Audio Research CL-30 Bias Drift

A client brought a CL-30 in for a re-tube because the amp blew a fuse when a tube shorted. While doing my usual voltage checks without tubes installed, I noticed that each half of the amplifier has a tube socket that has a tremendous bias drift that slowly drops from -50 volts to -19 volts. The other two sockets remain at around -49.
Not having studied the amplifier schematic, does this sound like a common problem associated with these amps? This would certainly explain why it cooked a 6550. Does anyone have any suggestions in sourcing the problem? I build and repair vacuum tube guitar amps, so this sort of hybrid circuitry isn't what I see very often.
Thank you in advance.
Btw, I obviously don't know where Ampman lives. If he is located in the Philly/New York/Delaware area, I have a great ARC authorized tech who makes house calls and is very reasonable. He works closely with Chris Ossana back at the factory.

If Ampman is located on the West Coast, Gary (Hifigeek1) is an ARC authorized tech.

If it's the heartland, there's always good old ARC.

Just trying to help.
My name is Billy Yates and I live in Venice Ca.
My client found me through several registries regarding amplifier repair. I fixed the amplifier, and is now back in service. Apparently, there are very few resources in Los Angeles to repair this amp according to ARC standards.
The amp sat for 2 years before it came to me. I can only assume he didn't have very much luck. Fortunately, I was able to get it running again.
Thank you for the comments.
Billy Yates
That's because the bias on 2 of the 4 tubes use a servo. The best way to check that is to wait a half hour while monitoring the tubes that are part of the servo. It has a slow time constant so you will have to wait. If the servo tubes don't pull in it will require service. BTW I am an authorized factory repair facility in L.A.