Transistor with highest power rating ever?

Something made me recall a hoax article I saw in PE or Elektor a long time ago. It announced a 1kW BJT and the accompanying picture showed a TO-3 style case that appeared to dwarf a wrench placed alongside it. Obviously they used either a composite image or a tiny special-purpose wrench.

In the late 60s and early 70s, manufacturers produced metal case

Ge. I think the power ratings went up to about 250W, maybe 300W,

Out of curiosity I just checked at the element14 site and they have two 400W types. What is/was the highest power rating ever for a production transistor?

Reply to
Pimpom
Loading thread data ...

There are several 1500W RF transistors, in production.

--
Never piss off an Engineer! 

They don't get mad. 
 Click to see the full signature
Reply to
Michael A. Terrell

Ah, I neglected to check RF transistors at element14. Now I see that they have two 1.8kW FETs there.

Reply to
Pimpom

I have some 300 amp 1200 Volt darlington modules, Fuji, iirc, that came out of an older motor control inverter. Measure around 4" x 3" x 1" each overall. The heatsink for the box measured around 18" square, 2" finned depth...

Chris

Reply to
Chris

There are some Unijunction transistors on ebay claiming 350MW, which is pretty impressive for a TO92 package.

Cheers

--
Clive
Reply to
Clive Arthur

There's a big difference between maximum package dissipation (likely taken by some large hockey-puck IGBT) and the amount of power the device can *switch*.

Low Rds(on) MOSFETs can switch huge amounts of power without much dissipation.

Which are you interested in?

Reply to
Clifford Heath

Nothing in particular as far as any foreseeable applications are concerned. I was just curious about how much rated dissipation levels have been raised.

I don't have a lot to do with high power switching. Even so, I have some general-purpose MOSFETs with impressive capabilities given their low cost and small size, like a 40V 54A 8-milliohm type that cost me ~10 US cents in small quantities.

Reply to
Pimpom

That tops what I found at element14 by a huge margin. One TO-92 BJT was rated for 625W.

It's common to see people write MW when they mean mW, but that one at element14 continued to use W on the details page.

Reply to
Pimpom

There are some TO247 mosfets rated for 1KW dissipation, like IXFH400N075T2, and I think I may have seen 1.5KW somewhere. Of course, that would take heroic heat sinking, and I don't think I'd be brave enough to try to dissipate a kilowatt from a single TO247.

IXTH110N25T is rated 694 watts.

I think there are mega-package (big multi-chip potted things) mosfets and IGBTs that will dissipate several kilowatts.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Mount inside some kind of finned copper tube, and use push/pull fans to shove air down the barrel?

Reply to
bitrex

That has to be the understatement of the Millennium. ;->

--
This message may be freely reproduced without limit or charge only via  
the Usenet protocol. Reproduction in whole or part through other  
 Click to see the full signature
Reply to
Cursitor Doom

IOW a chimney. That's what we call them in the UK anyway. More commonly seen on valves (toobs).

--
This message may be freely reproduced without limit or charge only via  
the Usenet protocol. Reproduction in whole or part through other  
 Click to see the full signature
Reply to
Cursitor Doom

You disappoint me, John. I thought you'd try to wring maybe 3-4kW out of one. At least. Then put it into production. ;->

--
This message may be freely reproduced without limit or charge only via  
the Usenet protocol. Reproduction in whole or part through other  
 Click to see the full signature
Reply to
Cursitor Doom

I was thinking along those lines myself and that's one of the reasons I asked here instead of just searching for sources of existing ultra high power devices on the net. I was wondering if there might have been some super-power transistor that some manufacturer had marketed and then discontinued due to lack of demand. Potential users could be put off by the challenges of dissipating so much power from a single small package.

I further surmised that some manufacturer might have single-mindedly tried to cram more and more power into a small package, somewhat like the way Intel once concentrated mainly on raising their processor clock speed.

Reply to
Pimpom

The problem is the immediate vicinity of the package, the spreading thermal resistance. You could probably bolt it to a half-universe of solid aluminum and still not keep it safe at 1KW. And forget trying to use an insulator. A half-universe of diamond might work.

Mosfet power ratings are often science fiction.

Here's a 17 KW peak output amp:

formatting link

The fets are clip mounted to nickel-plated copper heat spreader bars, no insulators. The heatsink is the electrical output and all the P and N fet drains connect directly to that. Lots of fets spread out the heat.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Maybe they only need to fire once.

Like this:

formatting link

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

We tested a lot of kilowatt-type mosfets to destruction, in the 100 millisecond sort of pulse range. Most failed explosively well inside their claimed SOAR.

formatting link

Sometimes you can safely push a part beyond data sheet specs. And sometimes you have to derate by 3:1.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

IR set the standards for cheating on mosfet power dissipation specs, and it looks like everyone else had to lie to be competitive.

Fets are cheaper than fans and heatsinks, so spreading out a few small fets is better than creating a hot spot with one giant fet. A fan-cooled heat sink might be 0.05 K/W across its entire surface, but it won't be that good at one spot.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

The IXFH400N075T2 data sheet says that the Rjc is .15C/W. At 1kW that puts the junction at 175C already (ambient 25C).

Now add in the Rcs. That thing will never be able to dissipate 1kW continuously as their SOA curve shows. Their curve must be based on an infinite heatsink and zero Rcs.

Reply to
John S

Any chance that was an April issue?

John ;-#)#

Reply to
John Robertson

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.