Need to identify transistors

Thread Starter

Jack_K

Joined May 13, 2009
143
I have an old ACDC Electronics 12 volt power supply model number OEM12N9.5. It's a 12 volt 9.5 amp supply.

It has died and I suspect the power transistors are bad. Anyway, one of them is open.
They are marked JIL (best I can tell) at the very top. Below that is 54-031. Next line is /017, and then 7521 (I assume is the date code).

Any help is appreciated.
Jack
 

ebp

Joined Feb 8, 2018
2,332
It is probably a house number.

The two output transistors are almost certainly in parallel (series used occasionally, but pretty rare and very unlikely with 12 volt output). There is some chance they are a matched pair (selected for similar Vbe and gain, primarily). If they are directly in parallel they would need to be matched closely. If each has a "ballast" resistor in the emitter circuit, the need for matching is lessened. You can make fairly badly matched transistors share reasonably well by using higher resistance in the emitter, but that means more wasted power.

A transistor going open-circuit is fairly rare - usually they fail short circuit but may subsequently open if there is enough power available. From the pictures I found on the web it looks like the supply has a crowbar at the output. This is an SCR circuit designed to short-circuit the output if the voltage goes too high.

If you are sure one transistor is still good, check the base-emitter voltage with a diode check function on a DMM, just to confirm it is not a darlington (about 0.6 V for single transistor, twice that for a darlington).

There are probably some off-the-shelf transistors that will work, but you would need to replace both because of that need for sharing. The 2N3772 (or 71) was at one time a very popular device and would probably work OK with decent safe operating area margin. The 2N3055 was extremely popular, but it is tight on SOA. There is always some risk of messing up the frequency compensation by substituting faster or slower transistors.
 

Thread Starter

Jack_K

Joined May 13, 2009
143
It is probably a house number.

The two output transistors are almost certainly in parallel (series used occasionally, but pretty rare and very unlikely with 12 volt output). There is some chance they are a matched pair (selected for similar Vbe and gain, primarily). If they are directly in parallel they would need to be matched closely. If each has a "ballast" resistor in the emitter circuit, the need for matching is lessened. You can make fairly badly matched transistors share reasonably well by using higher resistance in the emitter, but that means more wasted power.

A transistor going open-circuit is fairly rare - usually they fail short circuit but may subsequently open if there is enough power available. From the pictures I found on the web it looks like the supply has a crowbar at the output. This is an SCR circuit designed to short-circuit the output if the voltage goes too high.

If you are sure one transistor is still good, check the base-emitter voltage with a diode check function on a DMM, just to confirm it is not a darlington (about 0.6 V for single transistor, twice that for a darlington).

There are probably some off-the-shelf transistors that will work, but you would need to replace both because of that need for sharing. The 2N3772 (or 71) was at one time a very popular device and would probably work OK with decent safe operating area margin. The 2N3055 was extremely popular, but it is tight on SOA. There is always some risk of messing up the frequency compensation by substituting faster or slower transistors.
Thanks. I see ballast resistors on the PCB. The transistors do not appear to be Darlingtons. I may have a couple 2N3772. I'll try one at low current to see if that's the problem.
 
Top