Thermal drift in Darlington connected transistor pair 3

[labee]

I have a power amplifier circuit that has a darlington connected bjt pair. The voltage drop of the transistor goes from 1.5 volts to around 1.2 volts after the circuit has been turned on and is running for a few minutes. I was wondering if there is anyway to make the bias voltage to the transistor stiffer, so that the thermal voltage drift does not result in higher base currents.

 

[IRstuff]

Why is this a problem? The Vbe drop partially compensates for the increased base resistor value at high temperatures.

TTFN

 

[labee]

The circuit is a little bit more complicated I'm afraid. It has a drive control, and a biasing scheme with high precision 0.1% resistors. Normally, the drive control signal and the biasing voltage produce the needed current in the collector of the Darlington connected transistor. The control signal values and expected output has already been calculated etc and programmed in software.

However, this new problem I am encountering is that the output for a given drive signal is not what is expected due to the thermal drift. I cannot keep changing what I expect, to be based on the different temperature variations that the transistor may experience. However, if I can make the bias voltage somehow stiffer, so that it does not change very much based on temperature, I can keep everything else according to design.

I wish I knew how to upload a schematic of the circuit, so you would see what I am talking about. A little difficult to explain in words. Sorry for any confusion

Thanks

 

[IRstuff]

From you description, your fundamental approach seems dubious. You should not and must not try to control a BJTs collector current using voltage, it's bound to fail.

You'd almost be better off with a current mirror design. Current mirrors can be designed to be quite insensitive to temperature.

Alternately, you might want to consider some sort of emitter resistor and using the voltage developed there for your current sense feedback

TTFN

 

[cbarn24050]

Hi, you seem to be relying on the transistors vbe and hfe to define the collector current from some fixed base drive. Since both these values are variable, even worse for a darlington, you cannot do it. You must have a closed feedback loop somewhere in your control system.

 

[ScottyUK]

The old-fashioned way to provide feedback to keep the bias conditions stable is to mount a small transistor within the bias circuit in direct thermal contact with the heatsinking for the output stage. There are examples of this configuration in numerous published designs for audio amps.

As cbarn says, you can't keep stable bias conditions without feedback of some form or other.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[labee]

Great tips. I was leaning towards using a current mirror myself. Will get back if I need any more help with the feedback stuff.

Thanks again