Last year I created an 8 voice bi-timbral polysynth called “A Bit More”. It was rather clumsily executed internally with a lot of interconnect between the controller/assigner and the voice board. It worked well enough but could have been better. So using some of the ideas and implementations in the later synth “Much More” going to redesign A bit more as a 12 voice bi-timbral synth. I will create two self contained 6 voice boards with dual AS3340 VCOs under the control of an RP2350 to manage modulation, PWM, autotune etc. A slight difference is that now I’m doing the two LFOs in software on each RP2350 to generate FM, PWM and output TM and AM from there 8 channel DAC for the filter and amplifier. So the only interconnect to the voice board will be power, midi and stereo output, again each board will have its own Spin FV-1 chip with 32 algorithms. Filters are AS3372 and envelopes/noise are Electric Druid Envgen8D and Noise2 chips as before. The voice boards will still be sandwiches of VCAs, filter logic, envelopes and filters on the bottom and VCOs, RP2350 on the top etc.
Craig oh Craig … you make us all look silly when we struggle to get “usable” signals from a 555 …
I’ve commented before and I am in awe of the complexity of your projects and the skill you have. You raise the bar definitely, not for us to give up and hopefor a painless end, but as an inspiration.
I could not have said it any better!
All the love and respect for your (craigyb) superbe creations.
They will probably mature into unique collector pieces of unobtainium…
Even I struggling get usable signals from 555
So far I’ve made a pair of VCO voices, twin AS3340 per voice, upgraded to RP2350 stamps and create dual LFOs within each stamp, LFO1 is more FM and also outputs to the DAC channels g & h so they can be used for filter modulation. I mimicked the operation of the Taplfo3 in software so it responds to my existing front panel for rate, waveform, multiplier etc. the second LFO2 is used purely for PWM and outputs from the DAC channels e & f, this gives me independent PWM for osc1 and osc2 with depths for each. This reduces the component count as I no longer have to build 4 LFOs and add the VCAs for level controls.The VCOs have a sub output from the triangle via a 4013. Also soft sync is included between the VCO.
I had a bad DAC in slot 2, voltages were too low, I think I have a DAC8568A which I’m reality can only reach 2.5v max swing as opposed to the C version which is can swing 5v.
Anyway I sorted that and listened to the VCOs, probably would have been able to hear them if I was a dog, so high pitched, set the tune to its lowest point and still around 2 khZ on the lowest note, basically I had forgotten how to set up an 3340 chip. I brought the the scale right down to around 45 hZ on lowest C for each VCO and then adjusted the autotune trimpot to allow the pulse to get through cleanly. Hit autontune and it actually went through the procedures. I heard the VCO tuning adjusting and jumping to the next octave and trying again. So it all seems to be working, just need to manually calibrate the tuning before handing it over to the Pico’s to maintain.
More progress, my eyesight is slowly returning now the allergies are subsiding a little. I’m starting to run out of parts now, orders have gone in for caps, IC sockets etc. and I’ve run out of RP2350 stamps, more coming from AliExpress soon I hope.
It’s taking shape, I past the the point of no return today by cutting the Teensy section of the mux/demux board off and creating a small midi board from it. Hopefully all the connections stay the same and it will just work.
Pair of boards sandwich, vcos, LFOs, PWM on the top with autotune and filters, envelopes, VCAs, noise source, effects etc on the bottom.
I moved the polymod offset trim pots to the top board as on the last build I had to drill 8 holes to get to the trimpots underneath to calibrate the tuning, this makes life a little easier. See the 6 trimpots top right. They just fix the ADC inputs at 1.65v
Have you ever experienced crosstalk with so many devices so close to one another and all the wiring in between them? And what do you do to prevent is?
Not really, the only real noticeable cross talk was recently on a build where DCO2s output was being picked up by the effects processor, If I had relocated the chip further away I could have fixed that, but it turned out I didn’t need the output of DCO2 going to that part of the circuit anyways so it got fixed.
I think the key is good grounding runs I make sure I create a sort of plane around the board with runs that join each end together creating channels. But you never really know until you build a board wether there will be an issue to resolve.
Big headache
I decided to cut my 8 voice boards down to 6 voice and rewire them from bi-timbral mode to standard 6 voice mode. The VCO board was relatively easy as each pair of VCOs is in a column, so chopping the first two out wasn’t too difficult.
The filter/eh board on the other hand is a nightmare, 4x 16 pin connectors carrying upper and lower CV lines, cut the first two voices out but leave enough room for the new midi to CV section down the left hand side.
Now I have to untangle that mess of wires.
You are making this up to fool us?
You are in a secret chat group where you try to outcompete each other with the most “out of space soldering”?
Your skill and ambition are exemplary. I feel love, admiration and a bit of envy.
Thanks, it was either this or create two new boards from scratch and I really loath throwing anything away.
Only me and the drones here…
First rule of secret chat group … 
I told you I knew where all the wires went. The original board was split into two 4 voice synths, now it’s combined as one 6 voice and hopefully everything is where it should be. The one loose wire is the feed to the noise inputs and I haven’t added the noise circuit yet, but I cleared space in the corner for it to go after removing the upper FX section and channeling everything through the remaining lower FX section.
The perfect picture for this year’s Christmas card and wrapping paper. Bravo!
Just awaiting a delivery of AS3340 VCOs from Electric Druid and then all 12 voices will be live. I had an earth missing on one board which meant all the EGs were floating. Then on the second board I had to troubleshoot a lot of mistakes and things in the earlier rewire. I wired both ADSRs up backwards to the DAC so it was DARS for control, quite confusing. And some others were on the wrong pins too. But it seems to be working now, but requires calibration when all chips arrive.
Still need to build an effects section for the new board and add noise sections to both boards.
AS3340 chips arrived today from Electric Druid, so it’s getting closer, I made up the new power cables and drilled the chassis out to accept the new boards, it’s a bit tight because I re-used the old 8 voice board and it’s bigger than the new 6 voice card and pushed everything up a bit.
More work today, I managed to build the effects unit and noise source onto the end of the board. It was a bit of a struggle, not enough room for two CD4504 level shifters, so I elevated the FV-1 chip rather than fix it to the board and fitted a single 4504 underneath and the 1 extra level shift I did with a voltage divider because it is 5v down to 3.3v. That meant I had to cut down a section of the VCO board so it didn’t touch the effects chip.
I’ve ran into some issues on the 8 voice converted to 6 voice board. During testing I noticed some chirping on VCO B of voice 6. I thought it was a tuning issue and it got worse as you went up the keyboard obviously as it was more sensitive to any issues at higher frequencies. I replaced or swapped everything to do with VCO B to no avail.
Then it dawned on me that other B oscillators were having the same issues to varying degrees which I probably hadn’t noticed and thought it was isolated to one VCO. So having replaced all of these parts of course I had to recalibrate the VCO and it was when I noticed that there was no chirping at all until I got VCO B to 1760 Hz the same as VCO A and it was hard to keep it stable on this frequency.
So I deducted that the VCOs were locking or interfering in higher notes. Now VCO A doesn’t suffer at all with this issue, but B does, it could be down to signal paths or proximity of the two chips in each voice.
On my purpose built 6 voice card which has a different layout I don’t have this issue at all, A and B are clean at all frequencies. My only issue with that board is I used 100k trim pots for the scale instead of 10k, so scaling them was tricky due to the large resistance, I will swap all 12 scale trim pots out and replace with 10k ones when they arrive.
When the voice card was being used as an 8 voice in the previous build I never had this chirping issue at all, so I’m confused as to why it’s happening now.