Fast analog sw. as VC resistor

[Mark Smart]

I'm brainstorming about my GR-300 modification project again, and I have a
question.

In order to do what I want to do (modify GR-300 guitar synth for greater
sound flexibility and programmability), I need a way to create a
voltage-controlled resistor. Looking in the synth-diy archives, I see that
this has been discussed in the past. Ways which were suggested are:

1. OTA (LM13600 has floating vc resistor circ in data book)
2. LED-photoresistor pairs
3. Fast-clocked analog switches (way higher than audio range)

If I build what I want to build, there need to be sets of 6 VC resistors,
or in some cases pots, which all need to be controlled by the same voltage.
Otherwise I have to have 6 sets of pots to do everything!! So there need to
be 24 VC resistors just for the polyphonic envelope generators.

I saw the OTA circuit in the LM13600 data book, and there are also a couple
of circuits using the CA3080 in my Electronotes collection. These circuits
all look like they would be too complicated to do this, especially if you
want to make a pot. It would work, it's just that the circuits would be
hugely monstrous. For a pot you'd need like 4 OTA's, 4 op-amps, and a bunch
of other stuff.

Over a year ago, Scott Gravenhorst posted on here asking about the high-
frequency analog switch method. No one really responded. I'm basically asking 
the same question again...how would you do this?

The original post said that varying the pulse-width of the high-frequency
control line on an analog switch would vary its resistance, and that the
inversion of the clock signal could be used for another switch if you
wanted a pot-type circuit.

Intuitively, this seems like it would work...is the resistance of the
switch just equal to the on-resistance (on the order of 200 ohms)
multiplied by the 1/(fraction of the cycle that the switch is on)?

i.e.  200 ohms on-resistance x (1/0.5)= 400 ohms for 50% duty cycle?

I'm just wondering if it's more complicated than that. Does the resistor's
output have to be filtered to get rid of the high-frequency noise?

If this is as simple as it seems, it would save me a LOT of work on this
project. A master sawtooth oscillator could be sent to a whole bunch of
comparators, one for each VCR on a channel. The comparator outputs would be
used to switch all the analog switches. To make a pot, two different types
of analog switch chip could be used, with complementary clock polarities.

What are the potential problems with this approach? 

Thanks ahead of time!

************************************************
*     Mark Smart                               *
*     Network Technician                       *
*     University Communications Inc. (UCI)     *
*     smart at medusa.nn.com                      *
************************************************