Yes, the ground is shared, but i tried isolating the fan by using a B0505S isolator (worked great for other ground related noise problems so far), but the result was the same.
I used this simple circuit as a base for my amplifier:
I just added a TDA1524 for tone and volume control, a uPC1237 for speaker protection and some other tweaks, like relays for mono/stereo switch, extra fuses etc.
I'm also adding eagle .sch and .brd files if you want to look through my work: **broken link removed**
If you don't use eagle, i can post screenshots.
It's my first serious attempt at making an amplifier and i'm satisfied with the result, but if there's a solution for the problems stated above I'd very much like to fix them.
The fan is actually optional in my opinion, but i was quite surprised it would have such an effect on the sound.
Another idea that comes to mind is actually using a separate 12V power supply to avoid tapping into the transformer. I originally used a L7824 and L7812 regulators in fear that 24VAC after the diode bridge would be too high for L7812. However, the voltage appeared to be too low for L7824 which in turn generated awful noise and i had to remove it.
I also pondered how to separate audio input ground from the chassis ground, but didn't manage to figure it out.
I'm now trying to figure out a way to make a VU meter to be able to monitor load level, but didn't manage to find one which would be able to handle +-75V. I'm hesitant on making one at the amplifier input (after TDA1524), because i feel that it may not depict the real situation