[sdiy] Polivoks filter

[Dave Brown]

I just finished hand building a Polivoks filter using the Marc Bareille
schematics with the reversing attenuator.  I ended up making a couple of
minor changes to reduce the filter to near unity gain and softening the
self-resonance output level so it would not clip rail to rail.

I built this filter using NTE888M programmable op-amps since they are
readily available.  I did some frequency plots with my scope using XY mode
and was disappointed with the DC level shift I saw on the low pass output.
I got some of the УД1208 Russian parts to compare and discovered the same DC
shift.  I had repaired a Polivoks filter previously and the frequency plots
did not show any DC level shift.  I eventually figured out the DC shift was
occurring on the VCO output as I swept to high frequencies.  I tried
different VCOs and repeated the measurements with a VCO that had a constant
amplitude output with no DC shift over the frequency range of the
measurement.  Both the NTE888M and УД1208 Russian parts looked good with no
DC shift in the low pass output.

I saw only two minor differences between the NTE888M parts and the УД1208
Russian parts.  The NTE888M parts have slightly higher gain and the cutoff
or center frequencies was slightly higher at the same control settings and
input CV.  I assume the frequency shift is due to a slightly different Iset
response curve or a different Vbe bias point for the Iset control pin.

I have some uA776 parts on order and will also do the same comparison
(assuming the parts are genuine).

Now I need to finish this filter with a front panel and play with it some.
Details, photos, and scope images are on my site on my Polivoks Filter page.
http://modularsynthesis.com/kuzmin/polivoks/polivoks_vcf.htm

David J. Brown
Email davebr at earthlink.net
Website http://modularsynthesis.com