All right, so just in case you were wondering why an analogue monosynth from 1979 has a CPU board in it, it's because the synth has a really simple memory recall function. You can either use the sound live from the panel knobs, or you can save and recall the panel settings to one of eight memory slots which are on buttons.It's really very primitive, with the biggest restriction being that the panel controls are disabled when you're using the saved settings, which means you can't edit or fully manipulate your saved sounds. I generally used to keep it set to 'panel' for this reason.*But since all the controls pass through the CPU board, this is the next thing that needs to be working before I can move on to anything else. Initially, the lights are flashing nonsense and buttons are unresponsive which is not cool.So here I have replaced whichever generic logic chips on the PGM board that were cheap and easy for me to obtain in small quantities. Don't they look nice in their new sockets? Maybe it's overkill but these are nearly 40 year old ICs, they only cost a buck each to replace and the socketing should make future repairs so much easier. If this board starts working I'll probably do a bulk order and rechip all the boards. Capacitors have already been replaced in an earlier repair attempt. I also checked the connection to each nearest neighbour when putting the sockets in, so I know those parts of the circuit trace are all good.Plugged it in, fired it up... and the memory buttons still seems unresponsive. Dang. The lights aren't going nuts like they used to, but they aren't behaving correctly either.Well hey, now I get to use the oscilloscope to analyse it. I used the oscilloscope to observe a correctly-encoded 8 bit number being sent from the buttons on the panel into pin 4 of the CPU when that button was pressed. Neat! Another test point on the board shows that this value is being correctly registered and stored when the button is released. Good!I connected to another test point and saw the analogue controls being correctly multiplexed. A few of the individual knobs and sliders are responding a bit weirdly but most are good and the multiplexing itself seems to be working fine.And wait... I said earlier that the LFO trigger light was blinking and responding to the LFO speed knob. Well, the LFO speed knob is one of the parameters stored by the CPU, by golly. I check again and I see that the light does indeed respond to the speed control when the CPU is set to 'panel'. And when I push the memory buttons, the light becomes unresponsive to the knob and starts blinking at different speeds. It's using stored values! Furthermore, I discover that I can set the speed on 'panel' and then write it to the memory slots.Conclusion: the brains are working fine now. The lights on the panel board must be malfunctioning. Very good news. Over to the panel board next.* One possibility for a future project would be to replace the entire CPU board with an arduino-based system which I could program with a custom OS - one that supported live editing of stored settings, memory expanded from 8 patches to 64 or 128, and the ability to control all panel settings via MIDI would be three really useful things that could be done with that. You could even go nuts and program in additional software oscillators and patching options if you had an arduino in there. This is obviously a big step up in difficulty but it was something I was considering looking into if I couldn't get the original CPU to work, and it would still be fun to attempt one day.]

Stupid Photobucket. Now all my old image links are broken.Anyway what happened with this in the end was that I discovered the malfunctioning lights on the panel board were indeed coming from a problem with the brains (the panel board itself is fine), and then while I was checking this the brains got worse. And it looks like the keyboard matrix chip, which is another special Yamaha chip, might be dead too. These are the two chips you really need to not be dead.So it looks like the only realistic way forward is to knock up some kind of arduino-based thing that does the duty of the CPU and also the keyboard matrix. Which is potentially doable - somebody online has already done the keyboard matrix - but it's definitely back in the too-hard basket again for now. At some point I'll run some signal through the actual synth circuit and see if ANYTHING on the old boy still works.I hate to give up on any gear and me and this synth had some good times back in the day.

Have you looked into the Spin FV-1? http://www.spinsemi.com/products.html If the CS20 isn't fixable and you want to make something else out of it...

I haven't... it looks like a fun thing but I'm not sure it would help in this situation? Not many control inputs.I think all the analogue circuitry, oscillators and filters and so on, on the CS20 is OK, or if not, is repairable. The main problems are:1. the keyboard matrix isn't working, so no voltage signal from the keyboard. Somebody online however has made an Arduino drop-in replacement for this board which also gives you MIDI control and an arpeggiator, so there's a good starting point for that at least. I haven't tried triggering it from an external voltage yet but that's another thing to try.2. the CPU has apparently shit itself. The CPU is solely responsible for acting as a go-between from the analogue pots on the panel, to the control of the actual synth circuits. It's there to give you the ability to save and recall patch settings. It might be possible to jack it so the knobs on the panel simply pass through to the sample and hold board directly through a multiplexor, or recreate the functionality of the CPU with a programmable chip of some kind. The CPU really isn't doing much but all the panel controls pass through it.I'll have to put this up on the bench again soon and have another crack...

Anthony Flack wrote:Did I say the power supply was all right? Not really. I went to check the power lines on the PGM (CPU) board, only to find the -10v line was dead again. Yup, the fuse was blown again. I disconnected all the boards from the power supply board, replaced the fuse and fired up the power supply on its own. The fuse melted right away. Further investigation of the power supply board revealed a tantalum bypass capacitor on the -10v output which had failed and was shorting the line out. I replaced it, as well as its buddies on the other lines for good measure, and fired up the power supply again. Nice smooth DC power was once again flowing.Plugged in the boards again, turned it on and, yup, the fuse blew again. Replaced the fuse and started adding each board one at a time until I found the problem. The line that is blowing the fuse is the one that powers the MOD and FA boards - filters and waveshaping. I take those boards out for now and start it up again. It's not going to make any sound with those boards missing, but I should be able to check the CPU now.Power is stable and not blowing fuses. And now there are signs of life on the front panel!The LFO trigger light is pulsing, and responds to the LFO speed knob. It looks like the oscillators have come on, yay. The panel lights from the CPU are lighting up as well, but they're flashing nonsense and the buttons are not responsive. Something is likely not right on the CPU board.Using the oscilloscope I successfully verify that the clock is delivering a 100kHz square wave as it should, and the reset pin is doing its thing correctly, which was very gratifying. I like having an oscilloscope.I think I will do what the man in the video suggests, assume the main CPU chip is all good because I'm pretty fucked if it isn't, and replace all the cheap generic logic chips on the CPU board which are a more likely cause of scrambled brains. And with any luck the panel buttons may come to life.Hi I'm new here from the UK... I also own a CS20M in spanking condition, but it has started popping the little glass fuse in the back panel. I believe also the tempo/arp light on the left stays on rather than blink, but only for the 2 seconds that the fuse holds out.I'm not au faix with electronics, but I would like to keep the old boy running if I can... been reading your posts, but not sure what to do next.... any advice?

If the fuse is popping then something is shorting out. Stop powering it up for now before you fry one of the irreplaceable chips. It may be time to start to get au faix with electronics or befriend somebody who is. I'm not an expert by any means! But that service manual I linked to above is going to help out a lot. The first thing to do is unplug the power supply from the rest of the circuitry and make sure that is working well by itself and supplying stable power at the correct voltages. One of the nice things about the CS20M is you can plug all the boards in and out individually, so you can isolate which board is shorting out, pull it out and start hunting for faults.I didn't get that far. I think that my CPU chip and possibly my keyboard matrix chip are malfunctioning, which are the chips you can't afford to damage. Your problem could be something tiny though, and most everything else apart for those two chips is replaceable. Get somebody who knows what they're doing to take a look at it, give them the service manual and tell them to be careful with those Yamaha chips.Good luck, I'd like to hear how you get on.One day I'll open up mine again and see what can be done with it.

Hi Anthony, thanks for a prompt reply! I am thinking of doing this as well as visiting an old buddy who is an electronics guy... retired, but maybe I can talk him into it? I will keep you posted. Yes I saw the diagrams ... I'll print those off as it might help the electronics man decide to try.I used to have a Crumar Spirit - a similar synth made in Italy... ended up in a private museum as it turned out to have the serial number 00007. Wish I'd kept it... /