I’ve never been able to get the LED indicators to work well in any of the System 700 circuits I’d made. Even with no input, the green light would stay on and horrific screechy noise would get injected into the output.
Vowing to finally get it sorted out I connected up my 703E VCF clone, got the expected horrible noise and stuck my oscilloscope on the +14.4V LED power line. It showed that the power line was oscillating at about 32MHz, at an amplitude of about 200mV.
Then I tried some stuff like bunging different capacitors on the LED power line and different power supplies, but nothing did the trick.
Some of the schematics suggest that CD4001AE is the right version of the IC to use, although confusingly, not all of them. Also all the pictures of the boards I’d seen have AE or UBE versions of the 4001 in place. Found some old RCA AE variant chips on eBay and gave them a go:
DAMMIT IT WORKS.
I wasn’t keen on relying on 35+ year old CMOS chips, so I ordered a few 4001UBE, and they work too.
Still going, just about. Not sure what I’m doing with that speed switch yet, whether to go faster or slower, hmm.
While I scratch my chin about how to reinforce these daft plastic panels, I’ve been enjoying reading through Fumitaka Ansai’s page on the System 700 – http://anz.yokohama/Museum/Roland/System-700/.
Assuming that Google Translate is accurate (my obsession with the System 700 isn’t quite going to stretch to learning Japanese) it’s interesting to find out that the original prototype had internal speakers, like the ARP2600! Also, it seems that users had been asking for a voltage-controlled envelope generator so they’d planned to release one but never got round to it, even in the later 100M.
I wonder if it was designed and never released? The owners manual for the 700 lists the modules, with gaps at 719, 722, and then 724 – 728, almost as if they intended to fill those gaps at some stage. Actually, maybe the prototype speakers were numbered 719.
Anyway, there are some fun demos on the anz.yokohama site too – if you have a look at the patch samples and then scroll down to Sample Chart 5, you can hear a nice example of using a voice to control a squealing oscillator with the 714 Interface. I’m half-tempted to build that one but it’s a big board, and I’m not sure if I’d just spend ten minutes laughing into it and then never touching it after.
Prompted by a comment on my System 100 VCA stripboard post, for as much use as it is here’s a very quick comparison of the original System 100 VCA and a clone System 100M VCA, built on stripboard with a BA662 clone for the OTA.
No envelope tweaking, just a bit of VCF variation from CV2 on the MC-4, and then full-gas into whichever VCA.
Here’s the stripboard clone 100M VCA, behold:
This is the (non-clone) System 100-only demo:
And this is the System 100 played through the clone 100M VCA (set on linear response), using the System 100 envelope as a CV source.
The System 100 has a lower output envelope than the 100M, so the 100M VCA is super-quiet when triggered from the 100 when set to exponential response. Flipping the VCA to linear gives you a usable audio level to play with.
The VCOs in my 100M clone are a couple of Curetronic boards, both using a heated CA3046 in place of the ua726. I’m not convinced the scaling is totally right on either VCO, so try and ignore that. Here’s a quick test of the VCO waveforms played through the System 100 VCF and VCA. You’ll hear:
System 100 square, low VCF, mid res
100M square, low VCF, mid res
System 100 sawtooth, low VCF, mid res
100M sawtooth, low VCF, mid res
System 100 PWM’d square, open VCF
100M square, open VCF
System 100 sawtooth, open VCF
100M sawtooth, open VCF
and if you can take it, here are some sweeps, 100 then 100m, in the order
sawtooth
square
PWM square
triangle wave
That’s too much excitement for now, I’m going for a lie down.
