Ud2.7 help

[flyingperson23]

I soldered together a through hole ud2.7. I have my bridge pretty much done and when I tried to test it nothing happened. The ud had power but won't seem to start oscillations. It gives a mildly-triangleish waveform on one pole of the gdt but nothing on the other. Of course I'm not getting anything on the feedback transformer, but it's not giving a starting pulse to get feedback either. I don't have a function generator to test the ud with a known feedback signal. Any advice?

[flyingperson23]

Here's the board. Feeding it 24vdc hence the lack of lm7824

[davekni]

It gives a mildly-triangleish waveform on one pole of the gdt but nothing on the other.

More information is needed to be helpful:  scope image of GDT inputs, picture or specs of GDT, what IGBTs are used and what R//D values between GDT outputs and IGBT gates.  I presume you are sending UD2.7 enable (interrupt) pulses to cause these waveforms.

Debugging is easier with a signal generator, or with modified UD2.7 circuitry for self-oscillation.  Still possible without.

[flyingperson23]

Here's the gdt. I'm not sure the exact ferrite, but it came out of a welder where it was used as a gdt for similar igbts. It's driving a cm300 full bridge through 5.1ohm resistors with a 1n4007 and 30v zener diodes connected as shown here https://www.stevehv.4hv.org/drsstcdesign/ISSTC_sch2.JPG
There is no input signal to the gdt.
How can I modify the ud to self oscillate?

[Mike]

There's a description of the simplest method in faces post here: https://highvoltageforum.net/index.php?topic=1373.msg10197#msg10197

[flyingperson23]

Ok, I got one of those cheap diy function generator kits and fed a ~70khz feedback signal in. I'm getting this across the gdt. Does it look good?

[davekni]

Here's the gdt.

GDT looks great.  Nicely done.

It's driving a cm300 full bridge through 5.1ohm resistors with a 1n4007 and 30v zener diodes connected as shown here

1N4007 is way too slow to be usable for a TC.

There is no input signal to the gdt.

Then I don't understand your previous statement: "It gives a mildly-triangleish waveform on one pole of the gdt".

Ok, I got one of those cheap diy function generator kits and fed a ~70khz feedback signal in. I'm getting this across the gdt. Does it look good?

Are these the two different GDT inputs, both referenced to UD2.7 ground?  I'll presume so.  Just not certain what "across GDT" means.  Would usually imply a differential measurement.
Presuming I'm understanding correctly, there is an issue.  GDT inputs should be at ground potential between enable pulses, not floating.  Do you perhaps have non-inverting gate driver chips installed?  Correct ones are inverting.  Driver chip outputs should be high between interrupt pulses, turning on the NFETs, pulling GDT low.

[flyingperson23]

Sorry, my wording isn't great. I'll switch the 1n4007s for faster diodes, for some reason it didn't cross my mind to check that. The triangle waveform I was referencing before was what I was getting before I realized my interrupter was off. And yes, you're spot on, I appear to have a non inverting mosfet driver. Thanks for your help!