In regards to the cem-3340-diy-simple VCO circuit, what is the safe V/OCT CV range (+/-V) going through the 100K resistor to the freq cv summing node on pin 15. I’m experimenting with the 3340 and I’d like to not accidentally fry it.
Thank you.
The IC input is current driven, and sums up the currents from multiple voltage sources (CV inputs, tuning knobs, octave selectors, …). As long as you stay within reason (e.g. the supply rails) you cannot really overdrive it.
See here for the relation between voltage and frequency. Note that you need more than 12 V in total to get beyond 3100 Hz (~G7).
2 Likes
Thank you. I saw your post with the freq->cv equation. Staying within the rails (+/-12) is just the answer I was looking for. I was looking to put in an alternate (selectable) timing cap (~0.0125uF) to get LFO ranges as well as the Audio ranges. I’ve been working through the 3340 spec sheets and permissible CV range was the only thing I was unsure of. Going negative would be required to get the really low end knowing full well that the accurate V/OCT is limited based on the current range of the multiplier.
Thank you again … time to break out the bread boards!
2 Likes
Oh, I was thinking more in the other direction. Not sure what happens if the sum goes negative, TBH, but I doubt a few volts would damage anything, so “don’t be scared to try it” might apply here.
Couple things to take into account:
-
The negative supply is roughly −7 V, not −12 (assuming you use the usual resistor approach), so going below −7 V might cause trouble. But then you’re seven octaves below ~1 Hz which is pretty slow as LFOs go
-
The absolute max rating section says the voltage between frequency pin and ground can be max ±6 V but that pin is a summing node that’s held near 0 V, so that just limits the temperature compensation range iiuc, not the CVs before the 100k resistors (otherwise you wouldn’t get above 50 Hz or so…)
3 Likes
Yeah, it was the summing node held near 0V combined with the max +/-6V that was confusing me. There is another version of the schematics that uses -5V for Vee, skips the Ree resistor (since it’s less than 7.2V), and limits the thermal impact. There are also other 3340 based modules (that I don’t have schematics for) that allow -6.5V CV but I wasn’t sure if that was using an external op-amp to scale prior to the current conversion. It’ll take some experimentation but you’ve reassured me that I won’t fry anything under normal limits. I haven’t messed with circuit design in ages so this a bit of a re-learning experience.
Thank you again
1 Like
Yeah, a lot of commercial builds have a −5 V rail already, possibly for other purposes, or use a local 79L05 or equivalent to get a local low-impedance supply. I’ve considered using that in my own builds, but have been too lazy to redo the PCBs… should maybe try it some day.
I’ll give the specs another read while I wait for my next batch of parts to be delivered. I’m starting to wonder if the Tequila mule I was drinking had an impact on my comprehension of the specs 
I’ll post the results (of the circuit, not the drinking) once I get some parts to play with. The goal is a reasonably trackable VCO/LFO based on a 3340. Really the only changes to the reference schematics is changing Cf for a lower V/OCT tracking range and maybe change the freq control pots.
3 Likes
Tinkering with the capacitor sounds like a safer bet than pushing the CV into strange regions (frequency depends on 1/Cf so should be fairly predictable, at least if you hand-select some nice values). Hopefully you won’t get too much noise or random capacitances from the wiring. Worst case you can build a few more 
1 Like
I’m thinking about 2 or 3 different ranges by choosing (switch) one of three caps.
Based on the stable (V/OCT trackable) multiplier currents (Ieg) at 50nA to 100uA
f = 3 * Ieg / (2 * Vcc * Cf)
Cf = 1nF -> 6Hz to 12.5kHz (reference value)
Cf = 0.0125uF -> 0.5Hz to 1kHz
Cf = 0.125uF -> 0.05Hz to 100Hz
Going even -3V for CV would yield even lower rates but I’ll need to experiment a bit. V/OCT tracking isn’t as important for low LFO rates so I was hoping to stick with the 0.0125uF to keep stable tracking in the low audio range and then just go negative on the CV to get the really low rates … ideally down to about a 120s period.