Balancing power transistor output

Thread Starter

jrdoner

Joined Feb 21, 2012
14
I have a power supply arrangement using two D1088 (6A. PNP Darlingtons) the bases of which are controlled by an LM317. Thus the two transistors are carrying the current, not the LM317. It works very well, giving tight regulation over widely varying load. But here's the rub: one of these two transistors runs about twice as hot as the other, which I suppose is due to normal manufacturing differences from one to the next. If I were trying to balance the two to carry about the same current, would I put a resistor in series with the base of the hotter transistor? That should restrict base current and cut the heating of that transistor. However, I don't know if there is any downside to such an arrangement.

Any thoughts would be appreciated.
 

dl324

Joined Mar 30, 2015
16,918
Do you have a schematic? Technically the LM317 doesn't drive the base of your pass transistor. The current going through the regulator goes through a resistor that turns on the PNP.

I suspect you have the two darlingtons wired in parallel. If this is the case, that isn't recommended because Vbe variations will cause current hogging in one of them.
 
Last edited:

Hypatia's Protege

Joined Mar 1, 2015
3,228
I have a power supply arrangement using two D1088 (6A. PNP Darlingtons) the bases of which are controlled by an LM317. Thus the two transistors are carrying the current, not the LM317. It works very well, giving tight regulation over widely varying load. But here's the rub: one of these two transistors runs about twice as hot as the other, which I suppose is due to normal manufacturing differences from one to the next. If I were trying to balance the two to carry about the same current, would I put a resistor in series with the base of the hotter transistor? That should restrict base current and cut the heating of that transistor. However, I don't know if there is any downside to such an arrangement.

Any thoughts would be appreciated.
Without a schematic the best advice I can offer is that you consider use of a low-value ballast (cip 'degeneration') resistor in series with the emitter of each pass transistor...

Best regards
HP
 
Top