I went through a large number of HP8568B, HP8566B, spectrum Analyzers and found few common issues. I'm posting in case someone is having these issues.
1) Cables! Bus cable and its connectors are now 30+ years old. That means pins and prongs that rely on bends to apply pressures are now quite tired. IF YOU ARE GETTING INSTR CHECK I, try unscrewing BUS cable at Display section side. Pull lightly and see if there are any pressure. If they pull out very easily, CAREFULLY create an appropriate bend in cable side of finger stocks. Make sure thumb screws are tightened.
2) Causes of YTO unlock can be anywhere. YTO itself may be perfectly fine. YTO is the last thing in the chain. Thus, YTO unlock problem can be anywhere. One way to take reference oscillator out of the equation is to use external reference.
3) HP used good quality capacitors. There is no need to blindly recap RF sections. Even those big cans in power supplies are likely still good. Out of 12 RF sections I went through recently, number of capacitors I replaced = ZERO. HP8566B and DISPLAY sections are another matter. YTO driver boards and Z axis driver boards tended to have bad capacitors more than 8566s.
4) A to B conversions were pretty common? I have a batch of 14 units. About half of them were A upgraded to B. I understand HP offered an upgrade kit for several thousand dollars. That replaced in house CPU to 68000. I guess these were popular back then. A has two batteries, B has one. There is a vacant space around the reference oscillator. (silver can supported by rubber bushings). If you look into that area, on PCB, there are pads labeled "BATTERY". If battery exists there, it's an A, or it's a botched conversion. B doesn't have it there.
5) Handles are often CRACKED. Inspect first before lifting! Handle straps may have lose screws and each side is supported by only ONE screw. Be careful.
6) Have a Chiropractors and Massage Therapist appointment ready. You'll need them.
That's about all I can think of at the moment.