Some thoughts that have occurred to me after reading through this thread:
1) Since these amps are huge high powered monsters which draw lots of current and require a dedicated 20 amp line for each monoblock, I’m guessing that the electrician probably installed the two lines on opposite AC legs. Jim (Jea48), some years ago you had provided us with a link to a paper by Henry Ott and Bill Whitlock which explained that powering interconnected equipment from opposite legs is conducive to ground loop issues. And of course the grounds of the two amps are interconnected via the preamp.
If the additional experiments the OP will be doing don’t lead to a solution, based on your expertise on such matters do you think it would be reasonable to have the electrician put both amps on the same leg, per the suggestion George made earlier? Or is that likely to unbalance the legs to an undesirable degree?
2) An alternative approach that would certainly resolve a ground loop issue involving the amps would be to insert a single-channel Jensen transformer between the preamp and each of the amps. Based on numerous experiences with Jensen transformers that have been reported here and elsewhere I would expect the sonic effects of doing that to be very minimal, although perhaps not imperceptible. The two transformers would cost something like $250 each. A very short additional XLR cable would also be required for each channel.
3) I’m also a bit suspicious of the preamp. I infer from what the OP has said that the hum does not occur when the Mark Levinson 33H amps are used with the same preamp. However based on the last paragraph of
this page of Stereophile’s review of the 33H it sounds like the input stage of that amp may very well ignore the XLR ground pin (pin 1), while the Krell amps (like many and perhaps most other amps which provide XLR inputs) may not. So the response of the two amps to an issue in the preamp which would inject hum via that ground path could very well be different. Also, the preamp in question apparently uses a 6H30P tube and either a 6L6 or a 6550 (depending on when the preamp was manufactured) in a voltage regulator circuit in its power supply, which supplies B+ to its output stage. And it appears that the filaments of those tubes are powered with AC, which adds to my suspicion. So if at all possible I would suggest that the OP try a different preamp, as an experiment.
4)
I installed a "cheater plug" between the pre and wall outlet, I hear less hum than before, about 70% less. I cannot hear hum from my listening position (10ft) away.
Perhaps multiple issues are present, namely some combination of a ground loop involving the preamp (when the cheater plug is not used); the use of opposite AC legs to power the amps; and possibly also a problem in the preamp, perhaps involving excessive filament-to-cathode leakage in one of the tubes in the voltage regulator circuit I referred to.
Regards,
-- Al