[sdiy] Weird Yamaha CS40M problem

[Antti Pitkämäki]

Hi,

A small update: I think I solved why the "weird oscillation" reaches the
output even when the VCA is off. This synth has a feature labeled POA,
"Pass Over Amplifier", which sends a signal from the VCF directly to the
output amplifier, bypassing the VCA. It's controlled by a semiconductor
switch (TC4016BP) on the output amplifier card. I was able to detect the
weird oscillation at the input pin of the switch and also at the output pin
of the switch (in a distorted form) even when the switch was set off.
Apparently the weird oscillation (15V peak to peak) exceeds the voltage
limits of the switch and peaks get through. Also removing the VCF/VCA
connector from the output amplifier card stopped the weird oscillation from
reaching the output.

But what causes the weird oscillation? Well, I noticed the filters of both
voices share an IC, which is a dual opamp (NJM4558DV), each filter
utilizing one of the opamps, which are configured as simple buffers.
Incoming signal (from the oscillator mixer) goes through this buffer. The
buffer's output is connected to the input of the VCF through a resistor.
Also connected to the VCF's input is the pin of the VCF labeled "FB",
meaning feedback. Now could the shared dual opamp be the culprit, causing a
similar symptom for both filters? It's really the only shared component
between the filters. Well, there's another similar shared component, but it
buffers the VCF's envelope generator control voltage, so I don't think
that's the problem.

So I guess it would be a good idea to try changing the mentioned dual opamp
buffer at the VCFs' inputs (being the only common component for both
filters, which both malfunction in the same way)? Could such a component,
in theory, cause out-of-control feedback at the filter? It's weird that it
only occurs in very specific conditions. Somehow I feel a defect filter IC
would have made more sense, but it would be quite strange for two separate
filter ICs to malfunction in exactly the same way.

Regards,
Antti

la 18. syysk. 2021 klo 0.34 Antti Pitkämäki <antti.s.pitkamaki at gmail.com>
kirjoitti:

> Hi,
>
> I’ve got a Yamaha CS40M with a strange problem, which I’m unfortunately
> unable to solve by myself. I wonder if anybody could give any insight into
> the following:
>
> The synth works otherwise just fine, but certain settings result in a
> strange situation. I can make it happen by setting both oscillators to 8’
> square waves (50%), maximum volume for both in the mixer section, filter
> resonance at maximum and filter cutoff around 50%. With these settings, the
> resonance is mainly boosting the fundamental, and as the oscillators beat
> against each other, slight clipping occurs at peak volumes. And then… It
> sounds as if the filter breaks into total self oscillation (which the synth
> can’t normally do) with extreme volume (wasn’t fun to find out the first
> time when playing with headphones as the ”weird oscillation” volume is
> much, much louder than the synth’s sound otherwise). The resultant sound
> sounds like this screaming farty buzz, almost comical. No key needs to be
> depressed for this sound to keep sustaining, it gives a rat’s ass about
> anything the VCA is doing. The pitch of the strange oscillation changes
> when changing filter cutoff frequency, and it ends if the filter cutoff
> frequency is changed considerably. Also turning down the resonance ends the
> weird oscillation, after which everything works normally. Decreasing the
> volumes of the oscillators to zero from the mixer section does not end the
> weird oscillation, but if I do so, I can hear the weird oscillation more
> cleanly.
>
> It seems that clipping caused at the filters creates the weird oscillation
> situation, at least if resonance is set high, and only square waves from
> two oscillators with filter resonance at maximum create loud enough a
> signal to reach clipping - or at least that’s my theory. And yes, filters
> in plural - Both channels of this two-voice synth are able to create the
> weird oscillation, either together or either just one of them alone. This
> rules out a defect filter IC, I believe (would have been nice and easy,
> although costly…).
>
> I only have a cheap old analog scope which is in a dire need of
> calibration and likely recapping (even the calibration waveform looks
> slightly off), but nonetheless I did some poking.
>
> The schematics of the Yamaha can be found here:
> https://manuals.fdiskc.com/flat/Yamaha%20CS-40M%20Service%20Manual%20(%20HI-RES%20).pdf
>
> I checked the weird oscillation on the FA card (VCF/VCA) at pins A13 and
> A23, where the signal enters the VCA and where it also leaves to the PB
> card (where the two voices are mixed together before output). To clarify,
> pins A13 and A23 send a pre-VCA signal to the final output amp board (both
> pre- and post-VCA signals are being sent to the PB card). The weird
> oscillation is about 15V peak-to-peak, and looks a bit like a clipped
> triangle wave with a fuzzy component in the lower bottom part.
>
> On pins A01 and A02 of the FA card the weird oscillation can only be seen
> if a key is depressed - these pins carry the post-VCA signal.
>
> Looking at FA’s +15V and -15V supplies, I see nothing suspicious either
> during normal operation nor during the weird oscillation state.
>
> I’m really puzzled with this. Both voices malfunction similarly, but what
> causes it? Please let me know if you have any idea on what to do next, I
> would appreciate any help a lot!
>
> Regards,
> Antti
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20210920/17368a6e/attachment.htm>