Hey people,
in the digisound 80 VCO schematic and the raw 3340 chip datasheet schematics the audio outputs (triangle, saw, rect) are scaled to 0V-10V instead of ±5V-±10V (as proposed in the KOSMO format specification). Why? I thought that symmetrical audio is easy to handle with filters, mixers etc. or does every module have decoupled audio inputs?
It’s one of the ‘features’ of the AS3340, but not a very helpful one. Not all modules have AC coupled inputs, so the DC bias on the output of the VCO can be a problem.
I originally built my AS3340 VCO based on the Digisound 80-2 schematic but with op-amp output buffers to set all the output levels to 0-10V. Shortly afterwards I added a dongle board with 3 copies of this circuit -
Heyy again! Thanks for your answer! Finally I continue working on the VCO: Your circuit works flawlessly, if one neglect the bad cable connections from the insanely cheap breadboard lol.
I will include this circuit in my VCO! But I think, I still keep soldering points for the asymmetrical outputs…
do you recommend putting 100nF ceramic decoupling caps next to every TL07x chip btw.? or just for every module? Or is it sufficient to have the decoupling caps in the power supply?
And another question: Why do you add the 470 Ohm resistor at the output? Is it like a current limiter / protection?
Glad the circuit was useful! I think it’s a very good idea to have both the symmetrical and unipolar outputs and may well rebuild my VCO with a new front panel and all six outputs.
I use two 100nF capacitors for each op-amp chip, one on each of the power inputs. They should ideally be mounted as close as possible to the chip. I also include a 10uF capacitor on each power rail on all my modules. The 470R resistors are for current limiting in the case of accidental short circuits to either of the power rails. In the worst case such a short circuit would exceed the 1/4W rating of the resistor and you’ll see quite a lot of circuits that use 1k resistors to stay within that rating. I think 470R is OK as long as you don’t have any long duration short circuits.
So I made up a schematic for my VCO and extended it with the normalized & symmetric audio outputs - what do you think about it? My plan for the bottom row (the remaining opamps in the TL074 chips) is to include a sine-like output. I upload it to github soon!
That looks good to me. The only change I’d make is to use a separate offset trimmer for each of the outputs instead of your single RV4. The outputs from the U2A/B/C buffer amplifiers are unlikely to all be exactly 0-10V and you have to deal with resistor tolerances around U4 as well. Individual trimmers will ensure this isn’t a problem. If you’re going to include the 0-10V outputs in addition to ±5V, don’t forget to put 1k resistors between the outputs of U2A/B/C and the panel sockets.
Do you have a circuit for the sine shaper? I’ve used the shaper from this YuSynth circuit (U7d and associated components) successfully.
I got it working this morning. I tried a lot to calibrate and exchange some resistor values (Eddy Bergman’s circuit is vor ±15V), but I didn’t get very accurate - maybe, because I only have 50k trimmers laying around and again, its all on a cheap breadboard…
(its the lower part of the breadboard with the two trimmers)
I am still figuring out the YuSynth sineshaper, wow, it seems much simpler than the Eddy Bergman one! And I save two trimmers…I am going to try it too! By the way, is it allowed to share screenshots from the YuSynth schematic in here?
Bergman’s sine shaper is pretty similar to the one from the Thomas Henry VCO Maximus which I used in my VCO. (Originally designed for ±15 V but the link is to a ±12 V version.) I’m happy with it.
I wouldn’t worry about getting everything exactly centered on zero, a tiny DC offset isn’t going to affect much.
Your version probably works, but uses nine op amps, three to normalize the amplitude and six to offset them. I used a scheme adapted from the Kassutronics VCO 3340, in which the same three op amps simultaneously normalize and offset. You can see the whole schematic here but the relevant part is this:
It’s a good idea to reduce the OpAmp count xD. I already had a feeling that it could be like combining linear functions, but was overwhelmed by the idea to calculate all the values. I still have some questions regarding your schematic:
Are TRIANGLE_OUT_S, SAW_OUT_S, PULSE_OUT_S the 3340 signal outputs?
Are TRI, SAW, PULSE the normalized & offsetted ±5V outputs?
Why is there a cap sometimes included in the feedback loop?
It is probably bad to use approximate 20% resistors (eg. 10k for a 9.1k), right?
How do those transformation circuits work? Are they using like exponential diode / transistor amplification?
By the way, your VCO schematic is crazy impressive - thanks for sharing!
Stabilization against high frequency oscillatory behavior especially if the output is connected to a capacitive load. Kassutronics included it for the pulse and sawtooth outputs but not the triangle, presumably because those have sharp rising or falling edges corresponding to high frequency components while the triangle does not.
Well, for instance, substituting 10k for 9.1k would reduce outputs from ±5 V to ±4 V (I think). Whether that’s ‘bad’ is a value judgement. Anyway, resistors are cheap.
Is this about the tri to sine converter? I don’t know enough to explain it, I’m sure you can find an explanation online if you search.
Ok, thank you very much! Very interesting this cap in the feedback loop…
By the way I came across your YT channel, you have got nice vids! One sub more As I am still figuring out how I do the front panel, your experience is well appreciated! The differences between FR4 and Aluminium based PCBs are very interesting… I think, I will use raw 1.5mm aluminium plates and spray them black with pink UI elements through lasercut stencils, since I have access to a laser cutter