[sdiy] Choosing a DAC

[Tom Wiltshire]

On 7 Jul 2009, at 12:50, Csaba Zvekan wrote:

> Hi guys,
>
> This is really a highly interesting subject. Can somebody explain  
> what is exactly  happening after the AD558 DAC in the so called  
> "Glide circuitry".
> And why is the 8-bit DAC still capable of producing the 0-5V CV OUT  
> voltages pretty precisely ?

Because it's only got to be better than the VCOs!
Seriously though, tuning a Pro-One has a lot more to do with setting  
up the VCO Scale, Hi Trim and Offset presets, rather than getting the  
DAC to output the right voltages.
As long as the DAC is sufficiently linear, the reference voltage and  
any offset can be trimmed out. Even the Pro-One manual suggests that  
the initial control voltage at C0 isn't that significant (although it  
should be 0.083V).

> Especially the 3280 Glide OTA with them rectifying diodes on the  
> inputs  ? Could I use a CA3080 to experiment and replicate the  
> circuitry ?

I don't think those are rectifying diodes, since I doubt the input  
signal goes above the diode's forward voltage. Instead, I suspect  
they're linearising diodes like appear on the LM13700. I don't claim  
to understand the glide circuit, but I'd guess that the CA3280 is  
being used as a voltage-controlled resistor to provide simple RC  
smoothing of the CV. The mylar cap C141 on the CA3280 output would be  
the C in this picture. This is followed by an op-amp buffer to limit  
loading on this circuit. There's a current source to provide Iabc for  
the CA3280. Is this an exponential response to give the linear glide  
control a better range? Probably.

> Also how is DAC Scale being done with V SENSE and V SEL by feeding  
> the V OUT ? Could that be a VRef on another 8-bit type DAC? It  
> would be nice if somebody could explain a little further?

Sorry, don't know. The datasheet is at:
http://www.datasheetcatalog.org/datasheet/analogdevices/ 
263100547AD558_a.pdf

if you'd like to try and find out.

Regards,
Tom