[Mondial]

Yes, that's pretty much it. You want to find a replacement with equal or higher voltage and current specs, and comparable physical package. The other factor is the cutoff freq rating Ft, which should be in the 2 to 10 Mhz range. 

If you can't find any specs, then look for transistors with a Vce rating comfortably higher than the total DC power supply voltage. Ic current ratings are typically 8 to 15 A for small power transistors and should be suitable. If the amp is higher power, then parallel connected outputs are usually used.

Another thing is matching. Japanese output transistors specified for audio use are typically tightly matched into groups for beta and Vbe. This becomes important when you have sets of transistors in parallel for the output stage.

For the 170, I would be looking for transistors with Vce ratings of 60-80 v, and Ic around 10-15 A. What sort of package are they? Metal TO-3 or plastic case?

[Ron Ramirez]

Plastic, TO-220. (This is a low end, flea powered receiver.) That's why I hadn't popped them out for testing before now - it's a bit of bother to remove a TO-220, compared to a TO-3.

And thanks! That's great info.

[Ron Ramirez]

No dice, Mondial...I pulled Q15 and Q17, tested both with my transistor tester and tested both junctions on each transistor with the diode function of my multimeter. Both test OK.

I though I had found something when I tested R95 and it only measured 0.2 ohm in circuit, while R96 (right channel) measured 0.5 ohm in circuit. But I pulled R95 and measured it again out of circuit...0.5 ohm. ????? Yes, I checked...infinity across the points where R95 connects with R95 off the board.

I frankly do not know where to go from here.

[Mondial]

Hmmm. When you checked with the scope, what did the distortion look like with the speaker connected? Was it on only the top or bottom of the waveform, or did it distort equally on both?

Feedback amps are really hard to troubleshoot because a problem anywhere in the circuit will affect the signal from input stage to output. Sometimes the only way to fix something like this is to check each part individually, rather than try to trace the problem to an individual stage while operating.

Another complication is that it appears that the voltage measurements on the schematic were made with a sine wave signal applied and load connected. There is no way you could have .4 V across a .47 ohm emitter resistor from idling bias alone. That would be .4/.47 or 850 mA ! At that point the output transistors would be dissipating 35 Watts and cooking. Rather, under no signal conditions you probably should see 10 to 20 mV across R95 and R97.

I guess I would at this point remove the remaining transistors and test them. While they are out of the circuit you could check most of the remaining passive components without removing them, since most will be isolated with the semiconductors removed.

[morzh]

Ron

Have you set the initial idling current by VR7 yet?


Also,

It would be helpful if you gave the pics of your scoping the signal. When you see the distortion you can guess as what's causing it, otherwise we are kinda walkiing in the dark.

[Ron Ramirez]

Mike (morzh)

No, I had not set the idle current as of yet. I decided to follow Mondial's advice.

All transistors are OK. All of the associated resistors, caps, and diodes appear to be OK.

So I put some DeoxIT on the power-speaker switch, reinstalled the transistors, and tried again with headphones only first.

It sounded fine.

Then I connected the speakers, only to the "remote" terminals instead of the "main" terminals.

I turned the thing on and to the "Remote" speaker position.

It sounded fine!

I turned it off, and changed the speakers over to the "Main" speaker terminals.

Turned it on again, only to the "Main" speaker position.

It sounded fine for less than a minute - then distortion returned first to the right channel, then to both.

I then turned it off and moved the speaker wires back to the "Remote" terminals.

Turned it on again....distortion.

So I think I have found the problem - the speaker selector switch.

Sometime this weekend, I think I'll run a couple wires direct to the main speaker terminals, bypassing the switch and testing my theory.

[morzh]

I was wondering about the idle because all your voltages look just fine - fine for an amp whose idle has not yet been set.
If you scope the output, you are likely to see the crossover, judging by one output transistor not conducting.

However the crossover never sounds "heavily" distorted.

Anyway, you seem to have found the poblem, the small things could be adjusted later.

[Ron Ramirez]

It's official - the problem must be the power-speaker switch.

I disconnected the wires from terminals H1 and H2 that go from the receiver outputs to the speaker switch. I also disconnected the wires running to the left and right "main" speaker terminals.

In their place, I ran two wires direct from H1 and H2 to the "main" speaker terminals.

I turned the unit on and...no distortion.

I sat and listened to it for five or ten minutes, and it did well the entire time.

So now I can try and pull the switch for a deep cleaning, or just leave the wires in place. Of course, I will try to remove the switch as I will not be satisfied with a rigged-up solution. I would prefer to be able to switch between main or remote speakers or headphones as per original.

[morzh]

Still, after you are done, set the idle as at least in one channel, judging by the votages, it is zero. Depending what you listen to, you may or may not notice it, but when listening to a pure sinewave sound it is really noticeable.

[Mondial]

Yes, I agree with Morzh. From the readings you gave, there appears to be little or no bias on the output stage. Place your DVM across the .47 ohm emitter resistor and adjust the bias pot VR7 and VR8 so there is 10 mV of DC measured across each resistor with no audio signal. If you have the service info, set the bias as they specify.

This will provide a constant minimum DC current through the output transistors, keeping them on as they switch polarity of the output signal and minimize the class B crossover distortion.

[morzh]

Or another way, I simply would connect a scope and, while feeding the amp a sinewave, bring the idle while the crossover "step" disappeared.
If you feel the idle is too much, then bring it down a notch, as long as you cannot hear the step or it is barely noticeable on the scope.

There is a nice arrangement on an amp I built long time ago for myself with self-limiting of idle using feedback, I could set it as large as 300mA and there would be no thermal runaway.
An old and less efficient way to control the runaway is a thermistor in parallel with the trimpot that is mounted on heatsinks.

[Ron Ramirez]

All right,

I followed the Fisher recommended steps of setting the symmetry and idle (I do have the service manual).

The idle will not adjust in the left channel, although it did in the right.

And then the distortion returned...and now the idle will not adjust in the right channel either but remains at 0 volts as with the left channel.

So now I'm back to square one - distortion in both channels.

To say that I am very angry right now would be an understatement. I need this before bedtime?

I'm frankly tired of wasting my time on this D**n thing. Back on the shelf it goes...maybe for good.

[morzh]

Something tells me you have other problems. Power is one suspect.

[Mondial]

Ron, are you sure the distortion is in the power amp stage? Since you now have distortion on both channels, perhaps it originates in the preamp or control amp circuitry?

You might check the DC supply voltage to the preamp stages. If the regulated DC supply is dropping it would cause distortion on both channels. Check for constant 22 VDC on the emitter of Q20, the regulator transistor.

[Ron Ramirez]

Final thoughts before I put this back on the shelf for awhile:

I did turn it back on a little while ago and made a few voltage measurements.

At first it sounded OK, and then distortion slowly crept into the left channel. After a few minutes, distortion in the right channel would appear, disappear, and reappear.

Voltage at the emitter of Q20 was 22.5V and held steady.
Voltage at the emitter of Q19 also held steady, at 15.9V.

Now here's where it gets interesting. I measured the voltage across C55. When the right channel was clear, the voltage was 44.1V. But when distortion appeared in the right channel, it increased to 44.7V. The service data calls for 42V.

All rectifier diodes were previously replaced (by me).