Gate Driver Problem / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Gate Driver Problem

1 2 3 4 next last

Move Thread

LAN_403

Graham Armitage

Sun Jan 20 2013, 11:56PM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

Hi,

Have an interesting problem and desperate for some debugging suggestions. Building a DRSSTC based closely on the minibrute circuits and have fried 9 UCC37321 gate driver ICs. I am in test phase running a simple modulator with 200uS pulse width and 180Hz PRF. With the GDT connected to the driver IC's but NOT connected to the half-bridge circuit, everything works fine with good signal on the GDT secondaries.

When I connect the secondaries of the GDT to the half-bridge circuit and turn on the modulator the UCC37321 cooks and dies within 1 or 2 seconds. The UCC37322 has only fried once. Trying to understand if the half-bridge circuit is somehow overloading the GDT and causing the gate driver to be driven too hard or could it be something else on the driver circuit? The other question is how to debug this without frying a driver IC on every test? Getting expensive!

As a side note, when the UCC dies there is a direct short from Vcc to GND on the chip which in turns short circuits my 15v power supply.

Ben Solon

Mon Jan 21 2013, 12:33AM

Registered Member #3900 Joined: Thu May 19 2011, 08:28PM
Location:
Posts: 600

can you give us an accurate schematic of your system?

Graham Armitage

Mon Jan 21 2013, 01:13AM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

Here is the half bridge circuit and resonant driver circuit - pretty standard circuit and I know a lot of people on this forum have successfully this particular coil. So I don't suspect a schematic error (more likely my construction). Just not sure where the issue could be and how to go about debugging this without frying components every time.

Physics Junkie

Mon Jan 21 2013, 05:16AM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Hi bushman. I am currently trying to debug a very similar circuit as the one you posted. Actually its pretty much identical, from Dan McCauley's design book, except there are two drivers each in parallel for full-bridge configuration use. It seems we are having a similar issue here. Everything is fine and dandy until i connect the GDT's to the bridge. So far I have not fried any UCC's (yet) but my 15v regulator is on verge of catching on fire when GDT's connected to bridge. Have you noticed if your fried UCC's are whats causing the 15V short? Or perhaps your 15V regulator is shorting first causing UCC's to die. Are you getting a hot 15V regulator too i can assume?
Photos to go with this are the bottom of this page

I will do more tests tmoro, sorry i do not have any helpful insight at the moment. Though hopefully we can help each other out here...

Graham Armitage

Mon Jan 21 2013, 12:29PM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

Sounds very similar. The 15v regulator gets super hot as well. I added an LED on the driver board for a visual queue that the 15V rail is still on. When the UCC driver dies the Vcc and Gnd pins are shorted inside the chip. This creates a direct short across the 15v rail which in turn starts cooking the 15v regulator. It's definitely the UCC short that causes the regulator problem.

When all power is on and everything is hooked up there are no problems with overheating or anything. Within 2 secs of turning the modulator on the UCC shorts and dies. I did have the duty cycle too high at one point, but have that corrected now. I suspect it's something wrong on the half bridge, but not sure how to test it without killing more drivers. The other tesla I built with a very similar circuit is running fine. It's driving me nuts !!

teravolt

Mon Jan 21 2013, 02:29PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

baring a short those dip packages can't dissipate the heat with high frequency switching times for long periods. I prefer IXDN614CI, a TO-220 package.

maybe if C5 the 10uf cap was smaller like 1uf with a 100 ohm parallel resistor the load would be less on your ic. hears how I did it, look at the bottom of the thread

although I use fiber optics you could use a package where the driver and regulator are tied together on one heat sync

Gpete

Mon Jan 21 2013, 03:35PM

Registered Member #5137 Joined: Sun Jun 03 2012, 05:17PM
Location: Manhattan, KS
Posts: 1

Hi,

This must be a pretty common problem because I'm having the same problem. I have an almost identical circuit except the interrupter is different, and I'm running 12V instead of 15. This is my first attempt at an SSTC and it's becoming frustrating. I think Bushman is correct about it being something in the half-bridge I'm just not sure how to test it. I had an idea to use a differential probe in place of the load on the bridge but I feel like since so many people make these there must be some simple way to test the bridges that I'm overlooking. I'm going to go ahead and attach pictures of the output gate drive waveforms just so I'm not crazy. It's not running on feedback in these it's just off a signal generator. I'm sure we'll all appreciate any help thanks in advance.

Graham Armitage

Mon Jan 21 2013, 04:08PM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

I built a SSTC using Steve Ward's design (below) and it works just fine - none of the gate drive problems, even with 50% duty cycle. Apart from additional safety in the new DRSSTC circuits, it's the same concept (from what I can tell). Again - not doubting the circuit but how to debug it.

teravolt

Mon Jan 21 2013, 05:17PM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

I don't know if you noticed but C6 is .1uf vs 10uf. C6 is a coupling cap its XC is 100 times higher. make your coupling cap smaller and add a 100 resistor in parallel. there is probably more current going through the GDT than the IC can supply

HV Enthusiast

Mon Jan 21 2013, 07:19PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

There is definiately something going on with your particular board or gate transformer.

I suspect possible an incorrectly wired gate transformer on the half-bridge side or wrong connections. It definately sounds like either a solder bridge or short based on the symptoms you describe.

I have my money on an incorrectly wired gate transformer or incorrectly installed IGBTs. This would definately exhibit the symptom where the gate driver functions normally without the bridge attached but blows up the chips when it is attached. If the IGBTs are installed in reverse, it will look like a short across the gate transformer.

We have hundreds of these kits in circulation and this is the first time i've heard of this particular problem. (Are your IGBTs installed with the proper orientation???) We also just built (20) units of these for a local Venture Scout group using the same vintage REV boardset you have, so its not likely a board manufacturing issue as these are all working fine.

The number one symptom which you described via email that as a flag for me was that the UCC chips burn up in the absence of modulator signal. This seems to imply that something else is wrong somewhere else.

But again, there is no need to modify, or change values for this circuit. All the values, including decoupling capacitor were sized and optimized according to the specific parameters of the coil.

1 2 3 4 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob