Unique problem - one channel silent under different variables

Bought a mint (looking) Cary slp 98 on eBay, connected to two reconditioned Hatachi HMA 7500 mosfet amps, work perfectly for years since reconditioning - biamping NHT 4 towers, bought new. 
When the preamp is turned on, a series of pops through the speakers, then Only the right channel plays. Switched tubes, didn't solve issue. Changed inputs from preamp to left and right amp and immediately both channels worked. Switched off, waited 5 minutes and still both channels working. Came back a few hours later, switched on again (having reversed the inputs to normal) and again same sequence: left channel silent and after switching inputs, both work and remain working until I switch off and come back when the units are cold.
I can understand that that there may be a resistor/capacitor that need to be hot before working, but why would switching input cables from left out to right input of amp and vice versa immediately fix it!
so every time I want to hear music I have to go thro this rigmarole! Why does switching cables fix the problem????
any solution will be greatly appreciated! I'm sure it will end up back at Cary but I wondered  if anyone (Almarg???) has an answer?  Does the popping on switching on, both channels pop, provide a clue?

Post removed 
I’m not sure I understand this. In this latest experiment was it the speaker cables that were switched (per Yogiboy’s suggestion), or the interconnect cables (as seems to be implied when you said "the right channel is supplying the left amp")? Or was it both that were switched?

Also, previously you had found that switching the interconnect cables after a few minutes had fixed the problem. And that leaving the preamp in standby when it was not being used also eliminated the problem. And that popping had been heard in both channels. Not sure how to reconcile all of this.

Perhaps two problems are present. Or perhaps the preamp is the only problem, but its condition has worsened to the point that both channels are now affected in a similar manner, and the preamp is intermittently injecting either DC or some transient from its right channel output that is causing the left channel amp to which it was apparently connected during this latest experiment to shut down.

It all seems very inconsistent. But in any event I would not conclude at this point that the amp is necessarily the only problem, or even that it necessarily has a problem.

Best regards,
-- Al

Confirmation that it was the amp - replaced with another amp and no muting or popping.......