First of all very nice description of the problems in your web page, very useful for others to understand what is happening. I don't have a good theory but I can help explaining the HV loop.
What would be the best way to remove the HV feedback loop from the circuit (as
a test)?
Feedback loop is needed for proper base bias of Q714.
I don't quite understand how U740 (LM741 OpAmp) works here. It's
configured as a non-inverting amplifier with a 1.83 amplification factor. With
Amplification factor is (100k+12k)/12k = 9.33 based on the schematic on your web page. For all practical purposes you can assume U740 is trying to drive pin 3 to close to zero volt because of this large gain from pin3 to output. At nominal output voltage of -0.5V, pin 3 voltage would be at -0.05V. Voltage coming from R730 and -5500V should balance each other to ~ zero volts by the ratio of R734A and R734B.
the OpAmp removed, the circuit does NOT oscillate
Without opamp there is no base drive to start the oscillations.
and the voltage at pin 3 is between 11 and 15V, dependent on R736.
When the circuit is not oscillating there is no -5500V to pull R734B to -5500V. R736 provides +15V to +10.99 depending on its position (ignore the loading of R734A. Pin3 is connected to voltage divider of R734A and R734B, without -5500V pin 3 voltage range you see looks reasonable.
======
Possible root causes:
- bad/heating transformer: I think you eliminated it by heating/cooling the transformer.
- It is quite possible that HV supply is working harder than expected because of some fault/load in high voltage side. Your DC restorer could be the cause, may be the tube is drawing too much current. You can try turning down the -5500V supply to see if it gives you longer time to work on the HV supply. Once you fix the DC restorer you can correct the voltages.
Just to make sure: Are +120V UNREG and +30V UNREG OK (I am using 606 schematic, not 606A in case names are different)? Is there a 606A service manual?
== DC restorer
In your web page WF4 shows CR769 cathode doesn't go below 65V. It is not because CR769 doesn't do its job but voltage never goes below 10V so that it can conduct. This is consistent with your findings that changing CR769 doesn't do anything. It is a clamp, it will only work when voltage goes below 10V.
CR776 could be leaky, when cathode of CR769 (same as anode of CR776) starts going down it may leak current, then waveform doesn't have a chance to go below 10V. Since you already probed several voltages, here is one way to find out where the extra current is going: Look at cathode of CR776. It should go slightly up when it is clamping and stay steady when its anode (WF3) is falling from clamp voltage. If it is following the falling edge of the waveform it is leaking.
Next diodes CR782 and CR780 could be leaky. This is a high voltage area and best is to disconnect the diodes and test them outside. They should be able to work with a reverse bias of 100V+ so while testing check with a large reverse bias for leakage.
Another point to keep in mind is only 27pF is driving these nodes, your scope probe capacitance is a significant load.
Ozan
