I will build this. Hopefully I won't kill too many IGBTs. I will report the results in like a month

Great.  I think this is a good driver design.  Looking forward to seeing how it works for another user.  Of course, test initially at low duty cycle and low voltage.  And adjust R25 to get self oscillation frequency close to coil operating frequency even before powering bridge.  If R25 range doesn't include your coil frequency, then change C33 value, higher for lower frequency and lower for higher frequency.  (For a high-coupling QCW coil, self-oscillation frequency is useful to start at upper pole frequency.)

Quote
    Are the two unused 74AHC14 inputs tied to something?
Drat, oh well, nothing a bit of wire can't fix. An unfortunate case of out of sigh out of mind occurring there.

If using exact ECB layout, do remember to tie unused 74AHC14 inputs to something with patch wires.

If anything looks strange or incorrect, post scope captures and pictures.  I'd be happy to help before you apply enough power to fry IGBTs.

@davekni - The inductor you see would in the OP is ~100mH. I haven't properly measured it, but I don't think its exact value is super important. I have no formal electronics training (I'm a biologist), so I am basing my design heavily around Loneoceans QCW1.5.

That makes sense and is probably a reasonable size for your needs.  My question was about flyingperson23's need for several hundered uH and several hundred amps (which this core would not support).  I think that question is reasonably settled now.

Have fun building your QCW coil!

Gonna wire up a 3 stage Marx with them and the inductors I wound back then. I never tried using the Y5Vs back then with the inductors. Wish me luck

Yes, definitely wishing you luck!  Do you know how long (in time) output pulses need to be for your application?  You could run a simulation using your home-wound inductors to see if inductance is high enough.

.S. Output pulse timing is not critical, just figured it would be easier with a high voltage switch then trying to get the first spark gap's distance perfect.

I didn't find gap adjusting too difficult with my rounded-end copper pipe end caps.  They are friction-fit over PVC pipe using thin SS shim stock.  Made a very simple adjustment tool from a scrap of wood cut to proper gap distance.  Use that to set all gaps to same distance within reasonable tolerance.  If voltage needs to be a bit higher, I add a layer of plastic scrap to the tool to widen gap.  I don't find much difference in performance based on which gap fires first.

Did you ever hear anything else about that guy who used a UV bulb to trigger the spark gaps? LEDs don't seem to be there quite yet.

Haven't seen any specific examples.  I'd thought about suggesting UV triggering from a separate trigger spark gap close to bottom Marx gap, but don't have any information to say if that would work well or not.  Would be interesting to try.
Using a UV lamp is an interesting idea.  I have a short-arc xenon flash tube with quartz window.  Sounds like a good experiment to try if I ever get through my overly-long list of such experiments.  I don't know what wavelength is necessary.  If it must be short enough to ionize air, then it also won't travel that far in air.
Be careful with short wavelength UV.  It is quite hazardous, with enough photon energy to disrupt DNA in skin and eye cells.  So far I've used my short-arc UV flash lamp only with a polycarbonate window over it, as I needed only visible light.
Recall seeing 240nm UV LEDs somewhere, but efficiency was low and cost high.  May not be short enough wavelength still.

NCP81074A spec lists 10A peak reverse current for pulses.  So diodes are likely unnecessary.  If schottky diodes, could help since likely faster reverse recovery time than internal FET body diodes.

NCP81074A spec says nothing about requiring gate resistors.  I'd guess none is OK.

NCP81074A is very fast.  Look at all the spec layout recommendations carefully.

For your cap question, simulate it.

Shouldnt the gate startup pulses at igbt gates have the same amplitude of gate drive voltage?
In my case those pulses are very weak.
Is this normal?

Yes, start amplitude should be same as run amplitude, one Vge at +18V and other at -18V.  Low amplitude is not normal.
My guess is that your driver chip is the wrong part, not inverting.  Should be UCC27423, not UCC27424.
Either that or 74HC08 is inverting (74HC00), but I think that would cause other issues too.
Measure voltage of driver chip outputs when idle (not enabled).  They should be +9V.

Ah I see, I was misunderstanding how boost converters worked. I may still be misunderstanding, but It'll be boosting max to 1200v 25A which should be 80% duty cycle so 185A max with 100uH.

Using 1200V rated IGBT for 1200V output will be problematic.  There is no margin for parasitic inductance.  That is why gate charge etc. for 1200V IGBTs are specified for 600V.  That is typical use voltage, 600V nominal + 600V for spikes and tolerance.  Some list use to 800V.  Beyond that has a high probability of failure.  Given your 200A rather than nominal 600A, perhaps you could bet 1000V output.  Would require good low-inductance interconnect.

If 240V is rectified and unfiltered 240Vac, then calculate at 340Vpeak.  Ripple current will be a little higher than at 240V, but still fine presuming 20kHz.

Sorry to hijack your thread a little coilerer.

I'm the one who should apologize for hijacking this thread.  You're just answering my question.

I'm making a boost converter, hopefully at least 30kW. Switch is a cm300, for 20khz I'd like at least 400uH 200A. A few of these cores stacked would probably work, but toyota prius hybrids used to have buck/boost converters with beefy inductors, getting a few of those seems like my best bet.

To continue hijacking I'm still puzzled as to why 400uH is needed.  I'd think 50-100uH would be sufficient.  However, no requirement to answer.
For example (ignoring losses), if 30kW is boosting 240V at 125A to 600V at 50A, duty cycle is 60%, 30us at 20kHz.  100uH would ramp by 72A (= 240V * 30us / 100uH).  To average 125A, inductor current is 125A +-36A (89A to 161A).  50uH would ramp twice as much, 144A total or +-72A, so 53A to 197A.  That still fits within 200A maximum.

[edit] Actually we could get more DC energy in similar size inductor using a laminated steel core with air gap, able to handle more than 15000 gausses before saturation.  Not practical for a switching converter!

Tape-wound amorphous steel cores are used at 20kHz.  Haven't seen any such cores available surplus.

If I'm not misreading the datasheet, it looks like with enough turns for a few hundred uH, it saturates at under a hundred amps.

I'm curious about this QCW coil that requires such large buck inductance.  My QCW buck uses a 40uH 400A inductor, switching at 40kHz.

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.

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.

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.

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.

My DRSSTC switches poles when arcs get long.  It is a relatively low impedance design (~35k secondary impedance), opposite yours.  Relatively short on times and impulsive arcs.  I play music, but it isn't optimum for that.