Transistors Substitute

[ali_moinuddin]

Hi Guys ,

I have a Amp circuit in which use BDT86 & BDT85 as output transistors. (circuit is powered with +-25 volts DC)



I have with me 2sc2922
and 2sa1216

also 2n2955 2n3055



can i use these combination instead of BDT86 and BDT87 as they are not available in market

please advice

thanks

[lineup]

Datasheet:
http://www.halfin.com/comset/catalog...5-86-87-88.pdf

BDT85, BDT86:
20 MHz, HFE>40, Rjc=1.0 C/W, 125 Watt

Using 2SA1216, 2SC2922 is an upgrade
and would work just fine
----------------
do not use the other (2955/3055)
would be terrible downgrading

[jaycee]

I agree. I'd be skeptical of the original outputs anyway. 125W dissipation from a TO-220 package?? I think not!

The only issues with the sankens are a) are they the REAL thing, and not fakes, and b) will they physically fit ?

[lineup]

A usual MAX for some TO-220 is 90 Watt, 1.4 C/W
Like BD809/810 and BD911/912.

There are a few TO-220 packages rated 125 Watt. With Rjc=1.0 C/W.
Among them we have MOSFET IRF540/9540, IRF640/9640.
This DOES NOT mean you can use them continously at 125 Watt.
With very very good cooling, maybe at 50% ( 60 Watt).

This is something you know very well, jaycee
The difference of rated vs. usable power.
BDT86, BDT85 are rated 125 Watt.


Replacing 20 MHz 125 Watt devices with 40 MHz 200 Watt
is an upgrade.
What you may need, is increase compensation capacitance. Because 2SA1216/SC2922 is faster.
But if you are lucky they will not oscillate, because the original compensation was done with some margin.

Using 2SA1216, 2SC2922, of course, you have to check the PIN configuration.
I am sure they will no way fit in the original PCB

[megajocke]

"What you may need, is increase compensation capacitance. Because 2SA1216/SC2922 is faster.
But if you are lucky they will not oscillate, because the original compensation was done with some margin."

You got that the wrong way round, faster outputs might let you decrease the cdom cap if anything as there is less phase shift in the outputs.

Local instability is possible though sometimes with faster parts, base resistors may be needed if not already installed.

[CBS240]

I agree. Particularly with CFP output arrangements.

You must allow for derating of the device, especially for continuous linear operations. There is some TO-220 that are rated higher than 125W. FPQ47P06 is rated at 160W But, derate at 1.06W/degree C. So, if the die temperature is 150 degrees, then 125 X 1.06=132.5W 160W - 132.5W=27.5W.
FQP50N06 is rated at 120W but derate only 0.8W/degree C. So 120W - (125 X 0.8W)=20W @ 150 degrees. Not that big a difference. It is good to design for worst case scenario. Realistically, for a linear amp music signals will not be a constant output power anyway, 1/5 of full power or something like that.

[Mooly]

A lot of BDT types were darlingtons, used a lot in Philips stuff. Don't know about the 85/86's though.
Lineups data sheet seems to indicate not, dunno, changed a lot of these ( BDT65 etc types, ) over the years, and they were darlingtons.

[jacco vermeulen]

BDT85/86 were Philips(Mullard) stuff, but not darlington.
Typical fT of originals is around 10MHz, not 20.

[Mooly]

Wasn't sure on the BDT85/86. Thanks.

[lineup]

Originally posted by lineup
Datasheet:
http://www.halfin.com/comset/catalog...5-86-87-88.pdf

BDT85, BDT86:
20 MHz, HFE>40, Rjc=1.0 C/W, 125 Watt

Using 2SA1216, 2SC2922 is an upgrade
and would work just fine
----------------
do not use the other (2955/3055)
would be terrible downgrading

Originally posted by Mooly
A lot of BDT types were darlingtons, used a lot in Philips stuff. Don't know about the 85/86's though.
Lineups data sheet seems to indicate not, dunno, changed a lot of these ( BDT65 etc types, ) over the years, and they were darlingtons.

Yes, Mooly.
My first association was also DARLINGTON.
But when consulting datasheet I did see they are not.

I also know that some Elektor power amplifiers have used BDT.

Some BDX transistors were darlingtons, unless I am mistaken?
Maybe there were darlingtons among both BDT and BDX ...