Okay, so I was having some problems with the 20v transformer; all the expectable ones like no headroom and so on.
So I installed (with some trepidation) the 35v 5A transformer that resided on my test-bench. This had the immediate effect of pushing the speaker protection circuit straight out of it’s comfort zone and into latch-trip-off mode anytime it got pushed past a 6vRMS waveform. So Nurse, board surgery!
First job first, get rid of that awful latch. Option 1 would be to just link-out the output relay, but I kinda like the de-thump delay function, especially with that 6N tube warming-up. So, what controls it? There is a SIL UPC1237 right next door to the relay. The data sheet shows that the latch function is deactivated by an earth on pin three, so link-out C18.
That just left the laughably low level of the trigger. The cct is suppossed to be spec’ed for 68W/ch on a normal heatsink. Mine’s a bit bigger… So, I want to push this thing.
As a start on the trigger-level, with the two 56kRs monitoring the speaker channels summing into a 4u7F electrolytic, I tried a 25% reduction by discharging the 4u7 through a 100kR, but my eyes are getting so bad I got a 180kR instead. So I paralleled that up with another, but realised that it wasn’t transients setting it off, it was sustained bass power. That frequency-dependent characteristic is set by the 4u7.
Therefore, ultimate mod; stick a 10uF across that.
Job done. Now I can get 80w/ch out of it. Seriously, even the smoothing cans get warm. Never seen that before…