E-mu Systems Proteus/1
One day Greg came into the workshop and dumped a pale brown mud brick on my bench.
Well, actually it was more like a mud flagstone than a brick, thin, flat and rectangular.
I don't get much call to service flagstones of any sort, but on closer inspection this turned out to be a 19-inch rack-mounting flagstone!
Now there's something you don't see every day. Really solid-state. But I guess that made it my problem.
He is one of our self-selected volunteer tip crew who keep our local rubbish transfer site clean and tidy in their spare time.
It's dirty, physical, risky, and can have it's dead cat moments, but there are some real compensations - what is junk to one is another's gold and he had previously scored a welder with a very minor fault.
He had been pottering about, and he had happened upon something embedded in a dried-up mud puddle on the ground that looked like a 19-inch rack case.
Knowing I was interested in 19-inch rack hardware, he dug it out and brought it around.
Carefully chipping the dried mud off the outside I was surprised to finally uncover a E-Mu Proteus/1 MIDI-driven rack mounting music synthesizer.
Initially it looked like a write-off but as I dug I found that the mud had been so thick to start with that it hadn't flowed into any of the cavities such as sockets, and only a little had got inside.
The unit consists of a small power supply and a front panel control and display board connected to a single board computer around a Motorola 68000 CPU.
The electronics were virtually untouched by the mud and after stripping, washing, and reassembly it was almost as good as new.
Washing? Sure! With the exception of transformers, pots and switches, you can safely give most electronics a bath in warm water and dishwashing liquid when needed without damage. The key points are, get all the muck and soap off so there is only clean water remaining; shake off and dry gently but totally (sunshine is good). Plastic cases are even easier once stripped of metal but shouldn't be left in direct sun to dry.
The mains switch was broken but after getting some power connected it announced itself in the display, and could play its extensive demo okay.
With a single board computer such as this the moment the display started to do something sensible I was seriously chuffed because in that moment most of the electronics had passed muster and it was now unlikely that anything too serious could remain.
Deeper investigation showed the front panel rotary encoder used for selection in the display needed attention, variously not stepping, stepping the wrong way, or jumping a step or two. It could have been faulty interface logic, but no need to go looking for aggro when things were going so well - it may be Irish, but I'll check what I know I can fix first.
These stopless 360° rotary encoders are becoming more popular in digital gear to produce “go up” and “go down” pulses for selection.
They work on the two-phase principle where a pair of contacts or an optical encoder like in a mouse produce two skewed but overlapping pulses with the circuitry using the order of opening and closing to determine the direction of rotation.
In this case the encoder was a tiny sealed switch.
Sealed is good until dirt actually does get on the contacts, then it's a real pest to get open to clean.
Often these are held together with a heavy outside metal body with tabs bent over to hold the assembly together. These tabs can be gently un-bent and the switch disassembled.
In some cases the assembly is held together with long tubular rivets that can be drilled out and later replaced with long, thin nuts and bolts.
Actual disassembly should be done very slowly and carefully to observe exactly how all the parts go together, and to avoid any internal springs shooting some critical tiny part across the workshop into a parallel universe. The really paranoid can do this inside a large plastic bag to catch any flying bits.
Sure enough the gold contact wires looked pretty gungy and a gentle cleaning with alcohol on a cotton bud had them looking like new.
And reassembled it worked like new.
The piston-type mains switch was a special too but a rummage in the junk heap turned up an old cassette deck that had a similar one, and a bit of mix and match got that right (even if it is the only Proteus with a silver mains switch ;-).
Since then it has been hard at work in the heart of the keyboard pit.
Why did I go to the trouble of repairing the controls when I could have ordered the right parts?
The most obvious advantage is time, I'm impatient for a result, and I don't expect a special control like that to be easy to find or cheap.
And whenever you order parts for a job you have to put the job down, pack all the parts and screws of the disassembled whatever, together with the bench notes, into a box and put it aside until the parts come, perhaps weeks later.
Then you have to pick the job up again, come up to speed, remember what you were doing and up to, before you can complete the second half.
I have found it a constant across personal and professional activities that switching jobs mid-stream has an overhead of about 15+15 minutes, or half-an-hour of dead time. It's also very satisfying to do something entirely on your own resources.
“You amaze me” said John. “I look in a bin and all I see is crap. You look in the same bin and come away with an armload of useful parts.”
Greg reccons that I'm “A practitioner of curmudgeonly arts”, meaning I have a very tastefully selected junk heap. At least I don't have a freezer half-full of roadkill or spectrograph and an x-ray machine cluttering up my shed. ;)
http://www.ozvalveamps.org/proteus.htm | Last update: 22:37 14/01/06