Fast analog sw. as VC resistor

[Mark Smart]

> ...uhhh, please ignore my last post - I see it's already been covered
> <mental note: read all email thoughly before replying :)>
> 
> I would, however, still like to know what the frequency of your sawtooth
> was...
> I guess it was less than 10Khz?
> 

In my first experiment, with the 1458 for the comparator, I ran the function
generator at 10 KHz with a triangle output (this function gen. won't do
sawtooth). I found an LM339 yesterday and played around with it at 100 KHz.
This thing doesn't work as well as I had hoped. The output resistance is
pretty obviously non-linear, even at 10 KHz. Before when I said it was very
linear, I meant in theory! Graphing output resistance of one side vs. clock 
duty cycle, it looks like:

         100K              .
                        .
                     .
output             .
resistance       .
               .
              .
         130 .
             0%             100%    
              pulse width

The total resistance of the pot varies from 97K to about 103K as you
sweep the duty cyle.

I don't know what is causing the non-linearites. Some possibilities:
capacitance of 4053 clock input? switching speed of 4053? Comparator
speed? It might be possible to use the curvature to my advantage, since
the env. gen. schem's that I have call for log pots. The nonlinearity
gets worse as the clock speed increases. At 1 Mhz or 10 Mhz the thing
goes totally haywire.

Now that I think of it, I don't think the comparator is the problem.
Its output looks pretty good even when the resistance curve is totally
weird.

This is probably still usable for envelope generator control. I hooked the
output pot up as a voltage divider between GND and 12 V (the supply), and
sticking a 0.1 uF cap between the wiper and GND was sufficient to flatten
it out. I still haven't tried running an audio signal through it
yet...that's next!

I did try using the pot to vary the decay time of a simple RC decay
circuit. This seems to work fine even without any filter cap at all.

> > I'm going to get some LM339's and try them.
> 
> Reminder:  LM339's can SINK current, but cannot SOURCE it. You need to use
> a pull-up resistor on the output. I went nuts the first time I used one of
> these chips until I realized what was going on...
> 

Yeah, I figured that out after it didn't do anything the first time.
I used a 3.3K resistor. Would varying this help?

> It also might help to have a little positive feedback (like a resistor in
> the 560K to 1M range between the output and +in) on the comparator to speed
> it up.
> 

That's a helpful idea...I'll try it.

> A slightly better general purpose op-amp - like a TL084 or LF412 should be
> able to handle stuff up to at least 50 or 100kHz no problem - right?
> Actually, an LM358 would be better for driving the CMOS switch - how fast
> can they go?
> 
> I would think that the higher the frequency the better - because it would
> be easier to filter out - yes?
> 
> - CList

If the nonlinearities could be corrected, that would be the way to go. As
it is, it's not usable above 100KHz.

I would appreciate suggestions on what could be causing the nonlinearities
and what could be done to correct them.

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