AW: Re: Progress on ASM-1

Haible_Juergen#Tel2743 HJ2743 at denbgm3xm.scnn1.msmgate.m30x.nbg.scn.de
Thu Aug 8 21:38:00 CEST 1996


>     As the LFO rate pot is turned almost all the way down (attenuating the 

>     square wave down to almost nothing to the input of the integrating
>     op-amp), the LFO frequency will slow down towards zero hertz. Once the 

>     integrator's input is at ground, the LFO will stop and the triangle
>     output will probably slowly drift to either the (+) or (-) supply. The 

>     minimum frequency before this happens will be determined by a couple
>     things - first, the resolution of the rate pot near the lowest
>     setting, and second, the leakage currents of the TL-082. If you aren't 

>     using a TL-082 here, I suggest that you do because something like an
>     LM358 or 1458 will not go to nearly as low a frequency before
>     "stalling". Also remember to de-flux the circuit board because solder
>     flux residue will cause leakage currents and prevent really low
>     frequencies.

For a 1458, leakage (input current) would indeed be the limiting factor.
But not for the TLo82.
There is another effect, and that's the offset voltage of the opamp that
forms the integrator! Asume you have -1mV offset voltage, then positive
voltages below 1mV will not be able to change the integration direction
anymore! You can easily ckeck if offset voltage is the bad guy:
with decreasing frequency the rectangle would become more and
more unsymetrical, until you reach the 0% or 100% point and oscillation
stops.
In my modular, I have added an offset compensation to the LFOs.
Then I could use 10k log pots together with a very small "foot" resistor
(I think it was below one ohm!), and the LFO still oscillates.

(This is about general TRi/rectangle LFO's - I don't have Gene's
schematics right in front of me now, so I am not sure if it applies to this)

JH.



More information about the Synth-diy mailing list