This lovely, small fellow has been brought by one of our customers after some failed repair at one of Birmingham’s computer shops. Manufactured in 1992, it was working perfectly fine until a small liquid has been spilled on the keyboard and partially on the trackball.
Computer was starting up with chime, however, there was no picture on the screen and no HDD noise. Customer has stated that after the damage he could hear the chime and noise of a spinning HDD (this model has a very capacious 120MB IBM drive). Unfortunately, the previous repair company gave up on this repair after over 4 months of trying. We have then opened it to see what has been done by a previous repairers, and found that there was still a residue of some substance on the motherboard (how could that be overlooked?):
We have then cleaned a whole motherboard, so at least it looked way better. Unfortunately, after reassembling, the machine was still not willing to cooperate. Logic board has been then pulled out again and examined very carefully under the microscope (especially the area which had a liquid residue), and we quickly discovered something suspicious – one of the vias which was looking slightly different than others:
We have checked on the schematic diagram that this via connects pin 3 (source) of Q7, dual P-Channel MOSFET (IRF9953), with a cathode of D22 (10BQ100T), which provides 5V for supplying the logic circuit and… for HDD. Measurement of continuity has uncovered that there is no electrical connection between cathode of D22 and pin 3 of Q7. That was a turning point of our diagnostics. We had no other choice than repairing that broken via, to try and get the proper 5V voltage on pins 5 and 6 (drain) of Q7.
We have removed a soldermask from the surrounding area of broken via (both sides) and put a silver plated wire into the hole.
Now we could solder the wire to restore an electrical connection between both sides of PCB. We wanted to move D22 diode a little bit away of that via because it was slightly interfering with soldermask removal and wire soldering.
After restoring that broken connection, we have moved D22 diode back in place:
We have connected the logic board to our bench power supply and realised that it takes more current than previously. That was a good sign so we have decided to reassemble the machine. Fortunately, it was a good decision – machine has started up, loaded the operating system and we could watch this beautiful piece of electronics back in action!
It is always worth to examine the logic board under a microscope – especially when the machine has been liquid spilled. All signs of corrosion can be easily found if you do not try to limit the time of visual inspection.