Why does my system now have shrill top end.

I completely agree with Todd. I find myself befuddled by the persistent recommendations that the least likely culprit, the speaker cables, be given first attention. It's free and harmless to connect a DMM to the amp outputs and determine if the bias is off. 

I also agree, Todd, that the cables are not a particularly likely culprit. However, as Uberwaltz said just above, and as I said earlier, it is easy enough to perform a test that would rule the possibility in or out. And as I also said earlier there are very strong reasons to suspect that the OCOS cables have very high capacitance, and ...

Almarg 11-27-2017
...a possibility [note that I said "possibility," not "probability"] that occurs to me is that its [the cable’s] capacitance might be causing an ultrasonic oscillation in your [the OP’s] amplifier. And perhaps the condition is sufficiently marginal that minor aging effects or perhaps even a change in AC line voltage have put it over the edge.

If so, the volume dependency you [the OP] mentioned might result from the speakers being stressed to a greater degree when an oscillation and a high volume audio signal are both present.

So to be sure it’s clear, the **possibility** I suggested was not that the cables might be the culprit in themselves, but rather that the effects of their presumably very high capacitance on the amplifier **might** be. And, once again, it’s a possibility that is very easy to check.

Also, a reason that this possibility has been given what Kosst characterizes as "first attention" is that we certainly want to be sure that it has been conclusively ruled out before other findings may lead the OP to incur the trouble, expense, and risk of sending his amp out for a repair that may prove to be unnecessary.

So while Kosst finds himself "befuddled by the persistent recommendations that the least likely culprit, the speaker cables, be given first attention," I find myself befuddled by the objections that have been raised against that attention.

Regards,
-- Al

Absolutely, I completely agree. I posted before I had completed my thought..

Cables certainly are possible, but the evidence presented points to the electrical system as a distinct possibility as well. It would at least be worth shutting off the branch circuit at the panel to completely rule effects from the kitchen circuit out. Also, double checking the connections in the panel (after shutting off the power) would be warranted as well. I just wanted to put those other possibilities out there as well, to avoid needlessly shipping an amp around the country.

Kosst also offers a reasonably easy test as well.. lots of good ideas for more data points.

If Kosst reference to DMM means digital multi meter it is possible the op does not have a halfway decent one or even know exactly what he should be testing/looking for. If the latter is the case I do hope he will ask for help right here because that is an extremely valid test as well.

Kosst_Amojan 11-30-2017
It’s free and harmless to connect a DMM to the amp outputs and determine if the bias is off.

Uberwaltz 11-30-2017
If Kosst reference to DMM means digital multi meter it is possible the op does not have a halfway decent one or even know exactly what he should be testing/looking for.

I could use some clarification on that myself, Uberwaltz. I’m not sure how a multimeter measurement at the output terminals of an amp can determine if the bias is too low (or too high), the main possibility Kosst had cited being low bias. DC offset can be checked for in that manner, of course, but perhaps Kosst can clarify how such a measurement can determine if bias is too low.

Best regards,
-- Al