Help With Function Generators and VCR's

Thread Starter

to3metalcan

Joined Jul 20, 2014
260
Hello, everyone! First time I'm posting a thread.

One of my back-burner projects for a while has been designing a simple resonant 2-pole voltage-controlled filter with an onboard low frequency oscillator.

I got pretty good (and kinda unique...sometimes reinventing the wheel pays off) results up to the point of it being a simple resonant 2-pole POTENTIOMETER-controlled filter, but I seem to stall out when it comes to designing LFO's or voltage-controlled resistances, so I'm looking for advice, recommendations of favorite circuits known to work, and general theoretical pointers...I like understanding what I'm doing!

First the LFO...design goals are:

1) Perfect waveshapes are not that important...squares with no rise-time, for example, actually don't sound too great (they tend to POP really hard at the edges.)

2) I'd like a pretty wide range of frequencies without the oscillator stalling, ranging from "whoa, I forgot that thing was on!" (about .1Hz) to low audio-range (say, 55hz.)

3) Pretty consistent amplitude, both between waveforms and across the frequency spectrum...I seemed to have a REALLY hard time with this when I was attempting it, before.

4) Ideally, it should be able to run on a single supply, possibly as low as 9V. Ideally, it should also not need any weird components, or draw huge amounts of current. Ideally, it should also make me breakfast and iron my clothes.

Feel free to disabuse me of my delusions if any of the above is cripplingly impractical, particularly with respect to...

Voltage-Controlled Resistance!

Basically, I have yet to read an article on this topic that makes much sense to me. (I've read a few manufacturer's papers and felt like I learned a lot about the physics involved but not necessarily much that would help me make design decisions.) I'm looking for practical, hands-on advice about how to implement a resistance that will swing in something APPROACHING a linear fashion with the output of the above function generator. The specific range isn't necessarily SO important, since the caps can be scaled to match it.

1) Again, no weird or exotic components, if it can be helped. I'm open-minded about FET, LDR, OTA, or whatever...but I don't want a solution that's likely to be obsolete or hard-to-find in the near future, or one that's likely to bug out under slightly imperfect conditions. (ex: I'm skittish about the future availability of OTA IC's...they seem to be discontinued left and right.)

2) This may be asking too much, but I'd like to get a useable range of predictable results without relying on a bunch of trimpots or pre-screening of components. I could live with a compromise on the former; the latter is for the birds!

Thanks a bunch in advance to anyone who weighs in on any part of the above!
 

Thread Starter

to3metalcan

Joined Jul 20, 2014
260
Sounds like it's got the high-range covered! I don't need to go anywhere NEAR 100khz. :) How about sub 1-second times? Like, ideally (how many times can I use that word, today?), I'd like a period of t=10 seconds at the far end of the response (ten is slightly arbitrary...more like t=several seconds.)

Also, any advice on integrators? I'd need to at least extract a triangle wave from this thing, too, across the same kind of range. I remember having some pretty deep pitfalls, trying to do that, before...
 

Dodgydave

Joined Jun 22, 2012
11,285
Yes 10 seconds no probs, will go down to minutes easy, as for triangle wave is available across the capacitor,using an op amp buffer.

You can use an op amp to generate both triangle and square waves, using the circuit on page 9,figure 15 with an LM358
 
Last edited:

Veracohr

Joined Jan 3, 2011
772
The LM13700 OTA datasheet shows how to use it as a VCR, and also has filter schematics. For control all you'd need is your voltage source and a resistor going to the bias inputs. I built voltage-controlled 4-pole resonant low-pass and 2-pole resonant state variable filters for audio, with triangle wave LFOs, for my senior project, and they sounded fine.

Alternatively, you could build filters based around the THAT VCA ICs, although their voltage control is exponential rather than linear as you said. I assume you're building this for audio, in which case I think you'd want exponential frequency control, not linear. The THAT chips are bit more expensive than the OTA, though.

Also, attached is the LFO circuit I used. It went from about 0.1Hz to 25Hz, if I recall. C8 actually is pointless, and I ended up taking it out.
 

Attachments

Thread Starter

to3metalcan

Joined Jul 20, 2014
260
I tried the classic op-amp function generator, and it seemed like I was having tuneability problems...as the frequency changed, it looked on the scope like my amplitude was fluctuating a lot. Would the 555 circuit be less prone to that?
 

Thread Starter

to3metalcan

Joined Jul 20, 2014
260
Veracohr, that's awesome! Thanks a lot...it's really nice to see an example of a generator that's actually been used the way I want to use it!
 
Top