UD2.7C Debugging

[Saattvik24]

I finally populated my UD2.7C driver board and turned it on for the first time. I noticed it worked for a minute then the 12V and 5V voltage rails fell down to around 1V. So the next day, I started debugging, i removed the UCC27423 driver, and started probing around. The voltage rails were back to normal values.

I've attached the oscilloscope shots of the interrupter output received after schmitt trigger, the comparator output of the feedback, and inputs to the UCC27423 on pin 2 and 4.

I've set the function generator as feedback at 150kHz and the interrupter is to 20Hz. The interrupter is currently on a breadboard, will be put on a pcb. Its a oneTesla interrupter. Also there is nothing connected to the OCD CT terminal block (It should work without that during testing right?)

I wanted to know if my waveforms are good and what could be the possible issues. I haven't adjusted the values of the potentiometers RV1 & RV2 yet, but i don't think thats the issue, as those are for UVLO & OCD.

UPDATE

When reinserted the UCC27423, the voltage rails again fell down. I tested the IC separately and it seems to be working, so I have no clue what the issue could be.

[Saattvik24]

Turns out one of the irf540n mosfet had died. But after replacing it, the driver is still not behaving properly. It still tends to turns off after a few seconds of operation, also the LM7812 voltage regulator gets extremely hot. Even the time in which it is on, the gate drive output waveform doesn't look good and it keeps moving.

[flyingperson23]

Check the mosfet driver or other ICs for excessive heat

[Saattvik24]

No other IC seems to be getting hot

[flyingperson23]

The 7812 getting hot and voltage rails falling sounds like something is trying to draw too much current. If it's not the mosfet driver, are you sure all 4 mosfets are good? If you got them from a vendor like amazon or ebay they may not be authentic parts.

[Saattvik24]

The mosfet Q4 irfp540 keeps dying (2 have died so far). Also i got most of my components from element14, but the mosfets are from a different electronics seller, but its also a reputable brand in my country.

[davekni]

The mosfet Q4 irfp540 keeps dying (2 have died so far).

One likely cause for IRFP540 dying is that the paired IRF9540 is failed shorted.  Or that 1k gate pull-up resistor(s) are open (ie. bad solder joint) or 0.1uF coupling capacitor is shorted.

Even the time in which it is on, the gate drive output waveform doesn't look good and it keeps moving.

Scope traces of the two GDT primary leads (UD2.7 output terminals) would be more useful for debug.

It still tends to turns off after a few seconds of operation, also the LM7812 voltage regulator gets extremely hot.

Output FETs don't directly connect to +12V, only UCC27423.  Since it is still good, implies high load on UCC27423 output(s).  Most likely cause is a bad FET with shorted gate.  Another reason to suspect an IRF9540 having died (or both).

If still problematic, I'd remove all four FETs and test with UCC27423 chip installed.  Scope FET gate waveforms and verify supplies are not loaded excessively.

[Saattvik24]

After many hours of debugging and replacing components, i bought new mosfets and replaced the old ones, turns out they were bad quality all along. It looks to be working now. I've attached the scoped output of a test gate drive transformer attached to a 10k resistor. I wanted to know if the waveforms are good and correct

[davekni]

Looks good.
BTW, real GDT is best made with twisted pairs such as from CAT5 cable.  Manually twisting wire pairs works too.  Parallel one wire of each pair for primary, as in this tutorial:
    https://highvoltageforum.net/index.php?topic=1854.msg19355#msg19355