I have bought some Socket 7 boards with three Winco W-P55VT2 being among them. Visually they look OK, except one which has cache chips missing.

My focus is on the other two boards with cache chips. As mentioned before they both look visually OK and I couldn't find any significant damage. Somebody did some bad solder job on the voltage regulator though so my first suspicion was that some components went puff.

Checked voltages, crystals and bus speed, everything seems to be alright. But the northbridge of one of the boards got hot to touch after a while, signaling a short. Checked the resistance between 5V rail and GND and it was around 50 Ohms. In comparison, the other board, which doesn't have any overheating components, has the resistance in tens of kiloohms. So, I lifted +5V input legs from the nortbridge and sure enough, the resistance went to normal, at least in comparison with the second board.

The board without cache chips has a short too, except it's the southbridge that's damaged. What I want to do is use this board as a donor and transplant the northbridge to the other board with presumably shorted northbridge. Of course I have no idea if the donor northridge is OK. I'll probably start with removing the southbridge chip from the cache-less motherboard and check if the short went away. If all resistances on power rails will be OK, I'll consider the northbridge as functional. We'll see how it goes.

The second board with cache is a mystery though. There's no visible damage, voltages and frequencies are OK, but there is no activity on data/address lines in ISA slots. They're always high, probably pulled-up by some resistors. Chipset is probably damaged too but it's not creating a short. If I'm successful in reviving the first board, I'll use it as a reference and try to diagnose the second one further.