Linear current source

Haible_Juergen#Tel2743 HJ2743 at denbgm3xm.scnn1.msmgate.m30x.nbg.scn.de
Fri Oct 25 20:46:00 CEST 1996


>   They say the "Howland current source" is quite imperfect in the body =

What is a "Howland" current source? Is this that one-opamp circuit
that has a current sensing resistor and positive as well as negative
feedback? Where you have to match resistors to get a high
output impedance? Wouldn't exactly use *that* on OTAs.
Funny, Yamaha used something similar to control the current
thru the LED of a photocoupler in their CS-synths ...


>Driveing OTAs has become a recent little problem of mine as well...
>In the ASM-1 there is a op-amp and two diodes that drives an CA3080 with =
>linear
>control. I haven't put that schematic on the web.... yet!

If memory serves, this is the famous electronotes current source ...
it's reported to work very well (haven't tried it myself). It has one
severe limitation: Output swing is near zero in the negative range,
so you would only use it to drive virtual gnd summing nodes.
(Hey, or to emulate the SSM2040 behaviour ... now *that's* an
idea !!)

Personally I use the one-opamp-one-pnp-method all the time.
(as in  http://www.synthfool.com/hjvcaod.gif  )
It works nearly perfect. It has a slight scale error of about 1/beta
which doesn't matter at all, especially if you use a pnp with a
high beta (500 is no problem).
Another error is the offset voltage of the opamp, but with 1mV
its 0.02 % on a 5V CV range, and van be trimmed completely
to zero with a trimpot, or - even more handy, it can be set to the save
side (positive) so that at worst you have a 1mV threshold (of CV!)
before the VCA starts to conduct. Nearly perfect. And simple as it
can be.

JH.



More information about the Synth-diy mailing list