Yamaha M85 (Part 2)

Spoiler alert: I sold the amp. But here was the progress that I had made up until that point.

Gunk Patrol

First, I replaced a handful of parts that were affected by the Sony bond glue:

• Replaced the large 1000uF filter capacitors on the main amplifier board (not the behemoth AC smoothing caps); measured within nominal 20% tolerance
• Replaced the smaller caps below the larger ones; measured within 20% tolerance
• Replaced the two zener diodes by these caps, one was buried in glue; both measured identically

Diagnosis

I then replaced the parts that were measuring differently across channels:

• Replaced a set of BJTs that seemed to be measuring differently across terminals; both measured identically out-of-circuit (OOC)
• Replaced two 2.2M ohm resistors that measured very differently in circuit; both measured identically OOC

Seeing identical OOC measurements at this point implies that none of these components were causing the issues that were present in the amp (Part 1).

Next: I did measure different voltages¹ across the 2SK389 dual JFET used at the input of the poweramp board. It’s hard to find replacements for these because they’ve since become obselete in modern circuit designs, and folks are selling NOS equivalents on eBay at a markup. I do have two 2SK’s in my Yamaha P2150C, but I wanted to keep those in the amp because I was planning on repairing that one too (fingers crossed those aren’t broken either…).

I ended up simply swapping the current 2SK’s across channels. If these chips were the culprit to this amp’s problems, then I’d see the problem go from the right channel to the left.

Testing

After all of these changes, the same problems persisted: Amp stays in protection mode, and the bias of the right channel stays low (3.8mV) and is unrepsonsive to change in trimpot resistance, as well as changing to “Auto Class A” mode.

Meanwhile, the left channel performed as expected: The bias responds to trimpot and was set to 10mV, and would go up to 83.9mV in Auto Class A.

This means:

• The 2SK389’s are working fine!
• The culprit is either:
  • (1) downstream on the main board, or
  • (2) possibly on the long auxiliary “daughterboards.” I hope it’s the main board, because it looks like a pain to access the daughterboards – people just desolder the auxiliary power transistors altogether to remove the boards, then and resolder them after repair.

Apart from the parts that I’ve already replaced, I’ve only ordered replacement diodes and transistors for the first “upstream” half of the board. At this point, I’ll just start replacing these parts on the right side as a shotgun approach; others who have worked on the M85 have said that replacing parts that seem to measure fine often fixes things.

Finale

Often times, journeys end in different ways than we initially plan. I guess that’s what makes them exciting.

I found out through a Craigslist listing that someone is looking for M85’s. I reached out to the buyer out of curiosity and he offered to buy this amp, even in its non-working condition. Seeing that both of us would have made good profit from the sale, I decided it was time to say farewell. I sold it to him, he took it to his amp tech, and he got fixed! Last I heard, this M85 has joined his stack of other M80/M85’s and is now driving a circle of Heresy’s in his listening room. Sounds like a loud place.

What’s next? Probably onto the P2150C :)

Footnotes