[sdiy] Vactrol-based ring mod?

JH. jhaible at debitel.net
Sat Aug 23 10:23:08 CEST 2008


>It's certainly possible to do ring mod with a vactrol, because I have used
>an audio modulated light bulb & a cadmium sulphide photoresistor to send
>(admittedly very low grade) audio across a room, many years ago.
>So it should be a doddle, with a led!
>
>True, the LDR lags will roll off the top end pretty fast, and the
>characteristics of the LDR vary in a complex way with prior exposure - but
>in a ring modulator, most of the weirdness will cancel out.


I haven't tried it yet (another thing I want to try for a long time :) ), 
but given that these modules exist and work, I think I can understand _why_ 
they work.

LDRs are normally sluggish, so even though you can somehow modulate them 
into audio range, I'd expect the modulation index to drop rapidly with 
increasing frequency. The result would be something between poor and, as you 
call it, weird - being much more on the "poor" side than the "weird", I 
suppose.

But Buchla has put them into servo loops, and _there_ the slowing down 
towards higher frequencies is compensated by the servo amplifier. The 
Vactrol's "filtering" will be compensated by the servo amplifier's gain, and 
it will _only_ start to degrade in performance when the "headroom" of the 
driving system is exceeded. Then the system becomes slew rate limited.

It's not unlike our ordinary opamps. Without negative feedback, they'd only 
have a low pass cutoff frequency of a couple of Hz.
With proper feedback, you can even process audio signals with a 741; in 
addition to using a lot of feedback, you just have to keep the amplitude 
small, in order not to run into the large signal speed limitations (slew 
rate).

The only difference between that opamp example and the Vactrol is that the 
opamp has the required high open-loop gain already built-in, while with the 
Vactrol you have to provide it externally (servo opamp).

The multiplication relies on the tracking of LDR pairs, as you always have 
one LDR in the servo loop, and the actually multiplying LDR outside the 
loop. I can imagine this brings a bit of weirdness in as well.

JH. 




More information about the Synth-diy mailing list