Building new sound system

That's a ground loop, you need to get your amp power from the same outlet you connect your computer.

Have you considered a good pair of headphones?

I agree, it's a ground loop. To prove that, if you run your laptop on battery it should go away. The Behringer or an Ebtech hum eliminator should work, or you could also just put a 3-to-2 prong ground lift on the laptop and that would probably work as well, if you're comfortable not having it grounded.

I've had an Audiophile 2496 for at least 7-8 years with no issues whatsoever, and the sound is excellent (Windows XP).

No comment on the speakers, I'm not very knowledgeable.

Cheers,
Michael

Thanks for the replies.

Ok, I got Pianoteq Play working tonight, with no hum (see below).  I'm not sure the results are quite what I was hoping for.  At first, I played it through speakers, which wasn't quite satisfactory with any of the output modes for the K1.  This is peculiar, because I do like the sound of some of the demos that I downloaded and play via my Cowan MP3 player, through the same amp and speakers.

Then I tried headphones.  It is with headphones that the otherwise satisfactory output of my Kawai CA51 really sucks.  I had my wife switch between output modes and K1, C3, and M3, while I wore the headphones. 

With headphones, I definitely prefer the binaural output.  The K1 is a little too bright on the bass end, while the high end is Ok.  The C3 is preferrable in the bass, but too mellow on the high end.  The M3 is just too mellow in general.  So now I'm not sure what to do next.  I think I will spend some time alternating headphones with Pianoteq, vs. the Kawai with its built-in speakers.

Ok, about the ground loop.  I didn't want to wait for ordering a commercial device to fix this, so today I built a dual differential amplifier circuit to fix the hum, using a pair of AD620 instrumentation amplifier chips.  This cleans up the hum perfectly, leaving only a trace of high frequency buzzing and some clicks and pops, that seem to correlate with computer activity.

The next trick will be to see if a better sound card device will improve the noise level.  At this point, I don't know if the remaining noise is high frequency common mode, or differential mode.  If differential mode, it is from the sound card itself.  But it could still be due to the fact that while the low freq. CMRR of the AD620 is very high, at high frequencies it begins to deteriorate.  Nonetheless, at the volume settings I need to work with, the noise is actually imperceptible even without any notes going.  There is also the option of experimenting with some of the THAT 1200 series devices, which I should probably do, since they are remarkably cheaper than AD620s, and now are available from Mouser Electronics.

I will probably pick up a Jensen transformer, or perhaps one of the other devices mentioned.  I should put my AD620 circuit on a PCB and open-source them.  The PCB plus parts cost will probably work out competitive with commercial products.