[sdiy] Which opamp pins wiring is best for CV out ?? A , B or C ???

David G. Dixon dixon at interchange.ubc.ca
Tue Nov 9 17:59:25 CET 2010


OK, so what y'all seem to be saying is that the small output resistor is
critical for neutralizing capacitive loads.  Fair enough.  I'm glad we
finally reached a consensus on this issue, which we have been arguing about,
off and on, for the last year!

On that basis, I did some more simulations, assuming a 560pF capacitive
(Neil's 4m cable at 140pF/m) load, and found that the ideal resistor value
is 220R.  Assuming a 100k input impedance downstream, this would cause a CV
error of 0.22%, or about the same as typical tuning errors in VCOs.

Also, it bears noting that this small resistor would do very little to
protect the opamp from current surges, a rail-to-ground short-circuit would
generate about 60mA through this resistor, which is 2 or 3 times higher than
most opamps allow.


> > hmmm... when that input is on the end of a shielded cable. I've had LOTS
> of TL07x (08X)
> > opamps go NUTS when driving even a reasonably short cable, if there is
> no series resistance.
> > Even 10 ohms is enough to decouple the capacitive load...
> 
> Yes indeed, those darn cables!  If we imagine a typical example:
> output goes to a connector on the front panel.  This then goes into a
> cable, lets assume shielded cable for those who like to use jacks.
> Now I've seen cables a good couple meters long on modular systems
> here.  Then that plugs into something else, which might be a passive
> multi, that then breaks out to a couple more one-metre cables.  So,
> load capacitance?  Looking at the VanDamme catalogue I see something
> of the order of 100-160pF/m.  How many metres?  Lets say 4 (2+1+1).
> So that's 400-640pF.
> 
> That's a lot of capacitance for a lowly op-amp to drive without going
> "NUTS".




More information about the Synth-diy mailing list