Yamaha CS40m Sample & Hold mod, plus patch write not working

[1way2life]

Hi all,

I have a CS40m that I am restoring to the best of my ability.
Power supply rebuilt, and all caps replaced.

Got most of the unit working. However,

1. the sample and hold modulation isn't working. When selected on a VCO, and depth increased... pitch just decreases and stays, no change in pitch from LFO.
2. Also, writing a patch does not store. All the patches are silent, or have a noisy quip. After attempting a WRITE to a patch number, that patch just remains silent.

Any suggestions on where to start tracing these issues? I am a novice/intermediate tech.

Thanks!

[Mooger5]

I don´t have any experience with the CS40M, but the last two pages of the service manual seem useful. The block diagram and the overall circuit diagram.

Does the S/H modulate the VCF? If it doesn´t, maybe it´s not working at all, so better start checking from the source.

In the Block Diagram, locate the following:

MOD 1/3 PCB, divider IC5, noise source only if the noise isn´t working at all, then FET 2, 3, IC7 and TR8.
If all checks OK, proceed onto the next PCBs, always following the SH or S/H lines:

MOD 2/3 PCB, Modulation Function Selectors, Mod Depth, Buffer.
After these, the respective S/H line go directly to the destinations: the VCO and the IF 3/4 VCF PCBs.

Note the several "Sample and Hold" circuits in the PA board. Those only hold the CV from the keyboard controller and are not relevant to the problem in question, so just ignore them.

Regarding the patch memory there are two batteries for backup...

[1way2life]

Hi Mooger5,

Always great advice from you... thanks!

I have traced it to a point. I am NOT getting the S/H random square waveform out of the MOD 1/3 section where the other LFO waveforms appear. Instead, I get a steady ring pulse, with no variation.

I am getting noise out of the noise generator and it reaches the FETs, but after that I get this weird pulse again. However, I dont know if Im supposed to have it either at this point, Im only guessing I am.

So the big question is, should I be seeing the S/H random square pulse at IC7 pin 1?

RE: Patch writing... I solved this today. I found a trace break around IC 13 (pin 14) going to the 1uf cap at TR9.


Thanks for your help!

[Mooger5]

Hello, it´s alright.

There should be a plain output of the sample & hold at pin1 of IC7, yes. If not, and if there´s a pulse at pin3, then IC7, or the 33k resistor is probably faulty. Inspect the traces and redo the joints to be sure.

Good luck!

[Mooger5]

BTW congrats for fixing the patch storage yourself. It´s very rewarding isn´t it.

Now, to help figuring out what´s going on in the S/H circuit: the noise generates random voltages. The LFO sends a quick clock pulse through the 0.01 capacitor to FET 2. It´s just a brief thump but enough to activate the FET that serves as a gate which lets just a snippet of the random voltage pass through. In other words, it SAMPLES that voltage in a very short time. The 0.039 cap then HOLDS the voltage because the high input impedances of FET3 and IC7 make sure the cap will not discharge before the next sample occurrs. So the output is a steady voltage that only changes at regular intervals.

Hope this helps with the troubleshooting.

[1way2life]

Yes it is! I was dreading tracing through all the logic, as Im not very good at it. But then I saw the broken trace, fixed, and all was good with the patches. Very satisfying!

So, from your description of the S/H circuit, I suspect the FETs or the 0.039 cap, as the pulse that comes through Tr8 and D2 continue through the FETs and to pin3 of IC7.

Does that sound plausible?

[Mooger5]

From experience, it's the opamp that usually fails. One of the inputs has probably failed short and not presenting the high impedance for the cap to discharge slower. One of the power rails must now be on the loose, which would explain the ringing.
See if it's fried to touch.

[1way2life]

No heat, seems normal. I Swapped 2 of the 4558s and same issue. Really weird.

[Mooger5]

Yeah well I forgot about the other half of the dual opamp. Never seen one channel bad while the other channel was working. It happens with logic chips, though.

Well the pulses at pin3 of ic7, do they change randomly in amplitude? If so, it means the noise it being correctly sampled by FET2 and it could be the capacitor that isn´t charging, Replace.
If the pulses have always the same amplitude, then it´s the noise that isn´t passing through FET2. Replace.

If it still doesn´t work, then FET3...

Sorry for the trouble and better luck this time.

[1way2life]

Thanks... I ordered the replacements to try. Was a bit hard to find the 2sk105a FETs.

I'll let you know how it turns out!

[Mooger5]

OK. If it gets fixed, please don´t forget to report.

[1way2life]

Hi Mooger5,

I'm still waiting on the 2sk15a FETs, however I decided to start on restoring my CS30, and low and behold, a nearly identical S/H circuit!.

I popped it open, and started scoping its FETs. Sure enough, the S/H signal is right at the first FET, where on the CS40m just passed the pulse from TR8.

Im going to assume replacing the FET2 will do it. Of course, I will let you know as soon as they arrive and I replace it.

Cheers.

[Mooger5]

Well done! I mean, it´s not yet fixed but everything tells it will, so congrats!

[1way2life]

Mooger5,

Got the FETs last night. This morning, I desoldered both from the MOD board and tested them. Sure enough, FET3 was shorted. Replaced it, and the 0.039uf cap (used a MLCC) and it produced the cleanest S/H I've seen to date

Now the CS40M is totally restored!
Thanks for your continued help.

[Mooger5]

Glad to hear! Enjoy the Yamaha.
I guess the Sample & Hold module could be nicknamed the Vangelis module. Because it functions as a channel from which music emerges from the chaos of noise, and you had to replace la FET sauvage