Bad Transistor in Power Supply ?!?

For an unknown reason, the autolocator for my Tascam ATR-60 crapped out on me. I tracked it down, with the help of John LaGrou, to a problem the PS. The PS produces +5VDC and +14VDC. The loads are ok, and even unloaded the PS checked out, it just seems seemed like the +5 rail was crapping out under normal loading. I looked the bottom on the PCB, and sure enough, one of the transistors (2SB507) looked fried. So I pulled off both transistors (PNPs), and tested them by measuring from B to E and B to C. One of them checked out fine (a 2SA1015), and on the other one (the one that seemed fried), B to E and B to C were all showing open. When a transistor burns up, will is open, or short, or either one? I also lifted the end of the diode connecting from B to C on the transistor that fried to make sure it wasn't shorted, and it is fine. Before I order and place any components, I wanted to see if you guys can think of something else I should check.

Also, how do I test a bridge rectifier?

Hi Greg,

Yes, transistors die in all sorts of ways- short circuit, open circuit or a mixture of the two between pins.

Testing the transistor as a pair of back-to-back (NPN) or nose-to-nose diodes (PNP) can be a good go/no go test, but often a better test is to wire it as a switch with an LED and 2k2 resistor from a 9V Vsupply to the collector, with the emitter connected to 0V. Then connect a 10k resistor from Vsupply to the base. If all is well the LED will only light when you supply current to the base.

You can use the 9V battery/LED/2k2 resistor to test a bridge rectifier too- just check between each pin of the package so that you can check each individual diode for diode action- test both ways to make sure one hasn't gone s/c or o/c.

:thumb:

Mark

I keep a small box with this circuit in my toolbox at all times:



There's connections for probes/croc clips as well as a 0.1" female header to allow me to poke diode/transistor pins in- like on the Hfe testers on a DMM.

The LED can be a pair, or a single "bicolour" type. I've had this in a box for 15 years and it's never let me down!

You can test signal diodes and rectifiers too by connecting from c to e and switching the NPN/PNP switch to change the polarity of the test current. Watch out for some germanium diodes- some of these have very low Ifmax ratings.

Mark

Thanks that's awesome !!!

Also,

If you bring the c, b and e terminals out to 4mm sockets, and connect probes to c and e you've got a simple continuity tester. Always carry some form of continuity tester- they're just too handy!

Mark

Update:

Replaced a couple transistors... only one was bad but since I was in there I changed both.

All is happy in autolocator land !!!

Thanks for the useful info, Mark.

> When a transistor burns up, will is open, or short, or either one?

Failure usually starts with a junction going short.

But if large power is available into a short, then the bonding wires burn open.

Any way: if it isn't right, it's wrong.

> how do I test a bridge rectifier?

Use an ohm-meter with a 1.5V battery (analog, not digital). From + to - should read open one way, conductance the other way. The conductance will probably read near the top of the scale, no matter which Ohms scale you set the meter to. (You are actually seeing the 1.0 to 1.4V drop of two silicon junctions at various currents, not a true resisance.)

The ~ ~ terminals should read open to each other.

Or smoke-test: wire the AC terminals to a 12VAC CT transformer through a 100 ohm 0.5 watt resistor. If the resistor smokes, the rectifier is shorted (you usually know this because the original transformer smoked). If no smoke yet, wire 1K resistors from + to CT, from - to CT, and check DC voltage. Should be about 5.5VDC each side. Much less suggests an open diode.