Rev. 27 Jan 2015
Tue 21 Jul 2015 Drew & Zach:
We now believe that the Arctic Silver thermal paste is what is responsible for the burn-out. Although it's website states it's not conductive, it explicitly states that it should be kept away from “electrical traces, pins, and leads”. It may have conductive properties at high frequencies. We removed it from all other boards and haven't had this problem again since then.
We will attempt repairs on Board 3 if one of the six other currently working boards fails (Boards D, 1, 2, 4, 5, 6). The biggest problem during repairs will be ensuring that the charred PCB isn't significantly conductive. Reference material to repairing extensively-charred PCBs: here
Wed 15 Jul 2015 Drew & Zach:
Blue channel Test 4 works. The final output works fine and the HELA-10 stays moderately warm; works for extended periods of time.
Had to resolder R5 on the Red channel to get Test 4 working.
Red Test 7 failed; found that the upstream pad of C11 hadn't been soldered completely; afterwards, Test 7 passed, but showed saturated power levels: main signal flattened out at +20 dBm, 1st harmonic flattened at +8 dBm, 2nd at +3 dBm.
Red HELA-10 pin 10 VAPORIZED while testing final output; was removed; PCB has a scorched hole where pin 10 was; we cleaned it up as much as possible.
Green channel Test 4 worked for a while, but HELA-10 pin 10 VAPORIZED in a later test.
7/9/15 Boyd and Zach:
Construction on this board was recently completed. Testing reveals that the red and green channels are working perfectly, while the blue channel is not displaying properly. The problem may be just as simple as a faulty connector on this board, as the output does appear to be fuzzier when the DVI cable is plugged in. The fuzz seems to loosely suggest the shape of letters.