Main Power Failure
Despite all my joy about the new member of my vintage computer family, I was a bit disappointed when absolutely nothing happened after turning the machine on for the first time. But these days, there are no problems anymore, only challenges — so let's get to work!
I quickly located two blown fuses (F101 and F102) in the high-voltage section of the power supply. Replacing them with new fuses fresh from the hardware store restored the computer to a functional state at least long enough for its basic functionality to be confirmed. After a few seconds, however, there was a series of popping noises, accompanied by copious amounts of foul-smelling smoke pouring from the case.
Two exploded capacitors (C101 and C102) were identified as the culprits for this unpleasant phenomenon, although after more than 20 years of inactivity, one can hardly blame the components for no longer feeling up to the sudden stress and preferring to give up the ghost.
In sourcing suitable replacement parts, the Sharp MZ fan site sharpmz.org proved extremely helpful, thanks in part to the service and user manuals available for download. And indeed: After installing the new electirc charge storage devices, there has (so far) been no further development of unpleasant odors.
It should be noted, though, that another fuse, the F301, also failed later.
In the Augean stables...
...the device may not have been stored, but the previous owners did run a farm, and the motherboard looked as if it had seen the occasional batch of manure, slurry, or some similarly corrosive fluid. I only became aware of the full extent of this, however, when the Sharp suddenly exhibited a kind of garbled screen symptom — albeit even prettier than the PET, with changing patterns. A closer look at the circuit board then revealed a fluffy rash on the board as well as rusty deposits on the pins of numerous ICs, presumably causing fluctuating short-circuit currents, which ultimately led to the device's strange behavior.
The solution was a 35°C gentle wash in the dishwasher (with a few drops of hand dishwashing liquid instead of dishwashing powder), as well as a touch-up of the most stubborn spots with a little soapy water and a toothbrush. The drying process took place overnight on a radiator, then for a few hours in a convection oven (50°C, with the oven door open) and in front of a fan heater.
Even though the circuit board's protective coating came off in some places during the whole process, revealing bare circuit paths, the device hasn't exhibited any wild display extravaganzas since the cleaning.
The End of Eternity
The final problem — er, sorry, challenge — was the fact that about one-third of the keyboard's keys had no effect when pressed. Unfortunately, I initially had no idea how to open the keyboard for cleaning — it seemed to be built to last, with closed shafts around each key that firmly connect the circuit board to the base plate.
To avoid opening the keyboard, I initially tried the dishwasher trick again – but this initially didn't seem to improve things; after a gentle wash and extensive drying, only about half of the keys worked. The only solution – as threatened in the forum of the Verein zum Erhalt klassischer Computer (Association for the Preservation of Classic Computers) – was to desolder all 73 keys. After another wash without the offending circuit board, the number of dysfunctional keys had shrunk to around a dozen, and the stems of the remaining culprits could now easily be clicked out of the openings in the plate.
With a little patience, even the individual defective keys could now be disassembled (mostly non-destructively). The internal structure of the keys revealed in this process suggests that the engineers at Sharp back then were either completely crazy or absolutely ingenious.
After several attempts, including various combinations of pressing, scratching, and brushing, cleaning the contact surfaces with 70% isopropanol applied using a thinned-out cotton swab proved to be a simple yet effective method of reviving the troublesome keys. Unfortunately, I only discovered this after desoldering the entire circuit board for the second time. On the first attempt, my continuity tester thwarted my plans by deceiving me about the ability of many keys to bridge their contacts, even though the key resistance was actually still so high that it couldn't pull the potential at the corresponding inputs of the 8255 (IC50) responsible for keyboard sensing down to low level compared to the 10 kOhm of RA2. In the second attempt, I used an real ohmmeter to verify sufficient conductivity of the cleaned keys.
On the third day, however, when I began desoldering the keyboard board again because of two keys that were still not working, I noticed that a solder pad on each of the buttons in question had apparently come loose during previous disassembly, so that even the generous amount of solder I had applied to the key pin couldn't make contact with its corresponding circuit board track. Small wire bridges to neighboring solder points finally solved this problem as well.
Just Vegetating
Hooray, now the good old machine is ready for new tasks again! However, the MZ-80A — like a modern PC — has no operating system in ROM other than a kind of BIOS (the "monitor"), not even the BASIC interpreter common at the time. But the hunt for software for this classic machine is another story, and one for another time.