What is the fifth transistor likely for?

Temperature compensation for the bias? (This seems unlikely, as the devices would be silicon. But who knows?)

Might be nothing at all to do with the output stage. PSU regulator transistors are often fixed to the same heatsink as the outputs.

Arfa

The 2SB1020 (PNP) and 2SD1415 (NPN) are complementary pairs of power transistors... absmax around 100 volts 7 amps. Presumably the "usual four" that do the current amplification.

The 2SB1370 is a PNP in the same TO-220FP package. It seems to be spec'ed out as a driver.

I'd lean towards agreeing with the speculation that this is being used to regulate the output-stage bias voltages or currents somehow - to provide thermal tracking. This is more commonly done with one or more diodes in the bias string (one set per channel) but there may have been some tricky way of doing it with a single thermally-coupled transistor on the heatsink which was simpler and/or cheaper to implement.

Maybe (and this is sheer speculation in the fact of an acute lack of schematic) this one transistor controls two separate idle/bias strings via a current-mirror arrangement of some sort?

Or, maybe it's just a "Whoops, too hot!" emergency shutoff switch circuit, to keep the player from going into thermal runaway if the owner tries to bend metal and break walls by turning the poor beast up to 11?

Temperature measurement. Used to keep the bias where it belongs.

Not uncommon, and a B-E junction has a matching tempco, which a standard diode does not. RCA used to make (when they made transistors) a special "diode" which was a B-E junction in a two-wire package, for just that purpose.

After saying all that, though, my bet is that the fifth device has something to do with running the CD drive motor.

Isaac

Thanks all. Very interesting ideas. Maybe it wasn't as cheap as I thought it was. But I had no use for it. I just like to look at things and fix them when I can.

Doubtful. As Arfa mentioned, probably a regulator (voltage, not temp). Quite common to mount on the same heatsink as the output transistors.

Mark Z.

There is an dramatic element of keeping us in suspense here :-)

Could you give a clue to the make / model for some of the clever peeps to find a schematic on this?

Ha ! Thanks Mark. I was beginning to think that no one had understood exactly what I said ... :-)

Arfa

I knew what you said. There shouldn't be a need to compensate for temps in something that doesn't put out a lot of wattage in the first place.

Not so -- not for germanium transistors, anyway, which were very temperature-sensitive. Look at the schematic for any early transistor radio -- there's a diode or thermistor in the output stage.

I was the first to suggest temperature compensation -- but also pointed out it was unlikely.

Grow up Bill. Poster said circa 1998. How many combos from then had Ge outputs?

Many modern amps still have temperature compensation for the output stage(s). However, it's very unusual to see it made from any transistor package larger than a TO92 in a typical 10 to 60 watt hifi rig. These will often be clamped to the heatsink on their 'flat' side, with a sizeable dollop of white heat goop. They may also be encountered pushed into a hole in the heatsink, or sometimes just pushed or glued against it. Occasionally, they can be found clamped between the heatsink, and the PCB. You will also see the same mounting tricks used with glass (1N4148) silicon diodes, and sometimes with them just sticking up from the board and close to - but not actually touching - the heatsink. Very occasionally, you will find a tiny bead thermistor doing the job.

I have only seen flatpack transistors being used for this purpose where we're talking big power levels, such as in PA amps.

However, repairing this Chinese crap every day of my working life, I can say that it is extremely common to find one or more flatpack devices, additional to those being used for the output stage, mounted on the same heatsink. I guess this is to cut down on cost. Invariably, these 'extra' devices, are simple linear voltage regulator pass transistors, and that is what I would be fairly sure that the OP's fifth transistor would be.

Arfa

None. But that isn't what you said.

say

additional

There's also the fact that there's only one extra transistor. Wouldn't a two-channel output stage require two?

If the transistor was being used for bias compensation, then yes. Most likely one per channel. In the case of voltage regulator transistors, the rails are shared between both amps in most common hifi units, so only one transistor required per regulated rail, no matter how many channels are in there.

Arfa

Sorry. I stopped reading the thread. By the time I first posted, I had destroyed it. Don't remember the brand, much less the model.