Using Many Small Signal transistors for the output stage Update!!

[John Biles]

Hello Everyone,
Not long ago I brought up the idea about using many small signal transistor in parallel for the output stage of a amplifier.
Someone said to try 1 amp small power transistors like the BD139.
I built up a test circuit with 6 BD139's paralleled in place of each power transistor that I am using.
As you normally use wire wound 5watt resistors on the emitter of each paralleled transistor to promote current sharing, this would mean that I would need 12 resistors to do the job.
As I didn't have them on hand or want to buy them just for this test, I decided to measure 12 lengths of cooper wire until I had 0.4 ohms for each length of wire. This would take the place of the resistors. By the Pic Attached you will see that I have wound all 12 lengths of wire on to one former in the middle of the test board.
I know that the heat sinks are too small, but I will be running each test for seconds not minutes, just to hear the effect.
So before I wire the board up, I decided to put a pic on the forum just in case it all goes up in smoke.

 

[Chris Daly]

Did it work ?

Cheers / Chris

 

[CBS240]

What if current in one wire inducting current in another next to it affects bias and current sharing? is this a possibility?

you could get an ordanary resistor and wind the wire around each one seperately...making wire wound resistors.

 

[AndrewT]

Hi,
large diameter coils have large inductance.
Will this affect your assumptions?
A zig-zag pattern on a non conductive former will have almost no inductance.
Alternatively, wind the coil using half the wire length in one direction around your former and then reverse the winding direction to finish the second half of the coil. A long small diameter coil wound on a 5mm wooden dowel will have less inductance than your 20mm square coil, still using the two direction trick. Each coil would then be self supporting and independant.

 

[John Biles]

Hello Everyone,
When i connected it to my Amp the bias transistor and driver transistors in my amp went up in smoke in a big way and also one lot of the BD139's in parallel look like they bit the dust as well. Before I removed the output transistors, the amp had been working for months with no problems.

For this Amp I think I will run just 1 or 2 normal high ampage transistor on the output.

May build a low voltage single supply (10-16volts) amp to continue the tests?

 

[13DoW]

Don't forget that the ballast resistor for each BD139 should now by 6x the value of the single ballast resistor on the emitter of the single power transistor.
If you kept the value the same then it will have little effect at the lower collector current that each BD139 is working at and might not prevent one of the BD139s from thermally running away.

I would stick with discrete resistors rather than messing around with all that wire.

Regards
13th Duke of Wymbourne

 

[Dr. Photon]

I'd get some regular resistors and dump the jumble of wires. Looking at the circuit without knowing what it is it looks like some kind of inverter

Does the board simply have three leads and connect to the old amp's transistor pads, so that you have the amp board's individual emitter resistors AND the amp's old emitter resistors?

I would still scale the 6 individual emitter resistors up a little bit. I would also recommend individual base resistors on each transistors base. maybe 10 ohms or so?

 

[John Biles]

Hello Everyone,
Any advice is appreciated on this circuit, please keep your ideas coming.
Thanks for your replies.

 

[Ultima Thule]

Yes, trash that "ignition coil" and replaces with real resistors and I will look forward to take part of this thread!

Cheers Michael

 

[tlf9999]

one reason could be thost 0.4ohm "resistors": they are effectice inductors / transformers that send signals back and forth among those transistors. Not a good idea I would say.

 

[CBS240]

You know if you want it look similar to the picture, make a transformer coupled push pull for output like the tubeys do, then you could use more voltage and less current for CT primary and then match speaker with secondary. 20W audio output transformer? Also can use just one type of transistor since your whole row of BD139's went Poooffffff and the smoke fell out of them.

just another idea.


Still need emitter resistors....