FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 

Elka Synthex (CEM3320 filter core)
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author Elka Synthex (CEM3320 filter core)

Long time lurker, first time poster.

I guess it also helps to say what level I'm at, I'm an amateur with some experience building a few successful circuits. I've made a CEM3340 VCO pretty much half of TH's Maximus, left out sub octave and squarewave, I've made a few CGS modules and have adapted a couple of the designs, nothing inspiring, but has built my confidence in EaglePCB.

I've been perusing the electric druids site: ( to help inspire the next investigation, and I decided to pick one of the most complex designs the 'Elka - Synthex' multimode filter!

I've got the schematic to as far as I can, but I have doubts on some elements...

In particular the wiring of the pots in the reference schematic and the wring of the switches. Could anyone have have a look at the schematic and see if the VR's and switches make sense to them*? Only the net A connects after the switches and BCD are lone nets, I would have assumed they would be routed to ground/0V instead?

The other thing the electric druid post mentions is the control voltage mixer and components to allow for inverse filter env.
I'm guessing the CV mixer would just be a summing opamp mixer? But what would the components to allow for inverted filter envelope look like? I'm a bit lost on that one.

Any help would be amazing.

*Some shorthand VR symbols confuse me as to their wiring
One fellow wiggler already made a version of the synthex filter, and he/she shared schematics, gerbers, etc... I guess you will find useful information there:
Eurorack Synthex filter
Silesius wrote:
One fellow wiggler already made a version of the synthex filter, and he/she shared schematics, gerbers, etc... I guess you will find useful information there:
Eurorack Synthex filter

Thanks for the reply. Unfortunately the schematics aren't shared, just the gerbers, which is great, but in this particular project what I've tried to set out and do is make everything from scratch, as I have done with the VCO, EG, VCLFO and Steiner-Parker VCF clone, all adaptations from schematic just so I could add pots and knobs on front.

Happy to share what I've made so far, all working at 15v and 12v but they're version 0.1 so all risk is on end user. I've taking inspiration from the brunswick DIY kit so all the modules I've build are bare PCB with pots/knobs mounted to same PCB as the chips and all PTH because SMD is not so easy/quick for me.

*EDIT: BOM and Hardware Whoops found the schematic in the BOM and Hardware folder

EDIT, EDIT: Now found original Elka Synthex Schematics from MatrixSynth Elka Synthex Schematics
Just beware that there are errors/omissions in the original schematic.
Pando wrote:
Just beware that there are errors/omissions in the original schematic.

Thank you very much for pointing that out. very frustrating

I wonder how much they've left out, there were 8 of these filterblocks, could some of the interconnections have been left out?

Will likely skip the switch decoder, as that may cause some headache.

Think I'll try and reconfigure the circuit to take inspiration from this design but try and use just use minimal amount of switches.

If the schematic is fully jenk then the best approach could be starting from scratch from the datasheet diagrams individually making LP then BP and HP; then implement switching between the responses.

Either way, found this really useful yesterday:

First, we will explain the parts to ignore as signal paths. Diodes 239 and 185 are protective circuits. The resistor 187 is to bias the analog switch not to break it. Resistors 238, 246 and 241 carry a bias current that determines the DC operating point of the OTA. The CEM3320 is an OTA with a very special circuit configuration, but just think it's just an OTA here. 3Qa is a mixer of the input. 3Qb is a CV mixer, 3Oa is an inverting amplifier that buffers the output switching.

Analog switches 1Q and 3Q are turned ON / OFF by control signals named A, B, C, D.
According to the user manual, there are four operation modes. I do not understand the operation of the decoder which has the correspondence between the operation mode and ABCD well, but if you guess from the circuit, it should correspond as follows.

 LP Mode: 24dB / Oct LPF: B and A ON
 BP1 Mode: 2nd BPF: B and D ON
 BP2 Mode: 4 BPF Next: C ON
 HP Mode: 2nd HPF: D ON

The 3320 has 4 stages of filters, but the latter 2 stages are fixed to the LPF. These operation modes are realized by switching the characteristics of the first two steps to HPF and LPF, and switching the path from the output to the 3320 resonance input (Pin 8 of 2Q). (The resonance feedback path is not shown in the schematic as it is internal to the 3320.) For the

first two stages, simply turn on / off the two analog switches in each stage to get the HPF characteristics. And switching to LPF. 10k resistors (173 and 235) help to make switching easier.

Let's look at each row.

When 2 analog switches are ON, it becomes LPF. The BC559 (element 237) is a buffer. The output impedance of each 3320 stage filter may not be very low.

When two analog switches are OFF, the signal passes through 10k and 330pF, but within 20kHz that can be heard by hearing, 10k is small enough compared to the 330pF impedance and can be ignored. Therefore, it will be a beautiful HPF with enough characteristics.

So what happens to the overall characteristics:

o LP mode: If only A and B are ON, the first two stages of the characteristic are the LPF, and the output is taken from the fourth stage. It becomes the 4th-order LPF.

○ BP1 mode: When only B and D are ON, the first stage is HPF, the second stage is LPF, and the output is extracted from the second stage, so it becomes a second-order BPF.

○ BP2 mode: When only C is ON, the first stage and the second stage are HPF, and the output is taken out from the fourth stage through the LPF of the third stage and the fourth stage, so it becomes the fourth order BPF.

○ HP mode: When D is ON, the HPF in the first and second stages is the second stage because the output is taken out from the second stage.

If the filter configuration changes, the amount of resonance feedback will not change either, but it is clever to be able to switch the amount of feedback at the same time as switching the output output location.

With this kind of function, it seems that it can be realized by modifying a four-stage filter using an ordinary OTA.

Knowledge about the design taken from: Houshu 2007
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group