[sdiy] Phasors FET vs OTA?

patchell patchell at silcom.com
Tue Aug 6 04:43:25 CEST 2002


    I don't think it matters how you use OTA's, voltage controlled resistors or
not, if they are controlling the signal characteristics, they will introduce
noise.

    Some things you need to keep in mind about noise.  If you increase the
number of stages by 4, you only increase the amount of noise by 2, so adding
stages is not quite the big deal you may think (noise addes by the square root
sum of the squares, or something like that).  Also, the offset voltage on the
CA3280 is pretty good, you might be able to get away with quite a few stages
before you need offset trim.

    Also, on FETs, you can improve the voltage control characteristic of the fet
by using feed back around the fet.  This also improves the distortion of the FET
Voltage controlled resistor as well (talking about J-FETs here).  In fact,
didn't somebody just post a bunch of schematics of experimental JFET VCResistors
just a little while ago?  I was impressed with some of the novel solutions.

Jay Schwichtenberg wrote:

> Howdy One and All,
>
> I've been thinking lately about doing some work with phasors, at least 6
> stages maybe more. In general I'm more into quiet vs freq tracking for this
> type of device. After doing some research it seems to me there are about 4
> different general architectures for phasor circuits. FET resistors, OTA all
> pass, transistor ladders (Arp Omni) and LDRs. I'll ignore LDRs and
> transistor ladders simply because of size (I'm into small and SMT) so that
> leaves FETs and OTAs. In my mind (believe me it's a scary place) there are
> issues with both. With a number of OTA stages I think there would be issues
> of noise (OTAs aren't the quietest parts) and offset unless diode biasing
> (LM13X00 & CA3280) and offset trim was used. Using the diodes wouldn't be a
> problem but by the time you 6 or more trim pots on a board you've sucked up
> a lot of space. Using a FET to ground at the op-amp input starts to seem
> better noise wise. The major problem that I can think of with the FETs is
> the control/freq range. A lot of FETs have a very small gate voltage for
> full off/on operation some in the order of 0.1-0.2 volts. You can find other
> FETs that are more in the volt range which might be better suited for this
> type of circuit. Some other ideas for discussion might be AC coupling stages
> and using OTAs as VC resistors like the FET but not directly in the audio
> path.
>
> I know that there are people out there that have some ideas and opinions.
>
> Thanks for the discussion everyone.
> Jay

--
 -Jim
------------------------------------------------
* Visit:http://www.silcom.com/~patchell/
*-----------------------------------------------
*I'm sure glad Merry Christmas comes just once a year
* -Yogi Yorgensen
------------------------------------------------





More information about the Synth-diy mailing list