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 8 9 10 11 next last

Move Thread

LAN_403

Steve Conner

Wed Mar 12 2008, 10:58PM

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

Voila, melted bolt with around 1kW at 250kHz. At last I managed to tear myself away from the Internet long enough to do something practical

GeordieBoy

Thu Mar 13 2008, 12:16PM

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

Looks good Steve! ...but what did that poor bolt ever do to you!?!?

The switching waveforms look very clean on the scope - No doubt due to the actions of the board on the left?

Thanks for all the pics. Do you have a close-up of the ferrite power transformer used to couple power into the series-resonant tank?

-Richie,

Marko

Thu Mar 13 2008, 07:56PM

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

Steve that is awesome, welcome back

Oh, any transformer closeup pics for me too?

Are you using anything to control the current (hysteresis on-off control?) to regulate power and prevent the system from blowing up without a workpiece?

Hope to see that thing boxed up and put to use now

Just a noob question for the end, seeing Steve is using a series resonant system driving a practically resistive load at quite high frequencies with larger IGBT's, would there be merit to use a commutating inductor at the output of your bridge? Or maybe just put some small air gap onto the transformer core?

Keep this work up, all of you guys,

- cheers,

GeordieBoy

Thu Mar 13 2008, 08:24PM

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

seeing Steve is using a series resonant system driving a practically resistive load at quite high frequencies with larger IGBT's, would there be merit to use a commutating inductor at the output of your bridge?

I'd guess that if he wanted a slightly inductive load from the series resonant system he could just tune the driver slightly on the high-side of the tank circuit's resonant frequency, and the inverter would see a lagging current.

Or maybe just put some small air gap onto the transformer core?

...or like you said the magnetising inductance of the ferrite transformer will also act like a commutating inductor and contribute a small lagging component to the inverter load current.

-Richie,

Steve Conner

Mon Mar 17 2008, 10:20AM

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

Hi guys

Just an update on this...

For those who asked, I'll write this up as a proper project and give schematics and pictures of the transformer after Easter. I'm really busy just now.

I actually blew up the IGBTs trying to do levitation melting! I made a coil like the one Ameritherm show on their website, and it would just nearly levitate a small piece of aluminium. So I turned the current limiter off, turned my variac up to 260V, and I had the lump of aluminium floating and glowing a dull red for a few seconds, and then guess what happened.

The driver board is one of my Mk2 DRSSTC drivers, unmodified apart from the interrupter signal being jumpered permanently on. I found that it can't lock by itself when DC bus voltage is first applied. It needs the tuning pot to be twiddled until it locks. (this is best done at reduced DC bus voltage!) Once it's locked though, it seems to stay locked reliably as workpieces are moved in and out of the coil.

The DRSSTC current limiter seemed to work fine, untill I turned it off! When the workpiece was removed it would kick in. It made some audible noise, but didn't seem grossly unstable.

The transformer has 20 primary turns (IIRC?!) and one secondary turn of water cooled copper plate. The core is one of a selection that I bought off EastVoltResearch years ago, and I have no idea what grade of ferrite it is.

The IGBTs were some Fairchild FGH50N6S2D that I had left over from my SSTC experiments. I'm surprised they lasted so long, especially since they would have been running over 300kHz with the second work coil. I'll try again with a full bridge of IRFP460s. (These will probably be happier with a lagging current and some extra deadtime to give ZVS, as Marko and Richie suggested.)

More info after Easter! :P

GeordieBoy

Mon Mar 17 2008, 02:05PM

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

Thanks for the info. I look forward to seeing more details when you have time.

The IGBT failures may just have been down to poor heatsink contact and thermal resistance in the Silpads. With better heatsinking and positive pressure over the die you could probably have run them harder.

-Richie,

Steve Conner

Mon Mar 17 2008, 02:21PM

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

Yes, I forgot to mention that, the heatsinking was so bad I'm ashamed of it! The heatsink was pretty hot after it failed, and the devices were probably a lot hotter still due to poor thermal contact. Whatever devices I use next time, I'll be sure to try harder with the thermal management...

Marko

Mon Mar 17 2008, 02:50PM

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

Hi guys, thanks steve a lot for the post!

Are you using all of your celem caps, Steve? (don't see well behind that cup?) Your resonant frequency seems fairly high, I don't think you need closely that much unless you are for some special sort of surface annealing.

Are you limited with your capacitance vs. sane dimensions of the coil?

Under 50kHz seems to work pretty well, and people seem to do very well with common bricks like CM150's (I'm sure you have some around?), tons of power and thermal contact area.

Yes, I forgot to mention that, the heatsinking was so bad I'm ashamed of it! The heatsink was pretty hot after it failed, and the devices were probably a lot hotter still due to poor thermal contact. Whatever devices I use next time, I'll be sure to try harder with the thermal management...

Oh, didn't want to moan, I see you used a To3 sil pad under one IGBT? :O

Richie always recommended to actually use no sil-pads if possible and use separate heatsinks for each device.

ISOTOP and brick packages are OK; but due to their internal thermal insulator they may even have higher thermal resistance than direct contact of uninsulated TO247.

I even considered to solder the TO247 to copper but Richie didn't like that idea too much over simply screwing down the devices directly.

Experimentonomen

Mon Mar 17 2008, 03:16PM

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

I hate mosfets! They blow instantly in my circuit, only igbtÂ´s work reliably.

Steve Conner

Mon Mar 17 2008, 03:17PM

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

Hi Marko

I've only got one Celem capacitor, HFSSTC-Freak got to the Ebay auction before me

With a rating of 0.4uF, 500V RMS, 200A RMS, it would deliver its maximum 100kVA at 160kHz, but to get the frequency that low, I'd need either a very big work coil or some very small copper tube. I didn't want to make the work coil too big in case it failed to concentrate the magnetic field enough to melt or levitate anything. And I'm already using 4mm tube which is pretty thin.

first previous 8 9 10 11 next last

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