Yes, I have read a lot more of the manual than just the build instructions; I am exactly the kind of electronics n00b these kits are aimed at, and it's been very informative.
I agree it does sound like the problem described on page 10 - but the very next page then goes on to describe exactly how that problem is "fixed" in the final schematic! Then goes on to give no troubleshooting hints for if the final assembled module still has this problem. So the final circuit shouldn't be exhibiting such issues. I ran the module in loop mode with a multimeter attached to ground & +12V contacts on the case's PSU board, and could see the 12V rail swing up to 12.02V as the envelope opened - so not a big swing, but it doesn't take a big swing to produce the amount of difference in the tuning I'm talking about.
blassner wrote:I hearing an LFO type modulation on my VCO signal whenever I turn on the loop function on the envelope. However, this happens even if I don't have any patch cables connected to the module. The module can just be sitting in the rack powered up without anything connected and it will start outputting an audible signal as soon as I turn on the loop switch.
I built the breadboard first and assembled 2 kits. I didn't seem to have any issues with the breadboard schematics or the PCB builds. The effect seems to occur on either my built modules. I even tried removing the VCA and the other envelope module, so I just have the VCO and one envelope in my case. Even with just a single patch cable running the VCO saw tooth directly into a speaker, whenever I turn on the loop switch I start to hear an LFO tremolo.
Having spent some more time with my units today, I think this is precisely one of the two issues I'm experiencing. So it's not just me. (My other issue being how long it takes the output to drop to zero even with release turned fully CCW.)
However! Going back to the manual again, and mocking up the entire circuit (not just specific sub-circuits) in CircuitJS, I've managed to fix one of the problems!
On page 12, a pull-down resistor is introduced on the output of the gate's comparator, to stop the voltage floating when the input is below the threshold.
- pulldown.jpg (23.62 KiB) Viewed 315 times
Then, on page 15, we introduce the sustain pot - set up as a variable voltage divider, but it is still, at a fundamental level, a (variable) resistor sat in between the gate comparator's output and ground.
- pulldown-plus-variable.jpg (36.16 KiB) Viewed 315 times
So, does the first resistor here still serve a purpose? Quoting page 16:
Our input starts out low. This means that both our comparators’ outputs sit at –12 V. But because of the two diodes, this doesn’t propagate, and so our buffer’s input gets pulled down to 0 V through the 100k resistor and potentiometer. Giving us 0 V at the envelope’s output as well. Next, let’s assume that the input goes high. This will do two things: we’ll get a voltage spike after our high pass, which gets converted into a short 12 V burst by the top comparator. Simultaneously, the other comparator pushes out a constant 12 V that get scaled down by our sustain potentiometer. Let’s assume we’ve set it to about 50 %. This means that at the buffer’s input, we’ve got our 12 V burst coming from the top, and a constant 6 V coming from the bottom comparator. Since there’s no resistor in the top path, but a 100k in the bottom one, the burst will „win“ and push the overall voltage at the buffer’s input up to about 12 V.
Thinking about it... the potentiometer connected to ground itself already serves the function of a pull-down resistor, pulling the buffer's input low. So I guess its only function now is stopping the current from the top comparator flowing straight through the pot to ground, in preference to entering the buffer, right?
- preferred-path.jpg (34.2 KiB) Viewed 315 times
So... what happens if we just get rid of it? Well, I tried it in CircuitJS, and everything functioned as before - except there's now less voltage drop between the buffer's input and ground. In the final schematic, this resistor ends up designated R22. From page 42:
- schematic.jpg (22.05 KiB) Viewed 315 times
I soldered a bodge wire across R22 on one of my boards, and, lo and behold, not only does it still work, it now doesn't cause any measurable voltage change on the 12V rails during the attack stage, and - crucially - no more tuning change in the adjacent VCO, getting rid of the unwanted pitch LFO effect in loop mode!
Unfortunately, I still get tuning changes on long, sustained notes - but only when the sustain knob is set, roughly, 60% clockwise or more... which is easier to deal with, as I find that's about the position at which the decay stage starts to get drowned out by the sustain anyway, and you can still get fast, pop-free attacks with the right twiddling.
So, removing R22 is
a fix, but I suspect not
the fix. Does the 12V input coming in through R7 need to be buffered, or something? I really don't know, and this is already far more practical electrical "engineering" than I've ever done before
I have no idea what I'm doing.
As for the long, slow drops to zero... I'm wondering if that's simply how this circuit behaves? I
think the same behaviour seems to be present in CircuitJS, but it's hard to judge because it runs much more slowly than real-time. Anyway, with a loop mode that doesn't cause tuning issues, and a gate/trigger mode where the issues can be worked around, I'm feeling a lot happier now.
I've made this modification to both my kits.
I put some electrical tape over it after taking the picture... and yes, I have some contact cleaner on the way
Edit to add CircuitJS links
- Original circuit
- Circuit minus R22
- Hypothetical bonus mod - I wonder if changing C1 from 1uF to 3.3uF would give a bigger sweet spot for attack in gated mode? I don't have any spare caps to actually try this out, but this increases the length of the 12V pulse going into the buffer between gate/sustain & ADSR, causing it to rise higher above the sustain level over a wider range of attack times. Turn off the loop switch, toggle the gate input and watch the output. Would love to see someone try it out!
Also, for anyone else trying to troubleshoot issues, these should give you a better idea of what you should expect to see at the three test points.
as if the EG is somehow drawing too much power and causing a voltage or current drop on the VCO!? In case it was an issue with the PSU in the case (VCO & EGs were both hooked up to the edu DIY case) I hooked one up to my main rack, and got the same sustain problem.
