Discrete SSM2040 Filter Experiment

burdij · Post by **burdij** » Mon Apr 23, 2012 12:14 am

A long time ago, I built a synth that used the SSM2040 4 pole filter IC. I thought it had a very nice sound. Lately, I have been experimenting with a discrete implementation of that chip as it is difficult and expensive to get and not a good candidate for a production module. In designing a circuit using simple OTAs and buffers, I also wanted to make the filter more versatile by incorporating a method of selecting the arrangement of poles and transfer functions. This is the prototype of the design:

Each filter section has a 1V/Octave and an attenuated voltage input for frequency control and a voltage control input for resonance incorporating a reversible attenuator. The 4 pole low pass mode acts as a sine wave oscillator with the resonance control set in the maximum position.

The filter mode is selected by a rotary switch and each filter can be set independently. The modes implemented by the 4 pole core are:

1 Pole Low Pass
1 Pole High Pass
2 Pole Low Pass
2 Pole Bandpass
2 Pole High Pass
2 Pole Notch
2 Pole High Pass + 1 Pole Low Pass
2 Pole Notch + 1 Pole Low Pass
3 Pole Low Pass
3 Pole High Pass
3 Pole All Pass (Phase shift)
3 Pole High Pass + 1 Pole Low Pass
3 Pole All Pass + 1 Pole Low Pass
4 Pole Low Pass
4 Pole Bandpass.

The two filters are internally normalled in series. They would be usable in parallel with an external mixer. So far, the results are promising.

thetwlo · Post by **thetwlo** » Mon Apr 23, 2012 12:29 am

holy fukkin ....!!!
my favorite filter with its' full potential....
other formats would be cool! still impressed! and may find a a way to make it work.
Do you have a link to a higher res photo? I know one of my 2040 filters sacrificed it's RES CV for CV control over filter mode(hp, lp,bp) which is a bit better.

How close is this to the 2040?
we already have so many clones of the 2040 --even the chip.

The Hamburglar · Post by **The Hamburglar** » Mon Apr 23, 2012 12:30 am

This is exactly what I want.

Ockeghem · Post by **Ockeghem** » Mon Apr 23, 2012 4:38 am

burnsjed · Post by **burnsjed** » Mon Apr 23, 2012 8:02 am

spinach_pizza · Post by **spinach_pizza** » Mon Apr 23, 2012 9:08 am

Hmmmmm......

russma · Post by **russma** » Mon Apr 23, 2012 11:35 am

Dead sexy.

Questions from an idiot:

1. What are "transfer functions"?

2. Which famous synth(s) featured the SSM2040 chip? Edit: Prophet.

3. What are the applications for using both a 1V/Oct and Freq CV inputs? Or is it just to provide an attenuated freq input?

Dave Peck · Post by **Dave Peck** » Mon Apr 23, 2012 1:12 pm

Quite an elaborate filter! All of those available modes remind me of the Oberheim Xpander / Matrix-12 filter modes - except of course yours has TWO of them. Im curious about the fact that the attenuated cutoff CV is NOT reversable, and the resonance CV IS reversable. I'm wondering if the opposite might be more useful..? In any case, it looks great & I'm looking forward to seeing how this one develops.

burdij · Post by **burdij** » Mon Apr 23, 2012 2:13 pm

thetwlo wrote:How close is this to the 2040?

I went back to the David Rossum patent for the circuit.

russma wrote:1. What are "transfer functions"?

The transfer function of a filter "element" is the relationship of the input voltage to the output voltage as a function of the complex frequency s. Each type of filter has a characteristic function. The function for a low pass filter is:

Since there are only first order low pass filter sections in this filter, in order to create other types of filters, outputs are taken from each section and ratios of the voltages are mixed in a resistive mixer matrix to simulate the transfer functions of the other filter types.

thetwlo · Post by **thetwlo** » Mon Apr 23, 2012 2:31 pm

russma wrote: 2. Which famous synth(s) featured the SSM2040 chip? Edit: Prophet.

Also, RSF Kobal, later E-mu modular, Voyetra 8, at least one rev of the Octave Cat, E-mu Audity(never went into production), Symetrix Phase Filter.
I'm sure I'm missing a couple.

touyube · Post by **sduck** » Mon Apr 23, 2012 3:26 pm

MOTM 440 on steroids?

emdot_ambient · Post by **emdot_ambient** » Mon Apr 23, 2012 3:49 pm

Audio demo. Must. Have. Audio. Demo!

moogah · Post by **moogah** » Mon Apr 23, 2012 4:02 pm

Would love to hear how this sounds!

And I don't mean to take any wind out of people's sails, but this idea has been done by Papareil synth labs for the MOTM440 (or any other 4 stage lowpass filter).

http://m.bareille.free.fr/modular1/vcf_ ... rapole.htm

emdot_ambient · Post by **emdot_ambient** » Tue Apr 24, 2012 12:14 am

moogah wrote:And I don't mean to take any wind out of people's sails, but this idea has been done by...

...Oberheim, Digisound 80, and the others already mentioned above. It's not a new idea, but there aren't many examples of it nowadays for some reason.

I'm doing something like it with some CEM3320 chips I have: switches just for changing each pole between LoPass, AllPass and HiPass, restricting it to always combinations of 4 poles.

earthbound · Post by **earthbound** » Tue Apr 24, 2012 3:41 pm

Looks really cool! Can each filter mode be selected independently?

tuanuibi · Post by **tuanuibi** » Thu Apr 26, 2012 9:03 am

Looks Excellent!

earthbound · Post by **earthbound** » Sun Apr 29, 2012 8:03 pm

I think it would be cool if the Freq CV was normalled between filter 1 & 2 until you put a jack in it. That way you could control both filters with one jack.

MUFF WIGGLER

MUFF WIGGLER

Discrete SSM2040 Filter Experiment

Discrete SSM2040 Filter Experiment