[sdiy] SSM2164 Filter Cells

[Neil Johnson]

Hi,

David G. Dixon wrote: 
[snipped]
> > Also are there any words of wisdom that aren't in the data sheet or
> > obvious when designing with the 2164?
> 
> I can't think of any, but there are a few things in the datasheet worth
> heeding well.  The 7.5k impedance of the control voltage input, for
> instance, and the different modes of operation possible by putting different
> resistors on the mode pin.

As commented later, that should read "5k".

> As far as Aaron's comment about having to feed the output to virtual ground,
> just remember that the 2164 is a current source, and the best way to convert
> that to a voltage source is with

... a current-to-voltage converter, of which an op-amp is but one solution ...

> an opamp (where the 2164 takes the place of
> the input resistor in an inverting amp configuration).

> A resistor-to-ground and follower

... will *in theory* work, but not too well as the resistor will have to be very small to keep the output pin within the compliance limits.

For example, the maximum input current that can be fed in before distortion becomes too high is about 325uA.  Assuming maximum 0dB gain/attenuation the output current will also be 325uA.  The typical output compliance is +/-0.1V, so assuming you're prepared to push it to the limit that sets the maximum output load resistor to 307 ohms.  The voltage across this resistor would then need amplifying up to match the input levels.

> should also work just fine if you don't want to invert, but I'd
> probably always use the former technique, with an inverter after if you want
> to turn the signal the right way round again.

In the majority of cases the sign of the currents into and out of the 2164 makes life easier.

Neil
--
http://www.njohnson.co.uk