[sdiy] Capacitors in Moog ladder filter

[Tim Heffield]

Thanks to Jurgen, Antti, and all for the comments,


It looks like the answer to the question of which cap values are best
depends on your application and personal taste.  I have noticed that after
building a module (especially a filter) and verifying it on the test bench
you really can't get a good feel of it's operation till you drag it over to
the rest of the synth and try multiple signals, sweeps and conditions.
Unless you build up two of the same modules with the different caps
installed it's hard to do an A/B comparison (for my ears anyway).

Regards,

Tim Heffield

-----Original Message-----
From: owner-synth-diy at dropmix.xs4all.nl
[mailto:owner-synth-diy at dropmix.xs4all.nl]On Behalf Of
jhaible at debitel.net
Sent: Tuesday, April 27, 2004 9:19 AM
To: Antti Huovilainen
Cc: Czech Martin; Tim Heffield; Synth-Diy
Subject: RE: [sdiy] Capacitors in Moog ladder filter


> > It was stated several times on this list, that different
> > capcitor sizes will lead to different dc bias of the ladder,
> > if tuned to the same frequency.
>
> The control current will be adjusted, but the DC bias provided thru the
> resistor string will vary very little (millivolts).

"dc" refers to both, voltage and current, not just voltage.
(think of it, dc actually is an abbreviation for direct _current_
(;->) ).

It's true that the bias voltage will not change much (log
function of the pn junctions!), but for that very reason
the impedance of the circuit changes with current. So a trade
of ladder current vs. capacitor value would only then make
no difference if it wasn't for the loading of the ladder.


> If the transistors are assumed ideal (infinite beta, no early effect,
> perfectly matched), the control current & capacitors will only affect
> tuning of the ladder. Non-ideal transistors may affect this, but I don't
> know how much.

> I can't say offhand about the effect on the differential amp [...]

Be assured that the parameters of real transistors, the loading of the
ladder by various different differential amps, the choice of capacitor
values (and the linked choice of ladder bias current) have some considerable
effects of the sound (and the transfer function + nonlinear properties)
of a ladder filter. That's the reason why different Moog filters sound
quite different. For instance, the old Moog Modular VCLPF (904, wasn't
it?) has a rotary switch to choose 3 sets of capacitors, and the filter
will sound slightly different with each setting (cutoff correction
by applying a CV assumed).

This does not mean you need to copy a vintage circuit precisely to
cover all these little detailt - but IMO it means you must analyse
all the tiny contributions of special properties - like choice of
capacitor values, beta of transistors, architecture and component
parameters of the differential amplifier - and find a way to emulate
them all in your circuit, when you want to get that original behaviour
precisely.
(http://www.oldcrows.net/~jhaible/jh5/jh5.html)

Of course emphasis on every little detail is not needed, unless
your goal it to emulate a certain historic circuit precisely.
Moog himself has chaged the circuit of his ladder filters
from product to product - slightly changing the sound, and
probably always to make production easier and less expensive.
So if you just want "a" Moog filter, go for the cap values
most convenient to you, and use a high impedance BiFet opamp
differential amplifier a la Memorymoog. If you want to have
one specific Moog filter reproduced (or wonder about the
effects of capacitor values in these), expect some extensive
work of analysis.

JH.



-------------------------------------------------
debitel.net Webmail