[sdiy] Digital VCA - Choosing an ADC

[Matthew Smith]

Olivier, Tom, my thanks. I *think* I'm starting to understand this end 
of it.

With the input methods discussed, I am assuming there is no longer any 
need for a differential input ADC. That being the case, I get a LOT more 
parts to choose from. And probably cheaper. Having the simultaneous 
sampling/output probably isn't *that* important.

> To get something DC coupled is a bit more complicated, as a first
> guess I'd say connect the 5V rail to -ve input through a 2R resistor.

I think we'd have to be DC coupled - because one of those inputs (or 
possibly both, if someone decides to use the unit to modulate one 
control voltage with another) could have a signal so slow that it's as 
good as DC.

Negative voltages shouldn't really be an issue though - because DAC at 
the other end of all this will have a bipolar output and may well take a 
split supply.

> The only big problem I see with that is that you have no guarantee
> that a 0V input will fall exactly on word 2048 on the DAC -- if your
> mid-rail is not spot on. This might be a problem for a VCA - a 0V
> input could be translated into 2049 by the DAC, so you wouldn't
> totally silence the signal! This kind of trimming could be done in
> software, at no hardware cost.

Note to self: include calibration routine in the firmware. Put a pair of 
shorting jacks in the inputs, select 'calibrate' from the menu, and off 
we go. (User interface will be a rotary encoder combined with a push 
switch and an LCD. Might include some pots to tweak other parameters, if 
I feel the need. Hey, extra ADCs are cheap. Cheaper than pots, in fact.)

Cheers

M

-- 
Matthew Smith

Business: http://www.smiffytech.com
Blog:     http://www.smiffysplace.com
Linkedin: http://www.linkedin.com/in/smiffy
Flickr:   http://www.flickr.com/photos/msmiffy
Twitter:  http://twitter.com/smiffy