OMG Induction Heater / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

OMG Induction Heater

first previous 5 6 7 8 next last

Move Thread

LAN_403

kim_ladha

Sun Jan 27 2008, 08:04PM

Registered Member #1217 Joined: Mon Jan 07 2008, 11:46AM
Location: Leicester, UK
Posts: 11

Hi Tonskulus,
Nice work. I think you have pushed air cooled work coils as far as they can go!!

Hi HFsstc-freak,
Nice looking board. You should have no problems putting the 50 amp peak curents through your board. I would expect similar (probably better) performance to my pcb. The ultimate limitations is how you get rid of the excess heat. I was running mt setup at 3kw continuously and at just 40 degrees heat sink temperature I got an igbt failure- It takes 20 minutes to reach this temperature. I think the culprit is thermal runaway on the igbts. You might want to try a lower conduction loss version for a larger safety margin on newer designs like yours. The 20n60s are meant to be used for upto 3kw in this topology- I recon you need some kind of forced cooling of the devices to reach 3kw continuously though.

Regards
Karim

Marko

Sun Jan 27 2008, 08:54PM

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

Hello guys, Kim

oh do I hate mysterious IGBT deaths.

40 degrees C is really nothing I'd worry about, your cooling actually looks very good.
If I was really paranoid I'd just use water cooling on IGBT's too.

I don't know what you mean by thermal runaway... IGBT voltage drop actually decreases with increase of temperature. Paralleled devices can run away but even that is fixed well with mutual thermal coupling (bricks),

I don't really know why are your IGBT's dying sir.

Could it be that power dissipation just exceeds the limit (too high thermal resistance die->heatsink?) If that's it maybe it's time for you guys to move on bricks.

Have you tried to measure the output current of your inverter at various loads?

Still I don't think 70amp IGBT's should die so easily.

Richie wrote ...
At the risk of getting my backside kicked for turning this thread into a lesson on inverter testing I'll leave it there.

Hey, nobody will surly get you 'backside kicked', inverter design is like at least half of IH design so any of your help will be appreciated a lot!

Marko

Dr. Dark Current

Sun Jan 27 2008, 09:40PM

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

Sorry no experience with induction heating, but my 40A IGBTs died too with just little warm heatsinks and ca. 10-15A peak current. I dont know exactly why it happened but:
1) IGBTs usually survive worse than similarly rated FETs
2) Too fast gate drive (with too sharp fall edges) decreases the device's turn off current capability.

Experimentonomen

Sun Jan 27 2008, 10:47PM

Registered Member #941 Joined: Sun Aug 05 2007, 10:09AM
Location: in a swedish junk pile
Posts: 497

Kim, did you ever check your gate waveforms ?

I drove 20n60 igbts with IR2110 and the IR2110 got very hot very fast due to the high current needed to drive the igbt gates, do your IR2184 get hot too ?

Maybe you have insufficent gate drive.

Marko

Sun Jan 27 2008, 11:22PM

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

I'd really love to hear what Richie has to say on this 'too fast gate drive' thing...

I didn't ask, kim, you are saying that 40 degrees is heatsink temperature, have you actually attempted to measure it on IGBT's themselves? Maybe with some of those sticky probe thermometers.

Oh, and have you scoped your gates, how do they look like?

Marko

GeordieBoy

Mon Jan 28 2008, 12:59AM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

> I'd really love to hear what Richie has to say on this 'too fast gate drive' thing...

Do you mean for Induction Heaters or MOSFETs or IGBTs in general?

Textbooks usually say "as fast as possible", but in real life this isn't usually the case, and the designer messes about with (read "optimises"!) the switching speeds in order to get a decent compromise between efficiency, reliability and EMI. It's a big subject and depends on the switching topology, devices chosen and the application spec.

I can elaborate more if you want but I'm not sure the discussion belongs here.

-Richie,

Steve Conner

Mon Jan 28 2008, 11:21AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I think a discussion of gate drive would be great, but it's applicable to so much more than induction heaters, that it probably belongs in another thread, like this one

BTW, I've made a set of water-cooled copper busbars to hold my Celem capacitor, a ferrite-cored series drive transformer that also fits into the busbar assembly, and a new work coil. Will post pics when I get my internet working at home again...

Experimentonomen

Mon Jan 28 2008, 04:03PM

Registered Member #941 Joined: Sun Aug 05 2007, 10:09AM
Location: in a swedish junk pile
Posts: 497

I bought some parts:

Marko

Mon Jan 28 2008, 05:28PM

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

yeah guys, moving that to another thread is a good idea.

Steve: cool, is the bus bar a single copper plate turn incorporated onto the transformer?

looking forward for your project.

Hfsstc freak, what is that water-cooled-transformer-like thing in the first pic?

Marko

Experimentonomen

Mon Jan 28 2008, 07:19PM

Registered Member #941 Joined: Sun Aug 05 2007, 10:09AM
Location: in a swedish junk pile
Posts: 497

Those parts are stuff from an Ajax Tocco Meltmaster induction heater cabinet, same one as the celem caps came from.

The transformer is the matching trafo. The boards are the mosfet gate drivers.

first previous 5 6 7 8 next last

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