Fast analog sw. as VC resistor

brad sanders radioactive at geocities.com
Thu Jan 9 07:32:40 CET 1997


>The ARP envelope generator circuit I was planning on using has
>floating three-terminal pots on both the sustain and release function, as I
>mentioned in an earlier post.  Perhaps the sustain could be replaced with a
>reference voltage; I'll have to study the circuit some more to see if this can
>be done.

I see alot of what I mentioned has already been answered; when I
logged in this morning I sent that message before having time to pore
over the 40 or so messages in my inbox!

>> bigger limitation, 'tho, is the CMOS switch. Something better, like a
>> DG series, would help much here.
>
>I saw those in the Analog Devices data book I just got. How much do 
>they cost? 

Maxim has'em for about 3-5 bucks each. Visit their site and have'em
send you a couple!

>I haven't seen any prices yet and would be interested
>to try them out. I assume they aren't $80 apiece like some of the analog
>multiplier chips!

Yikes! I had no idea multiplier chips were that expensive. I guess
that's why Frank Burge ran that article on using DACs and A/Ds to do
multiplication. 

I can't resist pointing out here that a VC resistor could just as
easily be used to make a two quadrant balanced modulator. Never mind
that a switch is itself a multiplier and capable of multiplying any
wave by a square wave - if you use one of these  VC resistors in the
feedback loop of an opamp you could use it for any two arbritrary
waveforms! I dunno what kind of error and offset values you could get,
but for basic "bells and whistles" it seems it might work...

>I'm interested to know more about that. What's glitch energy?

Energy from glitches, of course! When one of those switches changes
state it briefly "shorts" (however, this "feature" is irrelevant in
this context). The junction capacitance causes feedthrough, and the
gates driving the switches themselves have feedthrough. Also, the
gates themselves have a finite switching time. Some devices are
optimized to minimize these transients (these might be best for your
VCF, or that balanced modulator I mentioned). 

Also, remember that in a balanced mod or VCF app, that HF switching
signal is going to itself cause sum and difference products. You may
need an antialiasing filter on the output, and definitely should
consider this when choosing opamps for that VCF (ie make sure the
opamp can handle these high frequencies with fairly low IMD; choose
one with decent bandwidth). 

If you want to play with DACs, you could also try "multiplying DACs."
Maxim has two quadrant multiplying DACs - four in a package - for
about 7-8 bucks. These essentially ARE "three terminal pots." Not all
have a floating ground node, but many do - and in the case of a
divider, it's irrelevant anyway: you just need a wiper and a CV in.

I haven't seen anyone mention'em here, but Maxim is a very "hacker
friendly" company. So long as you have a company and a position of
somerelevance to engineering, two samples of any of their products is
just a phone call away. From their web space you can even request
multiple devices VERY easily. I think they also have pricing info
online - and you can even place orders (uh - I think).

www.maxim-ic.com

Disclaimer: just a satisfied customer!



> http://radioactive.home.ml.org <-------------------------------



More information about the Synth-diy mailing list