Changing all the ICs on the KLM 356 board and a 2SA733 transistor that had shorted was the next task and with that done, it was time to test the synth to see if would work albeit still on life support with the bench power supply. With everything now wired back together as it ought, I applied power and tried the keyboard to see if anything, sound or triggering wise would happen. Nothing - it was a bit like Steve Race on 'Face the Music' with his dummy keyboard - a rendition of Bach's French Suite Courante in Cm sounding more like a 1960s typing pool. I also checked the key assign buttons on the front panel and they didn't change state so it looked like the processor wasn't scanning any switchgear, keyboard or otherwise. This assumption was only half right as the key assign switch scanning was cured by changing IC2 (74LS174) even though it had previously been changed. It may have been a faulty new part or it was out of spec. Why is IC2 and IC3 the only ICs that are socketed on that board
Anyway, that only left the inactive keyboard and it didn't take much to realise that the keyboard was on shared data and address lines that other switches worked OK on so could it be the keyboard itself
I've seen a lot of these membrane type key switches give problems but never one where every single note was out - but that is what the problem was - every single note was out. Removing all the rubber contact strips and then cleaning the gold plated contacts on the printed circuit board and the black conductive rubber 'pips' using isopropyl alcohol, cured the problem perfectly.
Although the keyboard now worked, there was something very much amiss. Scaling, tracking, calibrating
If the synth was working perfectly before the previous owner discharged the contents of the National Grid into it, then why was the calibration so far out now
I was still very suspicious of the SSM2033 ICs and convinced that they were faulty. However it always pays to completely ignore what previous owners say about a synth -
"honest guv', it was working fine before I ran over it in the car"
could really mean
"there was a fault with it and so I turned every trim pot I could find and it got worse - it was then that I decided to reverse over it in the SUV in a fit of frustration"
. I was to discover later that my Mono/Poly had been a victim of trimpot butchery at some stage.
Before carrying out a full calibration as given in the service manual and still worrying about those SSM2033s, I thought it was about time to do something about the power supply as I was still operating from the bench supply. I replaced the 110 volt mains transformer with a 240 volt transformer of similar dimensions - this in itself was a bad move as I will reveal later. Moving on to the power supply board KLM-376, it was in a bad way with burned-out components as well as all the semis. trashed. The M5230 IC is difficult to source and I don't like the design of the supply insomuch as that there is no protection whatsoever in the event of overvoltage, short circuit or low flying hang gliders. The + and - series regulator transistors have to sink around 14 volts at about half an amp (more on the -15V) and run very hot indeed especially with the inadequate heatsinking. So, the plan was to just get the power supply up and running as quickly as possible and design a new one to be fitted at a later date. I used the same original transistor arrangement for the + 15 and -15volt rails and used a fixed 78L15 regulator as a stabilised source for the +15v which gave an actual 15.3volts (calibration states + or - 0.5V) and a variable LM337L regulator with a 10 turn trimpot to set the -15v to within the required + or - 0.01v. These new components were easily accommodated on the existing board once the old components were removed. The existing arrangement to provide the +5v was used using all new components but with a ten turn trimpot used to make the adjustment to within 0.01v easier when setting up.
The power supply works fine although the power transistors on the 15 volt lines run hotter than I'd like and there's no real protection. It will suffice until the new improved PSU is built. It will be super accurate and have very fast built-in overvolts protection.
Next time... tuning, ranging, scaling, loud hum, stiff drink.......