IGBT flyback driver / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

IGBT flyback driver

1 2 next

Move Thread

LAN_403

Dr. Dark Current

Mon Aug 27 2007, 05:03PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Hi all,
after some problems I finally got my IGBT flyback driver working reliably. The gate driver is built around the IR2153 half bridge driver ic and its outputs are boosted by TC4452's. The halfbridge is made of IRG4PC40W Warp speed IGBTs and MUR1560 ultrafast clamp diodes (2 in parallel per igbt). Resistive inrush current limiting, shorted by a turn-on delay relay.
The bridge drives 2 ac flybacks with 13 turn primary each, connected in series. Secondaries also in series (with "midpoint" grounded). No airgap.

Total arclength up to 20cm (8"), ca. 1kVA input max.

Pics

Gate driver

Flybacks

I only tested it on a Jacob's ladder, since I feel a bit unsafe by drawing an arc with a hot electrode with 15kV on it...
Link to youtube video

Some captures from the video

Have a nice day

J.M.

Marko

Mon Aug 27 2007, 05:30PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

I fell sorry for those flybacks.. how much more can they take?

For those small IGBT's I don't se what's wrong with drive as simple as single SG3525 and GDT, actually I've driven larger gates that way without problems.
Again you may use bigger IGBT's with it if you come into need..

Have you considered building your own transformer for that circuit?

Dr. Dark Current

Mon Aug 27 2007, 05:38PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Marko wrote ...

I fell sorry for those flybacks.. how much more can they take?

For those small IGBT's I don't se what's wrong with drive as simple as single SG3525 and GDT, actually I've driven larger gates that way without problems.
Again you may use bigger IGBT's with it if you come into need..

Have you considered building your own transformer for that circuit?

Hi firkragg, I simply don't like GDT's. I think my driver outputs almost perfect gate drive signal free of ringing and is virtually indestructible...

I've often thought about making my own high voltage transformer, but a large ferrite core seems quite impossible to get here.

uzzors2k

Mon Aug 27 2007, 09:13PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Madness!

Guess what I'm doing once my bench supply is done. Is the awesome arc in your avatar also from this driver? Just out of curiosity, how does the performance differ from a mosfet half-bridge?

I had planned on putting 4 monitor flybacks in series under oil, but I don't know if they will beat this.

sparky

Mon Aug 27 2007, 09:26PM

Registered Member #530 Joined: Sat Feb 17 2007, 07:56AM
Location: Victoria BC, Canada
Posts: 178

Impressive - gotta build me one of those :) Looks dangerous.... and I like that!

Dr. Dark Current

Tue Aug 28 2007, 07:21AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Uzzors wrote ...

Madness!

No the arc in my avatar is from some pole transformer captured from youtube video

The performace really depends on how powerful devices you get- I think IGBTs are more suited for higher powers than FETs.

I have some problems with the arc starting, sometimes when it starts on the bottom of the ladder it sputters and bounces around like mad like if you were blowing at it from all sides, and it stays at the bottom. Restarting the driver (only the power part- gate driver is connected all the time) sometimes helps the arc to start again normally. This is really strange and I'm not sure whats causing it...

Edit- hmm strange, I rewired the flybacks with primaries in parallel and the sputtering seems to stop

sparky wrote ...

Impressive - gotta build me one of those :) Looks dangerous.... and I like that!

thanks, I think it outputs enough current to kill you

,it certainly is dangerous but nothing is dangerous when you know how to work with it

I've made a new video with a bigger Jacob's ladder here

Marko

Tue Aug 28 2007, 11:30AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

The performance really depends on how powerful devices you get- I think IGBTs are more suited for higher powers than FETs.

For low powers like two flybacks mosfets probably still win over IGBT's in price.

Those are quite large IGBT's from other side.
If you ever get larger cores you could use SLR and push like several kW of power with them.

Dr. Dark Current

Sun Sept 23 2007, 03:07PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

:-( Today my driver exploded, and I have no idea why. Current draw from the filter cap was not higher than 8A (8A fast fuse didn't blow) and the igbt's are rated at 40A cont. Anyway, I was getting up to 1' fat arcs from four flybacks before the driver died, here probably ends my hunt for the longest fattest arc from a ferrite cored transformer.

uzzors2k

Sun Sept 23 2007, 03:34PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Failures like that might be caused by a failure in the drive circuitry, I can't see an IGBT like that failing otherwise. If the gate signal sunk to low and the IGBT went into the linear range it would die pretty quickly. Did you take a picture of the monster arcs?

I put together a super flyback driver a few weeks ago, with over current detection, duty cycle control and the works. I had planned on putting four monitor flybacks in series, but I only got to two. In the process my oldest (and very first) flyback died.

It was 2 years old, and had survived every driver until then. That marked the end of my flyback experiments.

Dr. Dark Current

Sun Sept 23 2007, 04:37PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Uzzors wrote ...

Failures like that might be caused by a failure in the drive circuitry, I can't see an IGBT like that failing otherwise. If the gate signal sunk to low and the IGBT went into the linear range it would die pretty quickly. Did you take a picture of the monster arcs?

My gate driver works fine after the failure, so this was not the problem (with a TC4452 for each IGBT, I think it's rather undestructible

) . Sadly I have no pics of the arcs, the driver died just before I planned to take a video

The failure was pretty violent, the fuse is completely black from inside

so I guess one igbt went short circuit for an unknown reason and the other one shorted out the power supply (clamp diodes are fine).

Update- one of the gate drivers was blown but I believe this was the result of the failure (gate voltage went too high/ outputting to short etc.)

1 2 next

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