What about the LM339?
There are a lot of voltages present on pins of the LM339 (or at least higher values (eg: 5.8 v 1.1)) in the problem board when compared to a known good one in the same machine. In the good board, there are 7 pins that have no voltage present, whereas only 1 pin of these on the problem board has no voltage, the others range from 0.4 to 5V.
Does this indicate the problem is th 339, or prior to the 339?
Test the 339 on diode mode with the power off you might have a short internally
