IGBT signals problem / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

IGBT signals problem

previous 1 2 3 4 next

Move Thread

LAN_403

Physics Junkie

Wed Feb 13 2013, 06:14PM

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

HB wrote ...

I could be wrong but i think your signal looks messed up when only the interrupter is going because the gdt is designed for the frequency of the of the feedback signal not the interrupter frequency and that is why when you have both signals going its it looks better if that makes any sense.

Oh yes I get you. And thats basically what I thought too. But even with a signal generator simulating a perfect waveform sometimes the igbt gates are all messy, but sometimes its normal, and thats what I'm trying to figure out. Say for example that it turns out that silly waveform from the interrupter really is just the core saturating, then most likey I have a faulty component on the H bridge. I tested the 5.1ohm resistors and they read 5.7 on the multimeter. It would have to be a bad diode because they are oriented correctly and so are the IGBT's.

Graham Armitage

Wed Feb 13 2013, 08:32PM

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

Really like what HB said. So I tried to replicate what you have. I did not disconnect everything, but put the scope leads on the driver outputs (GDT primary) and ran just the interrupter. This is what I got - it was horrible.

Yet my tesla runs just fine. I probably would have stopped too if I saw this wave form, but never thought to test without the feedback signal. As long as the IGBT gate signal was looking good, I was happy. I guess ignorance is bliss. Hope this helps.

Physics Junkie

Wed Feb 13 2013, 08:48PM

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

Okay good so it turns out it is just the start up oscillation. Thanks a lot Bushman! Now I got to figure out why it gets disgusting as soon as I attach IGBT's. Anybody know in what case would a good IGBT be screwing up waveforms as soon as they are connected? I have 12 ixgn60n60c2d1 brand new, I've used 4 for testing, everything looks good on the secondaries of the GDT, but as soon as I attach IGBT's it goes all rolly polly on me. Gate capacitance maybe? Heres a video of attaching the GDT->half bridge board->IGBT's step by step showing the change in wave each time

It gets really bad with the IGBT's not really sure why. I tried this on another identical board also with the other cores, didn't make a difference.

EDIT: Also tried with 10 ohm resistor to try and diminish overshoot but it hardly had an effect. The 5.1 ohm's are still good they're not open or anything. Could a nasty waveform like that be happening by exceeding the maximum Vge of +-30V??? As far as I can tell, and everyone else can see by the vids I've posted, the circuit works, and the half bridge sections are most definitely built correct. So I'm at a complete loss here. And yes the IGBT's are oriented the right way...

teravolt

Thu Feb 14 2013, 06:36AM

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

It seems like your circuit is inductive thats what may cause your overshoot. Why do you have that big beed on the leg of the transformer. Try adding some series resistance to reduce your overshoot around 10 ohm to 100 on the driver side. I don't think that there is anything wrong with your bridge board. You could add turns to the core, 15 turns. Is your core the right material or is it saturating? it seem strange that you and Bushman had so much truble with that part of the circuit. Is the driver board a EVR artwork?

Physics Junkie

Thu Feb 14 2013, 07:33AM

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

Its not the overshoot I'm worried about. I've tried more resistance on the primary side, makes no difference. Cores are correct type from EVR, the bead is a noise filter that came with the core, you can see on his website. Removing the bead makes no difference. I've tried 3 cores, all appropriate materials, all give the same results. And like I mentioned above, the waveform gets silly only after I add the IGBTs. I'm trying my best to stay close to the original design here as it's known to work already. The design calls for 10 turns. I tried a core with 14 turns on my last board and from what I remember, it hardly made a difference, 10 turns gave better waves on the secondaries.

Teravolt, what do you think is causing the messy waveform with the IGBT'S in place? If you think an inductive board is the problem then why do I get perfect waveform on the secondaries without the bridge attached? Unless you mean that the bridge board is the inductive one. (Refer back to the first post for board artwork info)

Physics Junkie

Thu Feb 14 2013, 11:03AM

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

FINALLY THE PERFECT WAVEFORM!

+-20V give or take a few here with small overshoots, and 150ns rise and fall

All I did was try my other half bridge section (the one with the 10ohm gate resistors instead of 5.1) and I grabbed two new IGBT's from my stash, and there you have it....

1360839493 7267 FT150581 Perfect Waveform

I haven't thoroughly assessed what the problem is on the other half bridge board because it's 6am and I've been up all night doing tests but most of all I just want to go kill aliens on Halo4 lol.
But this is my conclusion:
a) The igbt's got damaged somehow unknown (probably when screwing them to the heatsinks and/or problems from my last circuit might have caused them damage)
b) A component, most likely a diode, is faulty or was damaged from prior tests.

Tomorrow (later today

) I will switch the IGBT's to see whether it is busted IGBT's or a components on the board.

Graham Armitage

Thu Feb 14 2013, 12:22PM

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

Hey, glad you got it working - will be interesting to see which component it is. Hopefully these threads will help the next person going through debugging like this. Hope to see you fire up the coil soon

Physics Junkie

Thu Feb 14 2013, 06:58PM

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

Much thanks to you Bushman! So far I've found that the other IGBT's are still good. I'm taking the components off one by one to see and will get back with results. I have no doubt these threads will help other nubs find their way.. Looking back on it myself I just feel like an idiot lol.

Anyhow, I took a close look at the rise and fall, and I'm just looking for a second opinion here, time base is 250ns. Does this look a little too slow? I might go back to the 5.1 ohm gate resistors regardless.

Graham Armitage

Thu Feb 14 2013, 07:47PM

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

Would not feel like an idiot - I think these are legitimate areas of confusion and exploration. Just think about how much you have learned from this - I know I have. If it was that easy everyone would be building them :)

With regards to the rise time, I found exactly the same thing. When I switched from the 5 to 10 ohm resistor (while debugging) it killed my rise/fall times. I was getting just under 200nS with the 5 ohm and it jumped way higher with the 10ohm. So if there are no wierd effects or heating with the 5 ohm I would go back to that.

teravolt

Thu Feb 14 2013, 08:02PM

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

the wave form looks beautiful. the gates of your fets are like capacitors and maybe the inductance of the transformer is resonant or making a odd wave shape plus your cores have a defined frequency response. hope you show us the coil when you are done.

previous 1 2 3 4 next

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