McIntosh C26 - 1st to totally stump me in 37 years

I suppose there are other technicians on this forum. I am at a roadblock here. I have my own C26. It had an intermittent low level "groan" on one channel. It had never been re-capped, so I went for it. I used ONLY Nichicon 105C long life caps. All caps are the same or higher capacitance. All are a higher voltage rating. Now I am left with a 120 Hz hum. Not horribly loud, but loud enough that the preamp is not useable. All four main electrolytics were substituted with other ones to no avail. I left in the new ones. ALL capacitors in the ENTIRE UNIT have been changed! I have spend DAYS on this one item. It is coming close to the "throw it in the trash" point soon. ALL components on the power supply pcb have been replaced with new 5% resistors and Nichicon caps. If I remove the lead from pin 2 on each side pcb and inject signal from a portable audio generator or an ipod, everything is NOISE FREE! When the source items are turned off and volume set at MAX on the C26 all is silent! ( I am feeding my sources to the C26 through .05 uF caps for isolation.) I am stumped to the max. This problem is probably too involved for a forum, but I am taking the chance someone might have an idea for me.
Any interest in contacting McIntosh or sending the piece to them for repair?

Check all grounds. Also, make sure you didn't accidently reverse polarity on one of the caps. Last I can think of, make sure you didn't replace a non-polarized signal coupling cap with a regular electrolytic type. There are free schematics on the net.
That would solve the problem, HOWEVER, depending on the technician or what the company policies are they MIGHT want to replace every Nichicon cap that I have replaced with factory caps. It leaves too much open space for me to worry about cost wise. It just bugs me that I can't fix it myself.
You could also measure every trace with your multimeter to check for zero ohms. There could always be a hairline crack in the trace. A lot of times, the trace will break close to the solder joint. Clamp test lead (if you have clips for your test leads) on the new caps wire leads and measure for zero ohms at the best place you could make contact, that will catch a hairline crack in the full run of that trace, if that may be a problem. There could possibly be a cold solder joint, and flux may be insulating a connection too. You can't just go by looks.

I've mentioned a couple of times in the past when someone ran into a similar problem, not to make a lot of changes at one time. I've had basic electronics ages ago, but never make that many changes at once. Those past threads seem to be gone. Even someone that does this for a living could run into this. So I just do new parts like that in sections, then test it. Then do another small section.

Also, when injecting a signal that way, use junk disposable speakers for testing. I must have picked up some RF while testing the way you mentioned 30-40 years ago. I wiped out a pair of tweeters that RF must have blown, I assume. Fortunately, they still had new replacement tweeters back then. After that, I always use junk speakers for testing first.

Another thing, especially replacing coupling caps, I'll try the new ones just one stage at a time. Then I bench test it, and if all is okay, then put the unit back in the system to see if I like the sound of the new caps. Plenty of times I didn't like the new caps. If I had done them all, that would have been a big disappointment for myself.

Hopefully you get it figured out. I thought I'd mention these things to save someone in the future from learning the hard way too. Maybe this will stay in the archives.

Old McIntosh units used rivets for grounding lugs, and willing to bet one or more of those has been compromised during the re-capping.
I have an old 'cattle prod' heavy duty soldering iron and acid flux from plumbing supply that I use to solder the ground lug rivets onto the frame they are attached to and this really helps the old Mac's. Never use acid flux on circuit boards, but it's perfect for soldering steel lugs to the steel framework.