Transistors in parrallel to increase collector current

[ikalogic]

Hello,

The answer in obviously yes, but i just want to be double sure it works that way in the practical world:

Putting Transistors in parrallel will increase their maximum collector current to the double, right?

 

[Papabravo]

Generally you don't want to do that with bipolar junction transisors. Biasing would be tricky, and current hogging would be a distinct possibility. It is however possible with MOSFETS.

 

[audioguru]

Audio power amplifiers use emitter resistors to balance the current gain of paralleled transistors and many circuits use them that way.
Here is one:

 

[Papabravo]

I never had to design an audio amp. Since the original application was not specified, I was thinking switches or motors so I stand corrected.

 

[ikalogic]

Generally you don't want to do that with bipolar junction transisors. Biasing would be tricky, and current hogging would be a distinct possibility. It is however possible with MOSFETS.

i made a search about current hogging, and as far as i understood, this problem can appear if the base's of the trasisitors are not connected together.. but as far as i am going to connect the exact same BJT fully in parallel.. is it still a problem.

Does using a tip122 is a darlington pair with a couple of resistors (used to stabilise the gate input, i guess, the datasheet doesn't say) will allow me to use it?

 

[ikalogic]

yes it is for motors... they shall be all connected to the same motor..

 

[Papabravo]

Experience tells me that you should size a single transistor to handle the load and not mess around with a parallel connection of bipolar transistors. I also question the need for a gain of 1000, typical in a darlington, for driving a motor. You will notice that in audiogurus amplifier there were complementary NPN and PNP transistors with emitter resistors. Are you planning a similar arrangement to drive your motor or did you have something simpler in mind?

Posting a schematic would save a bunch of text messaging back and forth.

BTW, what kind of motor are we talking about?

 

[audioguru]

Transistors from one manufacturer with the same part number have a difference of current gain of 3.5 times for a 2N3055. It is important to match the gain of paralleled transistors in linear applications.
The saturation voltage of switching transistors have a difference of about the same, 3.5 times.
Darlington transistors also have a wide difference.

Therefore emitter resistors should be used to balance the difference in current gain and saturation voltage to avoid hogging.

 

[audioguru]

BTW, what kind of motor are we talking about?

Is he the guy who wants to fry a toy car?
He said its motor smelled funny and got very hot when it had a battery with 3 times more voltage than its rating.

3 times the voltage creates 3 times the current, doesn't it?
3V times 3A= 9 times the power isn't it?

 

[ikalogic]

Ok i agree you are right it is not switable for a motor driver...

But, why the mosfets will accept to be connected in parallel without creating hogging...?

 

[dknguyen]

Because MOSFETs take "no" current through the gate? I'm not sure what he means by current hogging but I have an idea...

 

[audioguru]

Current hogging happens when the transistor with the highest current gain or with the lowest saturation voltage carries most of the load and fails from overload. Hogs push the others away from the food so they can eat the most.

When a transistor gets hot from powering a high load, its gain increases and its saturation voltage drops, so it hogs even more of the load. It is called "thermal runaway".

Mosfets turn off a little when they get hot so they share a load more evenly when paralleled.

 

[Nigel Goodwin]

i made a search about current hogging, and as far as i understood, this problem can appear if the base's of the trasisitors are not connected together.. but as far as i am going to connect the exact same BJT fully in parallel.. is it still a problem.

In ALL cases you MUST use balancing resistors in the emitters, doesn't matter if they are the same batch or not - you MUST use them!. Calculate their value from the current you're going to pass through them!.

 

[Analog]

with properly sized emitter ballast resistors, this is done all the time with bjts.

 

[ikalogic]

with properly sized emitter ballast resistors, this is done all the time with bjts.

how can i use emitter ballast resistors in an H bridge ?

can i use it for darlington pair transistor?

thx a lot.

 

[Nigel Goodwin]

how can i use emitter ballast resistors in an H bridge ?

In exactly the same way as you would any other transistors in parallel.

can i use it for darlington pair transistor?

Yes, exactly the same as a single transistor.

 

[audioguru]

how can i use emitter ballast resistors in an H bridge ?

can i use it for darlington pair transistor?

Darlington transistors have a wide variation in current gain and saturation voltage like single transistors.
They also have thermal runaway like single transistors.
They also have an emitter terminal like single transistors.

 

[ikalogic]

okay, but recalling the structure of an H bridge, there are no resistances in the path of the current going to the motor..

if i put resistances on the emmiter of the transistors, this will case serious problems.. am i wrong?

 

[audioguru]

okay, but recalling the structure of an H bridge, there are no resistances in the path of the current going to the motor..

if i put resistances on the emmiter of the transistors, this will case serious problems.. am i wrong?

The small loss of voltage across the emitter resistors and the small amount of heat from them are the small costs of reliably paralleling transistors.

 

[ikalogic]

Okay thanks a lot eveyone,

Learned a lot, and finally, i bought higgher currents transistors.