how powerful can a transistor be?

[little_carlos]

hey guys
yesterday, i was wondering about transistors power, i know some few power transistors, like tip3055, but, wich are the most powerful transistor in the market?

[denelec]

http://www05.abb.com/global/scot/scot256.nsf/veritydisplay/d746404db02cd15683257b5f004a8e33/$file/Anouncement_StakPak.pdf

[Bryan]

Wonder how long those IGBT's can handle those levels of current. Micro/milli seconds.

[denelec]

I think it is 2000A continuously.
Peak current is 14 000A for 10ms.
http://search-ext.abb.com/library/Download.aspx?DocumentID=5SYA%201430-00%2001&LanguageCode=en&DocumentPartId=&Action=Launch
Total power dissipation is 20 800W.  You must need a big heat sink.

[ovnr]

Wonder how long those IGBT's can handle those levels of current. Micro/milli seconds.

More like all day long. These are the kind of devices used in solid state HVDC converter stations - think multi-gigawatt systems running at over a megavolt (bipolar, so /2 relative to ground). They're generally watercooled.

[T3sl4co1l]

Most powerful in terms of dissipation in a single package, probably those.

There are also SCRs, GCTs and other esoterica used in the power realm, which may be higher power but I'm not sure.

Most powerful in terms of a single die, most likely the SCRs or whatever.  They can be made on single wafers.  IGBTs don't have the greatest yields, or good enough matching over the die, so they are usually sawn up into pieces under 25mm or so.  Medium sized modules (600-1800V, 300-2000A) are usually made of many dies, matched, connected in parallel, and bonded to a metallized ceramic base plate.

Most powerful in terms of linear amplification rather than gross switching capacity, probably BJTs.  The current density can be quite high, though that can also lead to runaway (second breakdown), which is true of modern density MOSFETs as well (and probably quite true of IGBTs, but they are never rated for linear operation or pulses longer than 1ms).  Though BJTs were available as power darlington modules back in the day (ca. 600V 300A capacities, and probably more -- but intended for switching only), probably the largest available (or useful) today are TO-264 packaged On Semi transistors (MJE29xxx or something like that?), a quad of them being good for maybe a kilowatt (or more with better cooling).

Most powerful in terms of sheer power density (watts per die area), probably something in SiC, because the operating temperature can be significantly higher -- you have to find a device packaged in something other than plastic to test that though.  The dies themselves are also quite small (it's hard to grow good quality SiC and get good yields).

I don't think there's anything "smart" that can dissipate more, physically speaking -- metal film or foil on a ceramic substrate makes one hell of a resistor, but it doesn't "transist" very well!  If you need both power handling and control, you're better off with a hybrid approach: switched resistors, or something like that.

Tim

[BillWojo]

Well I'm glad that those StakPak come in "Explosion Resistant Package" as per the spec sheet. Don't even want to be in the same building if one of those goes Poof!

BillWojo

[Jay_Diddy_B]

Hi Group,

Here are some pictures of the biggest transistor in my collection. I have included a T0-3 in the picture for scale. The base is 2.5 inches (63mm) square.





This is a single NPN power transitor, BJT. The ratings are:

Ic 300A continuous 450A peak
Ib 100A
Vceo 700V

Maximum Power dissipation at 25C case 1500W

The gain is about 8 -10

It was manufactured by GPS, which is GEC Plessey Semiconductors, used to Marconi Electronic Devices Ltd, MEDL, in the U.K. Westcode Semiconductors made similar devices.


Jay_Diddy_B

[DanielS]

100A base current... up to 70W dissipation for the base-emitter junction alone! Ouch.

[Jay_Diddy_B]

100A base current... up to 70W dissipation for the base-emitter junction alone! Ouch.

Note: There are two emitter connections, the main one on the right and a second one with the red crimp. The smaller emitter connection is the return connection for the Base current.

Regards,

Jay_Diddy_B

[David Hess]

Note: There are two emitter connections, the main one on the right and a second one with the red crimp. The smaller emitter connection is the return connection for the Base current.

I have seen big power MOSFETs and IGBTs (and thyristors) with a similar gate return lead connected to the source at the package.  It is needed when the change in current through the lead inductance is high enough to affect the gate to source voltage which can be a Ghostbusters level of bad.

[David Hess]

Most powerful in terms of linear amplification rather than gross switching capacity, probably BJTs.  The current density can be quite high, though that can also lead to runaway (second breakdown), which is true of modern density MOSFETs as well (and probably quite true of IGBTs, but they are never rated for linear operation or pulses longer than 1ms).

Toshiba made complementary IGBTs intended for linear audio applications for a while.  The datasheets include DC SOA graphs and show the secondary breakdown:

http://www.bucek.name/pdf/gt20d101.pdf
http://www.bucek.name/pdf/gt20d201.pdf

Though BJTs were available as power darlington modules back in the day (ca. 600V 300A capacities, and probably more -- but intended for switching only), probably the largest available (or useful) today are TO-264 packaged On Semi transistors (MJE29xxx or something like that?), a quad of them being good for maybe a kilowatt (or more with better cooling).

Some TO-3 bipolars and darlingtons are still made which reach 300 watts dissipation.  The various module packages are a lot larger of course.