I have an old power supply that uses a germainum 2N297A TO-3 as a pass transistor. When I looked up the price of replacing the 2N297A I was a bit surprised at how high they are! I realize that prices go up when parts are no longer made or rare, but is there something special about germainum transistors? I looked at several types. Is it difficult or wrong to substitute silicon transistors for germaniums?

So what is it about germamium transistors- is there something special about them? I know they have a lower forward drop voltage which is useful in some cases, but I don't know how else that effects their use.

I have several 2N3055's- they seem to surpass all the parameters of the 2N297A, but it is of course silicon. Would it work well as a substitute for the 2N297A in this case (as a power supply pass transistor)?

Thanks,
Keith

A 2N3055 is the wrong polarity.

2N297A is a germanium PNP 3Amps in a TO3 can - a 2N2955 will be a bolt-in replacement but there may need to be some circuit fiddling - they may be obsolete as well though! I see that MJE2955's are available here for only $2.40 - they will still bolt onto the heatsink and you can use the old mica washer. Bit overkill for the job though!

Can you show the schematic of the regulator in the power supply ?

That 2N297A transistor is a PNP transistor, the 2N3055 is an NPN, so it will not work as a direct substitute.
It certainly can be made to work, like turning it into a "virtual PNP transistor" using the Sziklai compound pair configuration http://en.wikipedia.org/wiki/Sziklai_pair
If you are using a silicon PNP transistor in front of the 2N3055 in the compound pair, then the bias voltage will be about 0.6V versus the 0.2 - 0.3V bias of the germanium transistor. Chances are it will not cause any problem.

Good luck, Peter

Oh... I don't know what made me think the 2N3055 was a PNP... got my letter mixed up...

The power supply is part of an old USM-116 meter:

http://www.strucktower.com/usm116-60.jpg

http://www.strucktower.com/usm116-61.jpg

After replacing the filter caps I am still getting 62 volts where I should be getting 55, and I was just thinking to sub out the 2N297A to see if that was the problem. The diodes and resistors seem OK (checking in-circuit). And I don't want to have to buy one of those NE-32's if I don't need to...$$$

Thanks,
Keith

The only concern is the part of the circuit that leads down through the precision resistors
to the special low resistance fuse. Where does point Z go to and is filament current critical
in the probe rectifier tube ?

You are getting 62V instead of 55V as nowadays the mains voltage is 120V nominal (in practice usually higher than that) as opposed to 115V!

Peter

Keith,

I have on hand what my old cross references show as a direct replacement. It's an International Rectifier TR-02. It's presently buried in one of a few boxes of old solid state stuff I have on hand. Double check to be sure if it is a direct sub and if so, get back to me and I can dig it out. Make sure the basing is the correct one to drop in the location.

Pat

Peter- The voltage at my house is generally 117, but I will look into that. The other levels are closer though, like the -26V is about -26.8 or so, and the 100V is pretty close to 100V (can't remember exactly). Also, this circuit is supposed to be regulated from between 105-125 primary input.

Steve- I would think the filament current in the probe rectifier tube is important, but not enough knowlege here. The "Z" goes to one of the main resistor divider networks: http://www.strucktower.com/usm116-89.jpg (sorry, bad scan)

Pat- Googling around I can't find a pinout for your TR-02, though I can see it is a TO-3 same of my 2N297A. I'll look around some more and PM you.

Thanks,
Keith

Either a Silicon or a Germanium PNP transistor with the same or better maximum ratings will work in the power supply circuit.

Voltage regulation is usually done by a Zener diode, with
the pass transistor being used as a current amplifier. It looks like
the -26 volt base reference is generated by a neon lamp (gas regulator)
and a pass regulator tube in that design.

It could be done a lot simplier these days.

Pete

The linked schematic shows the reference for Q1 being set by R54 and CR5 (can't read the details for either). In this type of circuit the output voltage should be the zener voltage less the Vbe drop of the pass transistor.

RRM

misread schematic, my post was in error... removed comments.

Peter

The drawing I'm looking at shows the banded end of CR5 connected to ground and the arrow end connected to the base of Q1 and a resistor to the -56? volt supply. The -56 V is regulated by the neon and the tube but the voltage on Q1's base is set by the voltage across CR5, which should be constant.

RRM

You are entirely right, I scanned over the schematic too quickly.
Thank you for the correction.

Peter

If your unit is as old as I think it might be, Ge transistors of both types were the order of the day. The only reason for this one being expensive now is that it's no longer made.

You can subsitute a Si PNP with appropriate ratings. This will lower the total voltage to the series tube filaments by maybe three or four tenths of a volt.

You might want to check CR5 and R54 before you change the transistor.

Ge transistors were a pain. High leakage and thermal runaway were no fun.

RRM

So where the first schematic shows -26volts you'll actually be getting -28v (55v - 27v for the zener [http://www.alldatasheet.com/datasheet-pdf/pdf/114404/MOTOROLA/1N3030.html}) so at the emitter of the series pass transistor you will get 28 - 0.3v (for germanium) and 28 - 0.6v (for silicon). IMHO that won't matter diddley squat for some tube heaters.

This is a negative supply. Things are "upside down". If the zener is 27 volts, the reference is -27 v. A Ge transistor will give about -26.7 volts out, Si about -26.4. Either way it's down in the noise since the zener tolerance is probably more than that.

RRM

My take on the circuit shown is that it is a shunt regulator which controls the entire voltage across the +100V to the -55V. The +100 primarily set by the pair of 30K and the 22K resistors, and the 6au6 trims the -55 volts. I don't think the PNP transistor, silicon vs germanium,makes much if any real difference, simply because the Zener is a +/-5% part, which means it can be anywhere from 25.5 to to 28.5 Volts.

Adjusting R-50 should get the -55 in line.

First unplug the power transistor and check the voltage at its base. It should be minus 26,plus or minus 5%. If it's not, fix that problem first.

If the base voltage IS okay, a silicon transistor will drop the regulated voltage by a small teensy bit, like 0.35 volts. That's barely 1%, and a lot less than the tolerance on the -26 volt supply that sets this voltage level anyway.

I would plug in a silicon transistor in a second. If it's really critical to get the exact voltage, you can bring it up with three resistors and a silicon diode, or one resistor and a germanium diode.