1222 Tuner VCO Module

I have three 1222 modules with AS3340 and each suffers from a very limited pulse width range (see osc images)
The voltage on pin 5 of TL072 ranges between 0V and 0.87V (the PWM pot fully open at 0Ohms) although I would expect 4V at the upper end.

I went through all troubleshooting advices, applied all suggested changes but wthout any finding or effect.

I also built the stripboard variant which provides full pulse width range with 0-4V at pin5.

I had a look at the schematics and obviously 1222 quite some different resistors for PWM than the stripboard version. Is this a known design issue?
I would be happy to get any suggestion to fix this.

1 Like

Well, I’m confused. The relevant part of the schematic:

With RV2 centered and no CV plugged in I find the voltage range at pin 5 should be 0.011 V to 1.04 V. In the stripboard version there is no equivalent to R10 or R13, and R11 and R12 are 100k. In that case you get much larger maximum voltage due to the absence of R13. Voltage range is 0 V to 7.2 V.

Per the datasheet, the voltage range for 0 to 100% width is 0 to 5 V,

So indeed it seems the PW setting is too limited in the 1222. Am I missing something? I thought maybe it was intentionally limited to avoid large frequency changes with the PWM/frequency problem, but there’s nothing limiting the modulation CV, so that doesn’t make much sense.

The voltages you report seem too low in both cases, but the design seems to be such that it shouldn’t be able to go a lot higher.

luckily it’ll only be a few adjustments to resistor values, even the R12 lowering that ill have a look this weekend and come back with some value adjustments :slight_smile:

7 Likes

That would be good, just about to start building V0.0.1 of my PCB’s :wink:

Rob

I was referring to the stripboard version “Square out with PWM CV”. My osc measured the full range between 0 and 4V. The pot had no “blind” range.

Yes, that’s what I referred to. I’d expect 0 to 6 or 7 V depending on the PWM CV pot setting.

Sorry, my fault. I measured voltage again on stripboard pin5:
0 - 5.9 V with PWM CV pot fully open
0 - 7.7 V with PWM CV pot fully closed
The full pulse width however is achieved with 0-4V

Repeating myself, but the PWM control range is 0−Vcc/3, see e.g. here. The values in the datasheet are for 15 V supplies.

2 Likes

Any news here? I would be happy about any suggestion

1 Like

apologies for the late reply on this!!! drowning in projects haha.

ok so I took a look this evening, the initial issue with the CEM chips is pitch drift when pulse width is adjusted it happens here and there and its a bit hard to predict in what situations it will happen, sometimes it does sometimes it doesn’t. in the 1222 it runs from 0%to about 33% duty cycle. from my understanding you want about 50 so then its the full amount.

issue with this was a slight pitch fluctuation. And in mk2 of the oscillator (currently sat with another 6 or so nearly finished prototypes, its a dual oscillator around 1 tuner) I went for an external comparator to make the square wave to make it not an issue

BUT!@!!! I just came across an idea that will sort it :smiley:
however I need a few guinnee pigs to see if this works for everyone!

ok so here goes. there is a reference image below. (this 1222 varies as its an old prototype hence it has bits hanging off it and chopped components etc)

FIRST swap R13 which is a 470k with a 200k, 220k is fine but 200k will get you closer to 50 percent duty cycle.

right that’s it, for the most part this is perfectly fine, in a set of 3 oscillators you won’t notice the minor minor pitch shift that happens when you adjust the pulse width

BUT I FOUND A PROPER PROPER DODGY BUT MAYBE QUITE A GOOD SOLUTION.

so it got me thinking maybe there is a way to counteract this externally to the CEM3340 reduce the pitch change well infact there is!!

put a 20M resistor (I didn’t have one so used 2 10M’s) between the two solder points circled in red.

what this does is it also sends the voltage going to the PWM pin on the 3340 to the pitch input too! just enough to counteract the pitch drop.

for some people thinking this may cause the pitch controls to bleed into the PWM yes it could do but it is so so small it isn’t noticeable, on my max zoom on my oscilloscope I did not see. but to be honest oscilloscopes are over rated on synths! like they say IF IT SOUNDS RIGHT IT IS RIGHT :D.

anyway! let me know if anyone dares trying this! I’m gunna continue testing it.

6 Likes

Glad to hear the Mk II is still progressing!

4 Likes

yeah! sorry, getting awfully bogged down with em all, im sure you know how it is!! as modules get more elaborate they get harder to get right. but I have been playing with the above and so pleased with this bodge job might omit the comparator pulse width adjuster completely.

but ill wait until someone else gives the above a go as I might just have got lucky on it.

1 Like

I soldered a 100k + 27k in series and put them in parallel to R13 to achieve 200k.
Then I put 2x1M 2x10M as you suggested. I used crocodile clamps, so tolerance might have been a bit larger.

The drift results my probe shows were:

  1. Original schematics: 3 Hz
  2. Use 200k for R13: 9Hz
  3. Use 200k for R13 and add the 2M 20M resistor: 6Hz

Listening to the drift, the 3Hz were tolerable, but 6Hz and 9Hz drift were quite much for my taste.
But the pulse range was much better, roughly 0-75%.

EDIT: Fixed resistor values

4 Likes

I don’t know if I can see well but it looks like it doesn’t match

1 Like

yeah and to be honest if you want to get uber fancy, try find a 5m preset pot ontop of some chunky M and M’s then you can exactly counteract the down drift with up drift. all in all you end up with drift wars.

but 20M seemed like a good number on a couple of my oscillators and made the drift somewhat unnoticeable even on the saw wave.

however at the same time im not sure wether the 6mhz with 2m is positive or negative drift. because that should be drifting the opposite way to your other readings with a 2m resistor. I tried a 2m at some point and it drifted up in pitch quite a lot, way more than that bend Gary Moore did at the start of whiskey in the jar

Sorry. I meant 2x10M.

1 Like

interesting i’m not completely sure what to suggest as this doesn’t match my findings on 2 oscillators. the drift has vanished. I’ll take another look tomorrow then. I even got out the oscilloscope to cross check with the ear holes.

IIRC the drift went into the opposite direction.

well if the drift went opposite direction. then there needs to be more resistance, ill give it a go on more of them in my synth tomorrow and see what happens.

ps I was quite pleased how this pretty dumb idea I had walking around actually worked surprisingly well to solving this problem… its not been an issue for me the pulse width before but its always cool to have a bit more range!

1 Like

I used 130Hz (C3) as ref, tested both pulse and saw (with ear and probe). RV2 centered no CV signal. I will spend some more time on this tomorrow.