your GDT 1:1:1 as per instructions?  (Schematic shows 1:2:2.)  There could be other differences between my friend's and yours.  My recommendations here are presuming your coil is similar to the one my friend purchased a couple years ago.  Schematic is copyrighted, so I can't post a copy.

Presuming C10 and C11 are good and both 2uF, GDT inductance must be quite high.  How many turns on your GDT?

Quote
I replaced the bad IGBT, went back to the original breakout point, and then tried again. Blew both IGBTs this time.
When one IGBT fails shorted, that stresses the other IGBT.  (Usually destroys other IGBT, but occasionally not.)  It is quite possible that other IGBT was partially damaged.  It then failed completely, which then destroyed the replacement IGBT.
Another possibility is that some of the 1.5KE220CA TVS diodes failed first.  Depending on avalanche energy capability of the IGBTs being used, TVS diodes can cause more issues than they fix.  If TVS diodes fail shorted, that causes opposite IGBT to die driving the short.  That may then cause enough current to burn open the failed TVS diode bond wires.

Quote
I decided to use my diff probe to see how VGE looks. Looks perfect to me?
Oh.  Was there any power to half-bridge during this scoping?  Presuming not, was scope probe ground clip connected to emitter?  Diff probe is a better way to measure, but I'm puzzled as to how the previous traces ended up the way they are.

Vge is low.  Perhaps there is still a driver chip issue?
« Last Edit: November 17, 2023, 01:15:17 AM by davekni »
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #11 on: November 17, 2023, 01:22:03 AM »
My GDT is 1:1:1 and I see is supposed to have 10 turns. I actually have 15 since that’s how many fit. Could that be a problem? The capacitor is two 1uF in parallel, so 2uF total.

I swapped both driver ICs and same result with diff probe.

Previous measurements, I actually had ground clip tied to board ground. I can clip to emitter to see if that makes a difference. No power to bridge during any measurements.

-Matt

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #12 on: November 17, 2023, 01:29:51 AM »
Here is the VGE with ground clip on emitter.

Forgot to mention that I’m not using the onboard MMC. I upgraded to a larger one…Having said that, my first blown IGBT was with the on board MMC, so I’m leaning towards something else. I’m also using a different primary and secondary (much larger and robust than what was called for).

Any chance the upgraded MMC is the problem? It’s 0.165uF rated for 6400v.


Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #13 on: November 17, 2023, 03:57:38 AM »
Quote
Here is the VGE with ground clip on emitter.
Looks much better, though still a bit low.  Microbrute does produce somewhat low Vge, ~14V.  Perhaps scope probe compensation or other calibration would verify if voltage is really ~12V as shown.

Quote
Forgot to mention that I’m not using the onboard MMC. I upgraded to a larger one…Having said that, my first blown IGBT was with the on board MMC, so I’m leaning towards something else. I’m also using a different primary and secondary (much larger and robust than what was called for).

Any chance the upgraded MMC is the problem? It’s 0.165uF rated for 6400v.
Can't think of any reason that would cause issues.

Quote
My GDT is 1:1:1 and I see is supposed to have 10 turns. I actually have 15 since that’s how many fit. Could that be a problem?
Extra turns does have a down-side of increased leakage inductance.  Slows down Vge edges a bit.  Probably not a big concern here.  If you want to improve GDT, winding with two twisted pairs will further reduce leakage inductance:
https://highvoltageforum.net/index.php?topic=1854.msg13949#msg13949

Quote
I was up to 23in arcs at 350A and 80VAC in on my variac.
Might be just a bit too high current for this design.  Especially if your Vge is even lower than typical.  Microbrute controller has no phase lead.  IGBTs switch slightly after zero current, so hard-switch off-to-on.  That adds some stress and power dissipation.  Low Vge could be causing IGBTs to come out of saturation at peak current.  (Did you measure current or base it on measured R23 POT voltage?)
There are other driver chips that can handle over 15V.  Also some with FET-only output stages that would provide closer to full supply voltage to Vge.  You could increase Vge slightly by tweaking 15V towards 16V absolute maximum rating of UCC3732X chips.  One easy patch is to add a diode in series with LM7815 ground lead.  Adds ~0.6Vf of diode.
This post lists a few 20+V driver chips.  Mostly duals for UD2.7 etc. drivers:
    https://highvoltageforum.net/index.php?topic=2400.msg17650#msg17650
There may be pin-compatible drivers for UCC3732X with either higher voltage capability and/or FET-only output stages.  I'd recommend a DigiKey search.  Even if purchasing elsewhere, I find DigiKey's search easiest to use.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #14 on: November 17, 2023, 04:45:57 AM »
Current was measured with a Pearson 110. I do have the OCD set and it did trigger a few times during tuning, but failures all occurred well below the limit. After I had things tuned and hitting around 23-24”, no more OCD triggering.

I guess the next thing is going to be getting new IGBTs installed and managing some scope shots with power applied to the bridge?

If all else fails, I do have a UD2.7 driving a cm300dy-12nF for my fusor power supply. I’ll probably duplicate that setup and use for this coil.

EVRs stuff seems very finicky and expensive to own.

-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #15 on: November 17, 2023, 05:13:52 AM »
Quote
but failures all occurred well below the limit.
That does seem odd.  Makes my previous guesses unlikely.
Possible that some previous event fried TVS diode(s) open.  Looks like IXGN60N60C2D1 isn't avalanche rated.  That could make TVS diodes important to clamp any switching transients.  (I forget how good or not the bridge ECB layout is.)
Or perhaps it is still possible that occasional startup failure caused IGBT failure.  Vge tends to look like a very-low-frequency ring (2uF with GDT inductance) if oscillation doesn't start.  That doesn't often fry IGBTs, but it does put them in linear range for extended times.  Resistor across bridge output helps make startup reliable.

Quote
I guess the next thing is going to be getting new IGBTs installed and managing some scope shots with power applied to the bridge?
I can't think of any other good option (except your next though).  Add the output resistor first.

Quote
If all else fails, I do have a UD2.7 driving a cm300dy-12nF for my fusor power supply. I’ll probably duplicate that setup and use for this coil.
UD2.7 is a better driver.  Phase lead and larger Vge swing allow pushing current higher for a given IGBT.  Bridge output resistor is still needed (unless IGBT leakage currents match well).

Quote
EVRs stuff seems very finicky and expensive to own.
My only direct experience with EVR was purchasing a MIDI interrupter kit in 2019.  (Was in a hurry to get something for demonstrating my DRSSTC at a local science fair.  Got my DRSSTC running 10 days before the fair, so had little time left.)  Quite disappointed with its performance.  Designed and built my own for the next year.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #16 on: November 17, 2023, 02:22:29 PM »
I’ve noticed at times that startup requires a much longer on time than I usually run. I usually keep on time below 100uS, but have to start up near 250uS (from time to time) to get it to start. The first set of IGBTs were lost at low power and I could hear a thumping from my variac. OCD was set to 150A, so it wasn’t that.

Second set was lost very inexplicably since they had been going strong for many runs.

Last two sets lost at low power, too low to see any discharge. (Last set was with secondary removed).

Maybe startup is somehow killing them?

What size/type resistor should I use for start up?

-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #17 on: November 17, 2023, 07:28:06 PM »
Quote
What size/type resistor should I use for start up?
I think 5k to 10k is common.  Generally lower is better until power dissipation becomes a nuisance.  Lower resistance can overcome higher IGBT leakage current.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #18 on: November 17, 2023, 08:49:25 PM »
I think I read that this would go from the CE connection (half bridge) to either positive or negative rail depending on which IGBT comes on first, right? So if positive rail IGBT comes on first, then resistor goes to negative rail?

-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #19 on: November 17, 2023, 09:49:11 PM »
Quote
I think I read that this would go from the CE connection (half bridge) to either positive or negative rail depending on which IGBT comes on first, right? So if positive rail IGBT comes on first, then resistor goes to negative rail?
As long as you know for sure which IGBT comes on first, yes, what you say is accurate.  Provides double the startup pulse amplitude.  However, I don't think Microbrute is guaranteed to always start in the same state.  (Unlike UD2.7 which biases comparitor input slightly negative.)  Without knowing initial state, resistor between out+ and out- makes output start in center.  Then either IGBT turning on generates a half-voltage startup pulse.  Even most UD2.7 users don't bother measuring or figuring out which IGBT(s) turn on first, so just use a resistor across bridge output.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #20 on: November 17, 2023, 11:40:01 PM »
I see the on board MMC has bleeder resistors and my upgraded MMC doesn’t. Maybe startup IS the issue after all?

-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #21 on: November 18, 2023, 06:43:41 AM »
Quote
I see the on board MMC has bleeder resistors and my upgraded MMC doesn’t. Maybe startup IS the issue after all?
That is also the best guess I'd have.  Increased GDT turn count will make slower ring when startup fails.  That might contribute to failures.  Though I haven't figured out exactly how startup failure would result in fried IGBTs.
David Knierim

Online Mads Barnkob

  • Administrator
  • Executive Board Member
  • *****
  • Posts: 2359
  • Karma: +78/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: DRSSTC low power testing killing IGBTs
« Reply #22 on: November 18, 2023, 09:15:32 AM »
I see the on board MMC has bleeder resistors and my upgraded MMC doesn’t. Maybe startup IS the issue after all?

-Matt

This is what is tried to be fixed with a 10K resistor across the inverter output, to get rid of any residual charge that could confuse the logic at start of a pulse, not just start-up, but it would sound sputtering and unstable.

Did you take a look at all gate driver resistors, TVS and diodes from GDT to IGBTs? Many components can fail "half" at IGBT failures, like gate resistors going from 10R to 100R-10K and you get all these weird behaviours from differently damaged components in the drive chain.
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #23 on: November 18, 2023, 06:24:23 PM »
Here’s a picture of the setup. It’s a pretty standard design…

As for TVS testing, I’ve tested for short circuit and none are shorted. To test open, I’d need to remove and connect to my hv power supply, right? Any other way to guess?

Zener diodes all test 0.6v, so guess they’re good. Gate resistors are 5.5ohms, so those should be good.

I’m itching to tweak something (easy that is) while I wait for replacement IGBTs :-) Should I remove some turns from the GDT? Or, should the bridge resistor negate any shenanigans from it having higher inductance?

Another thought that I had was voltage sag caused by smallish bus capacitors (1000uF each) and a much higher MMC (.165uF vs .075uF) than designed?

-Matt


Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #24 on: November 18, 2023, 07:39:36 PM »
Quote
It’s a pretty standard design…
Probably unrelated to failures, but I see IGBT Kelvin emitter connection is defeated by shorting the two emitter terminals together on the ECB.  That increases internal IGBT die Vge spikes, especially given hard off-to-on switching of this driver that doesn't support phase lead.  Perhaps the relatively-low normal Vge of this design prevents internal Vge spikes from becoming problematic.

Quote
To test open, I’d need to remove and connect to my hv power supply, right? Any other way to guess?
Vce TVS diodes can be tested with IGBTs removed.  Apply some voltage to half-bridge supply, say 40V.  Then verify that output (Out+) gets clamped around +440V and -400V.

Quote
I’m itching to tweak something (easy that is) while I wait for replacement IGBTs :-) Should I remove some turns from the GDT?
Your decision.  I'd remove turns at least to make it like the original design with 10 turns.  If it were my coil, I'd change to two twisted pairs as well.  But there may be some comfort in exactly matching original design using three wires.

If changing GDT, you could check Vge diodes while GDT is removed.  However, no way to know if zener diodes or TVS diodes are conducting.  Both are nominally 33V.

Quote
Another thought that I had was voltage sag caused by smallish bus capacitors (1000uF each) and a much higher MMC (.165uF vs .075uF) than designed?
You may get voltage sag.  I can't think of any resulting issue beyond reduced performance.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #25 on: November 19, 2023, 12:01:31 AM »
I hooked up my Bertan to the board and applied +500v (gradually) to each set of TVS diodes. Set one clamped at 435v and set two clamped at 425v. Guess they can be ruled out?

-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #26 on: November 19, 2023, 04:47:45 AM »
Quote
I hooked up my Bertan to the board and applied +500v (gradually) to each set of TVS diodes. Set one clamped at 435v and set two clamped at 425v. Guess they can be ruled out?
Yes, presuming you biased the half-bridge supply.  If 0V across supply, then no way to determine which pair of TVS diodes is conducting, since they are bidirectional.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #27 on: November 19, 2023, 05:32:49 AM »
I put the hv power supply leads directly on each set of TVS diodes and then reversed the leads.

-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #28 on: November 19, 2023, 10:00:16 PM »
Quote
I put the hv power supply leads directly on each set of TVS diodes and then reversed the leads.
If the TVS current at breakdown was sufficient to charge bulk capacitors, then this would be valid.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #29 on: November 21, 2023, 02:02:52 PM »
While I keep waiting for my replacement IGBTs, I had another question about the voltage drop and calculation for half bridges…Should bus voltage (400v approximately from voltage doubler) be half (200v)? So the calculation for energy in capacitor vs MMC energy would be (BUS/2)^2 instead of BUS^2?

-Matt

Online Mads Barnkob

  • Administrator
  • Executive Board Member
  • *****
  • Posts: 2359
  • Karma: +78/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
Re: DRSSTC low power testing killing IGBTs
« Reply #30 on: November 21, 2023, 07:34:43 PM »
While I keep waiting for my replacement IGBTs, I had another question about the voltage drop and calculation for half bridges…Should bus voltage (400v approximately from voltage doubler) be half (200v)? So the calculation for energy in capacitor vs MMC energy would be (BUS/2)^2 instead of BUS^2?

-Matt

Yes, you only have the energy available from one of the capacitors, at a time. Which you can also see is substantial as its the figure squared that you are halving.
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #31 on: November 24, 2023, 11:26:04 PM »
Got my new IGBTs in (the C2D1) version which can handle another 100A.

Added a 7k resistor across the bridge output, and now I’m able to get things going with barely any voltage in from the variac. Here is a screen shot of current and voltage across the bridge. I am testing at very low BPS and on time with just the primary in place.

Edit: Excuse the scaling. I swapped my Pearson 110 CT for a triad CT (easier to fit in there) and forgot to change scaling around.


« Last Edit: November 25, 2023, 12:02:15 AM by FonziDaytona »

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #32 on: November 25, 2023, 04:29:13 AM »
Quote
Added a 7k resistor across the bridge output, and now I’m able to get things going with barely any voltage in from the variac.
Glad resistor worked as expected.  Hope IGBTs survive now!  Scope captures look fine.
David Knierim

Offline thedoc298

  • High Voltage Engineer
  • ****
  • Posts: 221
  • Karma: +2/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #33 on: November 26, 2023, 05:14:20 AM »
I see the on board MMC has bleeder resistors and my upgraded MMC doesn’t. Maybe startup IS the issue after all?

-Matt

This is what is tried to be fixed with a 10K resistor across the inverter output, to get rid of any residual charge that could confuse the logic at start of a pulse, not just start-up, but it would sound sputtering and unstable.

Did you take a look at all gate driver resistors, TVS and diodes from GDT to IGBTs? Many components can fail "half" at IGBT failures, like gate resistors going from 10R to 100R-10K and you get all these weird behaviours from differently damaged components in the drive chain.

Hello all, enjoying the chat, where (and I looked) would I find more info about the 10k resisitor and what problem it might fix. I know it's not necessary 10k. 

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #34 on: November 26, 2023, 08:04:56 PM »
I see the on board MMC has bleeder resistors and my upgraded MMC doesn’t. Maybe startup IS the issue after all?

-Matt

This is what is tried to be fixed with a 10K resistor across the inverter output, to get rid of any residual charge that could confuse the logic at start of a pulse, not just start-up, but it would sound sputtering and unstable.

Did you take a look at all gate driver resistors, TVS and diodes from GDT to IGBTs? Many components can fail "half" at IGBT failures, like gate resistors going from 10R to 100R-10K and you get all these weird behaviours from differently damaged components in the drive chain.

Hello all, enjoying the chat, where (and I looked) would I find more info about the 10k resisitor and what problem it might fix. I know it's not necessary 10k.

It fixed issues where my coil wouldn’t oscillate at low power or act “weird” and then kill a IGBT. Lost a lot of them…

After adding a 7kohm resistor across the half bridge outputs, the coil starts up at very low input voltage to the bridge. No weirdness either.

Now back to tuning and seeing if I can push my IGBTs higher :-)

Here it is hitting around 21”

 




Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #35 on: November 26, 2023, 09:57:50 PM »
Quote
Hello all, enjoying the chat, where (and I looked) would I find more info about the 10k resisitor and what problem it might fix. I know it's not necessary 10k.
Here are several posts where I describe startup behavior and need for a resistor across bridge output:
    https://highvoltageforum.net/index.php?topic=1934.msg14466#msg14466
    https://highvoltageforum.net/index.php?topic=2330.msg17044#msg17044
    https://highvoltageforum.net/index.php?topic=2388.msg18296#msg18296
    https://highvoltageforum.net/index.php?topic=2624.msg19237#msg19237
    https://highvoltageforum.net/index.php?topic=1914.msg17230#msg17230
    https://highvoltageforum.net/index.php?topic=1350.msg9973#msg9973
If you still have questions, feel free to ask.

Quote
Now back to tuning and seeing if I can push my IGBTs higher :-)

Here it is hitting around 21”
Great to see it working!  Good luck with pushing just the right amount, not quite to IGBT failure.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #36 on: November 27, 2023, 03:12:43 PM »
Thanks again, David. That’s two of EVRs products that needed outside help and modifications to make them work.

-Matt

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #37 on: November 30, 2023, 12:03:30 AM »
Managed to kill another IGBT. At least this time, I was pushing the coil near the max. No explosion, just sudden silence. I was near 400A, which seems low for these IGBTs. I’m thinking that this might be due to running 15v on the gates instead of something closer to 20v since the max pulsed rating is 300A.

Any reason not to modify the GDT to output 20v instead of 15v? I could wrap a 10 turn primary directly over the existing windings without much hassle…

-Matt


Offline thedoc298

  • High Voltage Engineer
  • ****
  • Posts: 221
  • Karma: +2/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #38 on: November 30, 2023, 12:38:37 AM »
Quote
Hello all, enjoying the chat, where (and I looked) would I find more info about the 10k resisitor and what problem it might fix. I know it's not necessary 10k.
Here are several posts where I describe startup behavior and need for a resistor across bridge output:
    https://highvoltageforum.net/index.php?topic=1934.msg14466#msg14466
    https://highvoltageforum.net/index.php?topic=2330.msg17044#msg17044
    https://highvoltageforum.net/index.php?topic=2388.msg18296#msg18296
    https://highvoltageforum.net/index.php?topic=2624.msg19237#msg19237
    https://highvoltageforum.net/index.php?topic=1914.msg17230#msg17230
    https://highvoltageforum.net/index.php?topic=1350.msg9973#msg9973
If you still have questions, feel free to ask.

Quote
Now back to tuning and seeing if I can push my IGBTs higher :-)

Here it is hitting around 21”
Great to see it working!  Good luck with pushing just the right amount, not quite to IGBT failure.



Thanks a million, be hard going without you guys on this site.

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #39 on: November 30, 2023, 05:50:53 AM »
Quote
Managed to kill another IGBT. At least this time, I was pushing the coil near the max. No explosion, just sudden silence. I was near 400A, which seems low for these IGBTs. I’m thinking that this might be due to running 15v on the gates instead of something closer to 20v since the max pulsed rating is 300A.
15Vge may contribute, along with hard off-to-on switching (no phase lead provided in this design).

Quote
Any reason not to modify the GDT to output 20v instead of 15v? I could wrap a 10 turn primary directly over the existing windings without much hassle…
Two down-sides to changing GDT winding.  One is that leakage inductance increases so Vge waveforms will have slower transitions and perhaps more overshoot/undershoot.  Another is that higher Vge can make hard off-to-on switching more stressful for IGBTs.  Increases switching current, forcing opposite diode reverse-recovery to occur faster.  When it does turn off, rapid end to higher current leads to higher Vce spikes.  Slower Vge edges can reduce this issue some, but also delay switching so normal primary current is a little higher by the time the off-to-on transition occurs.  Hard to say what the final trade-off will be.  The original design of EVR Microbrute appears to have 1:2:2 GDT winding, which was changed to 1:1:1 for later versions.
If you do change to a higher GDT ratio, I'd recommend dropping primary turns to 10 (per Microbrute instructions) to avoid adding even more leakage inductance.
(Higher GDT ratios also increase power dissipation in driver chips.  There may be margin to handle that.)
Another option is to trade out driver chips for versions capable of 20V and with direct FET outputs so GDT drive gets closer to full supply voltage (rather than Vbe less).  That would maintain fast Vge edges by keeping GDT 1:1.

David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #40 on: November 30, 2023, 06:10:34 PM »
I’ll wind another GDT (less than 1:2:2 since I’m only needing another 5v or so) tonight to see what the waveforms look like. This would be much easier than reworking the EVR design.

My planned upgrade (ud2.7 with a cm300 half bridge) won’t have this problem. Just need a few more parts to make the swap.

-Matt

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #41 on: December 01, 2023, 05:07:24 AM »
Rewound the GDT for 10 turns on the secondary and 7 turns on the primary.

Look okay? It boosted the voltage up a few volts but also has a little more overshoot.


Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #42 on: December 01, 2023, 06:14:43 AM »
Quote
Rewound the GDT for 10 turns on the secondary and 7 turns on the primary.

Look okay? It boosted the voltage up a few volts but also has a little more overshoot.
Overshoot looks OK.  Not nearly enough to damage IGBT gates, and dip after overshoot isn't significant.

One puzzle:  Normally falling edge has quite a bit more overshoot (undershoot) than rising edge due to diode across gate series resistor.  This looks so symmetric that I'd guess either diodes are missing or shorted or gate resistors are so low value that diodes don't conduct significantly.
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #43 on: December 01, 2023, 02:03:32 PM »
Looking at previous scope shots, the undershoot looks to be close to or the same as these do, and slightly more than the overshoot. Maybe just a coincidence?

The gate resistance is 5.4ohm (measured). If diodes were shorted, I should see a short across the gate resistance, no? I guess they could be open…

Guess it’s clear to let it rip? I’ll see if I can push higher peak amps this time.

-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #44 on: December 01, 2023, 08:48:38 PM »
Quote
Looking at previous scope shots, the undershoot looks to be close to or the same as these do, and slightly more than the overshoot. Maybe just a coincidence?
Captures in your posts #4, #9, and #12 look more like I'd expect, with undershoot on falling edge and slower rising edge with little overshoot.  Difference may be due to just increased GDT leakage inductance combined with gate resistor not increasing to match.

Quote
Guess it’s clear to let it rip? I’ll see if I can push higher peak amps this time.
Probably.  There is always some risk.  Hard to tell how dead-time compares with original waveforms.  Didn't spot any previous captures so nicely comparing Vge of the two IGBTs as this last capture does.  Probably OK.

Hope power testing goes well!
David Knierim

Offline FonziDaytona

  • High Voltage Technician
  • ***
  • Posts: 108
  • Karma: +0/-0
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #45 on: December 03, 2023, 06:50:36 AM »
So far so good. Up past 24” arcs now.

I’ve noticed that the OCD trips (set for 400A) yet the scope doesn’t show current going above 320A. What gives? Would this be transients that are too fast for the scope, but somehow long enough for the OCD circuit? Or, is the EVR OCD design up to some shenanigans? Should I set the OCD per the scope shots or stick to EVR settings chart?



-Matt

Online davekni

  • Executive Board Member
  • *******
  • Posts: 3000
  • Karma: +142/-2
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: DRSSTC low power testing killing IGBTs
« Reply #46 on: December 04, 2023, 05:06:19 AM »
Quote
I’ve noticed that the OCD trips (set for 400A) yet the scope doesn’t show current going above 320A. What gives? Would this be transients that are too fast for the scope, but somehow long enough for the OCD circuit? Or, is the EVR OCD design up to some shenanigans? Should I set the OCD per the scope shots or stick to EVR settings chart?
As long as you are confident in your scope path, then set it based on that.  EVR OCD circuit places diode bridge after CT burden resistor.  )After most of burden resistance, 1.6ohms.)  100 ohm resistor after bridge along with 1uF filter cap make for relatively high peak diode current for 1N4148 signal diodes.  Diode forward drop is part of OCD setpoint calculation, but is not very precise at high current peaks.  1uF filter cap minimizes sensitivity to transients.
David Knierim

High Voltage Forum

Re: DRSSTC low power testing killing IGBTs
« Reply #46 on: December 04, 2023, 05:06:19 AM »

 


* Recent Topics and Posts

post Re: Measuring the coherence length of a laser
[Light, Lasers and Optics]
haversin
Today at 05:48:16 PM
post Re: Big Coil Build Log
[Dual Resonant Solid State Tesla coils (DRSSTC)]
flyingperson23
Today at 03:20:08 PM
post Re: Measuring the coherence length of a laser
[Light, Lasers and Optics]
Uspring
Today at 01:50:30 PM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General Chat]
shBLOCK
Today at 07:39:21 AM
post Re: First DRSSTC, Full Bridge PCB & IGBT Selection question.
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Benjamin Lockhart
Today at 04:14:03 AM
post Re: First DRSSTC, Full Bridge PCB & IGBT Selection question.
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
Today at 01:12:38 AM
post Re: Testing and teardown of a commercial induction heater
[Electronic Circuits]
dejuli2
December 11, 2024, 09:58:07 PM
post Re: Push Pull VTTC
[Vacuum Tube Tesla Coils (VTTC)]
myoniwy
December 11, 2024, 09:26:40 PM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General Chat]
myoniwy
December 11, 2024, 09:08:24 PM
post Re: Measuring the coherence length of a laser
[Light, Lasers and Optics]
haversin
December 11, 2024, 08:05:20 PM
post Re: First DRSSTC, Full Bridge PCB & IGBT Selection question.
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Beggernator.
December 11, 2024, 03:48:40 PM
post Re: APC Back-UPS Pro 1500 Watt Teardown
[Electronic Circuits]
Mr.Cas
December 11, 2024, 05:27:48 AM
post Re: First DRSSTC, Full Bridge PCB & IGBT Selection question.
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 11, 2024, 01:41:52 AM
post Re: First DRSSTC, Full Bridge PCB & IGBT Selection question.
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Beggernator.
December 10, 2024, 04:18:23 PM
post Re: Ferrite transformer from Microwave
[Transformer (Ferrite Core)]
davekni
December 10, 2024, 04:05:07 AM
post Attracting (and picking up) nonferrous metals with an electromagnet
[Electronic Circuits]
davekni
December 10, 2024, 02:49:03 AM
post Ferrite transformer from Microwave
[Transformer (Ferrite Core)]
Simranjit
December 09, 2024, 11:37:35 PM
post Re: Measuring the coherence length of a laser
[Light, Lasers and Optics]
alan sailer
December 07, 2024, 09:41:37 PM
post Re: Measuring the coherence length of a laser
[Light, Lasers and Optics]
haversin
December 07, 2024, 07:07:49 PM
post Re: Measuring the coherence length of a laser
[Light, Lasers and Optics]
alan sailer
December 07, 2024, 06:33:32 PM
post Re: Measuring the coherence length of a laser
[Light, Lasers and Optics]
haversin
December 07, 2024, 06:13:56 PM
post Re: Measuring the coherence length of a laser
[Light, Lasers and Optics]
alan sailer
December 07, 2024, 02:40:55 AM
post Measuring the coherence length of a laser
[Light, Lasers and Optics]
haversin
December 07, 2024, 01:07:46 AM
post Re: Single board for SSTC and DRSSTC operation
[Solid State Tesla Coils (SSTC)]
davekni
December 07, 2024, 12:18:49 AM
post Re: Single board for SSTC and DRSSTC operation
[Solid State Tesla Coils (SSTC)]
Simranjit
December 06, 2024, 11:59:05 PM
post Re: First DRSSTC, Full Bridge PCB & IGBT Selection question.
[Dual Resonant Solid State Tesla coils (DRSSTC)]
davekni
December 06, 2024, 11:33:05 PM
post Re: Single board for SSTC and DRSSTC operation
[Solid State Tesla Coils (SSTC)]
davekni
December 06, 2024, 11:28:25 PM
post Re: Welcome new members, come say hello and tell a little about yourself :)
[General Chat]
GaelJaton
December 06, 2024, 09:47:04 PM
post Re: First DRSSTC, Full Bridge PCB & IGBT Selection question.
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Benjamin Lockhart
December 06, 2024, 08:06:27 PM
post Re: 160mm DRSSTC II project | Questions
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Late
December 06, 2024, 07:54:01 PM
post Re: Single board for SSTC and DRSSTC operation
[Solid State Tesla Coils (SSTC)]
Simranjit
December 06, 2024, 03:54:00 PM
post Re: First DRSSTC, Full Bridge PCB & IGBT Selection question.
[Dual Resonant Solid State Tesla coils (DRSSTC)]
Beggernator.
December 06, 2024, 02:38:56 PM
post Re: GU-81M Hartley Driven VTTC
[Vacuum Tube Tesla Coils (VTTC)]
janno288
December 06, 2024, 12:20:40 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