Mark Levinson protection circuit - Need help

Yes for me for sure it would have been impossible. I don't know how you do it.
I measured again the LM394 with clips, without touching it.
Left:
Vbe 674 reverse 814
Right:
Vbe 679 reverse 797

Still acceptable?
My meter is really cheap so I'm not sure how reliable are the measurements.
 
Yes that's what I was thinking, they replaced all the transistors (except the LM394 that was however pulled out for testing I suspect since it was installed slightly tilted), plus most 1n914 diodes, but couldn't solve the problem.
Yes it's a pity to wait for the resistor but with mouser I should get it in a couple of days, at least I will put the original Vishay Dale resistor back in place. Plus I had to order the caps for restoring the ML28.
 
So I mounted the resistor and now all voltages are correct. I can set the offset of the inverting input to 0V with the trimpot and the LS output behaves like the other channel, going to under 10mV after a couple of seconds from switch on.
It seems that the amp is fully functional for the first time after probably at least 10 years!
Can't spot any big differences in sound though, but it isn't calibrated yet and still open with wires connecting the power rails etc..
I have the impression that the left channel has a slightly lower volume, but I'm not sure yet. Probably the bias are set very different between the two channels too..
 
Oh ok, didn't know you could do that. So far, however, after adjusting speaker positioning I'm not noticing significative imbalance between channels. The amp seems to be working and sounding great, thanks for your help! Thanks to you I was able to bring the ML back to its former glory!
 
This is completely and absolutely harmless, altough from a pure technical point I wonder how this can happen because as you can see in my simmed model that the input is AC coupled.
Could you measure what DC voltage your preamp produces in the XLR plug ?
Pin 1 is gnd.

Hans