Firstly, apologies for the long delay since the last post, but life
had other demands on my time, and I could only to get to work
on the 2270 the last few days.
I think I finally have it licked. Both channels are working fine now,
and I can set the bias correctly now. A longish report -
1) The resistors at R769/R770 should be 680 ohms as per the
service manual, but both channels had 380 ohms at these locations.
Replaced them with 680ohms, 1/4W.
2) On both channels, diff pairs H751/H752 (2SA640) replaced by
a pair of 2SA992, matched Hfe of 500 each. The pinout was the
same, so it was an easy replacement.
H753 (2SC945) was replaced on both channels with a BC546
which has a different pinout, but it fits nicely with no bending of pins etc.
3) Current protection circuit - H754/H755 (2SC735) and H766/H767 (2SA562) were replaced with 2N3904, and 2N3906 respectively.
I found that a couple of the old ones were bad (tested with
a AVR transistor tester after removal). Also, one of them was
replaced in a previous repair, so made sense to replace them all, on
both channels.
4) On both channels, H756/H757 (2SA484/2SC484) had been changed
during the previous repair to SK100/SL100s. These
were replaced with 2N3440 at H757 and 2N5416 at H756.
There really are very few TO-39 transistors available, and even
these were not easy to locate locally.
5) The output transistors were unmarked and rusty and suspect.
I changed them to MJ21195G/MJ21196G
Now the left channel worked fine, and the bias could be set correctly.
But the right didn't want to cooperate, and the bias was
still not working.
The only devices left to replace were the TO-66 drivers -
H758 (2SC680) and H759 (2SA566).
I used TO-220 drivers MJE15032/MJE15033. These fit well into the
TO-66 layout and the heat sink can be reused too.
It would be easier to fit TO-220s into the old heat sink,
if the holes for the pins in the old heatsink are drilled out to
a slightly larger size.
There were a couple of zener diodes H764/H765 which are
WZ-071 (7.1V, 0.5W). Considering I had almost rebuilt the whole
board, I replaced them with the nearest match available which
is 7.5V, 0.5W (1N5236).
Thankfully, the right channel worked after all this.
The bias pot is highly sensitive, though. A slight touch and there is
a big change in the bias. Is that normal?
I think it would be worth replacing the original pots with Bourns
multi-turn trimmers to have better control.
Any suggestions for the suitable part number? TIA!
Catrafter noticed that the amp had a pair of 10 ohms, 10W resistors.
Apparently, this is an old repairman's technique used for
debugging - during normal operation, there is little
voltage drop across these resistors (I measured 4 Volts
across them during normal use),
but in case of any short or malfunction, there is a big drop
across these resistors, which hopefully will prevent further damage to
other parts on the amp.
To sum up, it was a lot of work getting this to work properly
and rather frustrating at times, but it was fun too, and I learnt
a great deal in the process. There is a wealth of information in
the old posts on this forum.
Thanks to everyone here, and specifically to ecluser and Catrafter
for their help.