Getting the Midibox CV to work with Yamaha CS-series

[sonicwarrior]

Hi,

in his build report of the all-in-one Midibox CV PCB rvlt reported a problem when using Yamaha CS-series mono synthesizers (CS-40 and below).

I’m opening this as a new thread because else the posting may get lost without any notice:

the only synths I couldn’t get to work with mbcv were two yamahas (CS-5 and CS-40m). They have a “kind of” Hz/V behavior (CV doubles with every octave), but control voltage starts real low (C1 = 250mV, C2 = 500mV, C3 = 1V ..), whereas Korg synths start at C1=1V, C2=2V, C3=4V …

I don’t know, maybe the resolution of max525 is too low in that range? maybe someone could enlighten me..

cheers,

rvlt.

It does interest me, too, as I own a Yamaha CS-30 (but I haven’t built my Midibox CV yet).

According to sequencer.de the cv/gate data is:

CV in/out: 0,0835 V to 4 V (exponential)

Trigger in: Note off 15 V to 3 V / Note on 0 V to -10 V

Trigger out: Note off 3 V / Note on -7 V

So Gate Polarity should be set to negative and Output Curve is Hz/V.

As far as I understand the documentation the Hz/V curve has to be calibrated.

Or is it better to use the Oct/V output and a converter like this one:

http://home.swipnet.se/cfmd/synths/schematics/oct2hz.pdf

?

Almost two years ago I designed a PCB for this converter including the simple V-Trig to S-Trig converter

from the Doepfer homepage (see attachment).

So if it is really better to use such a converter and if there is enough interest here I will make

a PCB run of these (and previously test it as it as it currently untested). Q1/Q2 can alternatively be LM394 or SSM2210 which

are cheaper than the MAT02.

[bugfight]

… Q1/Q2 can alternatively be LM394 or SSM2210 which

are cheaper than the MAT02.

why use super-matched?  this is essentially an inverter, no?

check this:  http://www.retrosynth.com/docs/trigger/triggermod.gif

edit* oops i see you are doing expo too, thus the matched

[nebula]

I have built a MIDIbox CV and I haven’t had time to test it with my CS40M and CS15D.  I have messed around VCF’s and multimeters so far.  I’m away from home right now, but testing MIDIbox CV with my CS-synths (my only two monosynths) is on my short list of things to do when I get back … probably 1-2 weeks.

I will definitely be very interested doing whatever it takes to make it work.

[sonicwarrior]

edit* oops i see you are doing expo too, thus the matched

That is the main function of the PCB.

[rvlt]

although I found another solution (reed relays, see below), I had some time today to do some tests (midibox cv <> cs40m).

1. I double checked internal cv-calibration of cs40, but it was spot on

measured at cv-out:

C2 = 250mV

C3 = 500mV

C4 = 1V

C5 = 2V

2. I tuned the cs40 again, got a pretty stable tuning

3. I calibrated the mb-cv according to the manual (Hz/V, max setting = 10,24V, then fine-tuned the 1V-setting to get exactly 1000mV)

4. then I measured the cv-out at the interface:

C2 = 247mV

C3 = 498mV

C4 = 1V

C5 = 2.01V

I connected mbcv to the synth and, as one might expect, C2 & C3 were too low (about 40-50 cent) and C5 too high. a difference of 3mV doesn’t seem much, but it is really noticeable.

5. I tried messing around with the onboard trimmer, but the behavior always stayed the same: “low keys too low” and “high keys too high”.

6. I re-adjusted mbcv and connected my ms20 (starts at 1V, 2V, 4V, 8V), and tracking was perfect. hmm..??

7. I connected a friends’ kenton pro solo to the cs40, which has a “fine tune” and “scale adjust” parameter. finally it gave me a good tracking, it was off by a few cents, but that doesn’t matter..

so…

I think the best solution would be a software fine tune & scale adjust that would allow us to adjust the “steepness” (?) and offset of the output curve.

But I would really like to hear the results from other cs-users, maybe my cs40 is just funky and other synths work just fine.

====

in my case I found another solution: I had a midio128 (with reed-relays) installed in my juno 6. I put that in my cs40  and will install a mungo-enterprises interface instead. works perfectly, attached some pics..

====

@sonicwarrior:

if you decide to make the converter-pcb, I think I’ll take one or two, I could use that for other stuff.