just to let you know! rob over at molten modular is building his first kosmo stuff today 8pm GMT+1 I think
thats the link, he says this is the first project he has had to source the components and stuff, ill pop on the stream at some point im not sure when yet! to chat about it. maybe if you have some insight pop over! 
LIVE NOW
,ā¦
v;sd,v;ldamfv;lb,adf
didnāt catch it live but it was fun to build along to . maybe do a group build some weekend ?
I would totally be up for a group build/hangout. Tough to match up schedules, but if we donāt schedule anything, nothingās going to happen.
Im sure there are groups that can get together.
If people interested in cooperative streams post here saying when theyāre available (from, to), additionally giving time zone and daylight savings switch date if any, it will encourage matching.
sounds like a a new thread set up for something like this to happen ?
That seems wise for reusability.
During that stream, someone mentioned a 79L05 -5V regulator being used in the Kassutronics VCO
@lookmumnocomputer started looking for it in the schematic, and I got curious too, but as it seems to be only used as the 3340ās VEE negative power rail, (and the schematic is sideways in the pdf found on the Internet) we didnāt find it immediately so Sam didnāt get to comment on it.
The buit-in Zener diode shown in the CEM3340 datasheet and relied on in Samās VCO design is fine, but does anybody know if using an active regulator could help with the issue of the pulse width affecting the frequency?
The Kassutronics documentation only says: āthis module uses onboard voltage regulators to reduce sensitivity to power supply fluctuations and limit self-heating of the VCO chipā, so it doesnāt seem that the pulse width issue was a factor in choosing to use the 79L05.
I built the Kassutronics module, but I have no idea. I tend to leave the PW in the middle, so I havenāt noticed the issue on the Kosmo VCO either. Would I be able to tell the tune drifting if I dial it to the extremes?
Iāve believe Iāve read somewhere that using a regulator helps with the PW problem, but I donāt know if thatās well substantiated.
Iāll be fat box building on stream tomorrow, btw!
I havenāt witnessed the effect myself (I havenāt looked for it) but, yeah, that is my undestanding: as you change the pulse width towards an extreme, the oscillator frequency changes, even for the other waveforms, so it should be pretty easy to hear, if you listen to only the sawtooth or triangle while twisting the pulse width knob.
Iām chipping in here as pwm as an effect can affect perception of frequency. My physics is well dusty but I believe the pitch change you hear is similar to the Doppler effect and is a matter of observational ābiasā (ok so now im winging it while making a plea to the physics crew out there that always swoop to my rescue when I overreach like this ā¦ guysā¦any time nowā¦any secondā¦ you wait and seeā¦ confirmation is but a moment awayā¦ oh itās Sunday, so soon ā¦)
Mine arenāt built yet, but for what I have read the frequency really changes.
If you just listen to the triangle out, change PWM width, and the freq of the triangle changes.
As for pitch perception, when you change the PWM width, you change the harmonics content of the signal, this can (will ?) change your perception of the pitch.
Frequency only applies to a periodical signal, which most of traditional instruments DONāT produce.
Pitch is perceptual by definition/necessity.
I can test this in a builders stream later in the week, I dont think it will help as it feeds the same zener diode. but who knows.
in the second version (got a dual cem3340 oscillator Sith 1 tuner and wideness instead) itāll have comparators doing the PWM so the square wave in the chip isnāt touched:D.
as for the -5v in the kassutronics, a lot of Roland gear does that. truth be told I didnāt bother as it added more components, as in 4 extra components to bypass 1 resistor didnt make sense to me.
edit. might be more the actual capacitor wired to the negative supply. plonking a larger like 470n onto it even. that might be worth a go for Pwm drift. ill devote a builders livestream to this issue within the week we can try some stuff out 
Itās a 6.5V Zener, so if you feed a regulated voltage that is less than 6.5V, the Zener is essentially inactive and itās like it wasnāt there.
Regarding the 3340ās decoupling capacitor, in the #1222, the capacitor is at the wrong end of the resistor.
It should be directly at the VEE pin.
(I know Iām repeating myself, but I wasnāt sure if you had seen my previous mention of it.)