Oscillation Explosion!

[KeithTr]

Did you ever check those capacitors? Play a square wave into a dummy load with 100 W DBT just to be safe and tap away while observing your scope.

Update #3

Well, I took all suggestions to heart and did the following in order: changed out the 10pf cap and made it 20 on c363. Checked all other miller caps and added 10pf. Could not reproduce oscillation. But when I started tapping, I got no change in the scope tapping anywhere in the amplifier section. But I got some jittering in the scope when I tapped in the other side of the board. I looked very carefully and found some suspect solder joints in the areas of the inputs and speaker terminals. Reflowed all those joints and tested. This time got just a tad when I tapped in a very specific area. Reflowed all joints there and went back and very carefully examined the board for cracks etc. I had previously found and repaired a burnt trace when the left channel failed before I undertook an attempt to fix this sucker (and it has been a time sucker for sure). I think I finally got it. Not seeing any signs of jitter or oscillation in testing and square waves look normal. Will just have to let it cook now for several hours as a final test. Thank you to all who chimed in. This is why I hang out here and how I learn. Even an old dog can pick up a few new tricks!

 

[gadget73]

Have been bitten by bad solder more than once. Problems that appear completely random are such fun to chase down aren't they?

 

[KeithTr]

Have been bitten by bad solder more than once. Problems that appear completely random are such fun to chase down aren't they?

This one was a deliberate learning experience that turned out to be a more difficult challenge than first expected. Someone gave me one of these that was going to be trashed. Although hardly TOTL, I like Technics stuff and also the form factor - quite slim for an integrated amp. I found that the main circuit board, which had many very small traces, was cracked in three places and not worth the attempt. So I figured I had a parts unit. Then I went and bought the same unit cheap with a non-working channel off the "e" site (I think it was around $60) and I figured it would be fun to try to fix.

The non-working channel had 2 blown outputs, several blown driver transistors, and, as I figured out after many many hours of checking components and continuity, a trace that had become discontinuous - probably when the outputs blew and ran too much current through it. That was a bear to find and took weeks of on and off looking and testing. Once I replaced that trace with a jumper, the channel worked.

Then, in testing, I had oscillation, which I didn't recognize right away, and that blew up the RC (zobel) components (a combined resistor and capacitor in series near the outputs) but apparently did not other damage (some nice magic smoke, though). After a bunch of consulting with a friend here and AI, I determined that some of the replacement transistors in the amp section were creating oscillation as the modern parts were much faster (higher frequency) and added capacitance to the "miller capacitors", which caused me to read up on what they actually do and how they work.

Then, after I fixed that, I started tapping on the board while hooked up to the scope and got more jitter on the signal - and started checking solder joints (which I had done before, but without seeing obvious failures). But I reflowed all joints in the area where I had been tapping and now both the oscillation and the jitter seem to be gone. Fingers are still crossed, but I think I got it.

 

[KeithTr]

Cleaned up, on for several hours playing, no excessive heat, no distortion, nothing unusual. Another save from the dump. Quick demo attached.

 

[Sansuiman]

Nice save. I really like the aesthetic of that era Technics gear in silver. They were kind of aping Yamaha with those rectangular knobs.