Biggest Complementary Transistors

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
"In terms of power handling a transistor rated at 250 watts might only be good for around 20 watts at Vce = 100v. In this case you would be far better off deploying a mosfet solution if you wanted fewer output devices."

Unless they are MJ21195/96 - just down from 250W to 200W at 100V. 5 times better SOA at this voltage than the 2SC5200/2SA1943:s! :p And this is at room temperature, the hotter the heatsink the worse the Toshibas will be in comparison....

I don't see why you would want bigger devices than this - getting low enough thermal resistance will require an enormous package and heatsinks with insanely thick baseplates.

I haven't seen any bipolars more suitable for high power than the MJ21195/MJ21196 pair in TO-3.
 
megajocke said:

I don't see why you would want bigger devices than this - getting low enough thermal resistance will require an enormous package and heatsinks with insanely thick baseplates.

That's what I prefer anyway. :D Brute Force Power! Even small transistors get hot with a thin-base Heatsink.

They make IGBT in big rectangle flat packages, and they mount easy. IMO they could do the same with bipolars! See the package of this IGBT: http://www2.scut.edu.cn/scutco/main...uments/datasheets/standard/BSM200GA120DN2.pdf

The cost-savings would be better with a few big ones, because more parallels requires more emitter-sharing resistors and sometimes more base stopper resistors too. Also, most folks (I see it all the time here) have trouble getting an amp really stable with a ton of outputs in parallel. If the big transistors have low gain, then just drive them with some high gain drivers.

Now for REALLY high power, a few pairs of monster transistors could perform as well as 10 pairs of smaller transistors, without the complexity.


Thick base heatsinks is where it's at baby! :cool:
 

Attachments

  • basefins.jpg
    basefins.jpg
    30 KB · Views: 324
That's because these heatsinks are indeed meant for forced convection cooling, and not natural convection cooling.

A: If the base has a 1/2" thickness, then the fin spacing is 1/4"=> petty much standard for forced cooling.
B: The fins are VERY thin=> Forced.
C: No anodising=> Forced.
D: ratio of fin depth and fin thickness >>40=> Forced.
E: Schedule an appointment at the optometrist.
F: I've just had my eyes lasered, but i can cancel the appointment.
 
Depending on how hot they get IMO determines if you need a fan or not.

I've gotten one of these sinks somewhat-hot with an 80W TEC to test, and it does just fine without a fan, but a fan makes a huge difference, and hardly gets warm. These heatsinks to me look like they were meant to be fan cooled because the fins are close together.

I've used fans on heatsinks in an amp that are meant just for convection cooling, with a bit wider fin spacing, and the fans always make a significant difference. A little wasted air IMO isn't a big deal.

Now for the amp I'm building, I'm using regular NJL/MJL4281/4302's, however, if I build a large power supply, amp, or whatever, I'd like to mount some monster transistors with the other sinks I have.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.