[Ron Ramirez]

I went back to the bench and ran several more tests, took more voltage readings, and pulled out the signal tracer.

I had set aside the original TR-110B output transistors as they seemed to be OK after testing. This was before I cleaned the switches of my "new" transistor tester with DeoxIT. Well, I tested them again tonight - yes, they are good! So, back in they went. This raised the previous voltage readings of 19.6 and 19.7 volts to above 22 volts.

For some reason, I felt suspicious of the GE-3 driver I had installed in place of the bad original TR35144, so I installed a NOS 40626 Ge transistor (same specs as NTE121/GE-3/GE-16/TR35144). No change, right channel was still lower than the left.

I measured 29.99 volts at B+ (the common point where my most recent revised schematic states To Power Supply).

I replaced C14 and C26 with the only things I had - a 4.7 uF, 450V cap for C14 and a 100 uF, 100V cap for C26. No change.

Then I hooked up the signal tracer. This was quite revealing.

The sound in the right channel just before the volume control was louder than that of the left. This was because I had the balance control set to favor the right channel as I listened to a local FM station while doing my tests.

Now, bringing the tracer down through the audio output circuitry, I compared the left channel to the right by using the probe on the base of Q2, then the base of Q1, then the collector of Q2, then the collector of Q1.

I think I have found the problem.

The sound was louder at the collector of Q1 (left channel) than at the collector of Q2 (right channel).

The sound was equal, more or less, at the base of Q2 and at the base of Q1.

When I set this chassis on the bench, I found that someone had replaced Q2 with something other than a 2N2924. I did not have a 2N2924 on hand. So...I stuck in a transistor equivalent to an NTE123.

I think this was my big mistake.

Oh, I restored audio in the right channel, all right, but not to full specs. And I had been going around in circles ever since.

2N2924 transistors are 59 cents at Mouser.

So...since I need to order some electrolytics anyway, I'll also order a genuine 2N2924. We shall then see how the right channel performs. I will go ahead and replace all power supply and audio output electrolytics while I'm at it.

Edit: I now see that 2N2924 (NTE85) has a much higher hFE (min. 100) than the NTE123 (hFE min. 30). So it now makes sense to me that using an NTE123 equivalent was a poor choice. Kind of like trying to substitute a 27 for a 6J5.