This year Lower West Side Studio has produced over 50 hand-built synthesizer modules. Of all those modules the CGS49 from Cat Girl Synthesizers was the trickiest build. Let me explain;
Ken Stone, maker of Cat Girl Synthesizer modules, has some outstanding circuit designs that make for awesome synthesizer modules. A lot of times though the documentation leaves a lot of interpretation and options. Many times the information required for a successful build must be looked up and cross referenced in order to determine the best course of action of a build. I’ve found the best sources for this information are on Muff Wiggler’s DIY forum and on Electro-Music’s Ken Stone Designs forum. This is, of course, on top of the documents provided by Ken for the modules themselves.
The CGS49 Dual CMOS Filter or Wasp filter is one such module. The filter was designed based on Electronic Dream Plant’s 1978 Wasp synthesizer. The filter features the raspy, buzzing quality of the Wasp’s built-in analog filter. Even better, the CGS design features two filters which run in parallel.
To make the filter even more fun to use a Spread control voltage can be added to sweep the filters control voltages together. This can produces some very bizarre and unexpected sonic results. It also makes this filter far more flexible than the single filter versions of the module which are available.
When building the CGS49 I began with two PCBs which I built simultaneously as I often do when building multiples of a module type. Doing so helps me cross-check each board to ensure the correct parts were used and to verify if a bridged solder connection was accidentally done. I do this by saying “shit! I think I bridged that connection” and then checking the other board(s) to ensure those connections either should or should not be bridged. In some cases solder can flow across a trace line which connects two points so a solder bridge is fine in that case.
The build of the boards went smoothly although on board #1 a solder pad was lifted during the build process. This connection was repaired by jumpering the two points together with a short piece of insulated wire. I have worked on a large variety of PCBs over the past year of building and while I love CGS designs I have to admit the board quality is lacking. Compared to other boards with clear, dual sided solder pads the CGS boards come up a bit short. The CGS49 also has some portions of the silkscreen doubled and therefore unreadable. One must refer to the documentation in order to determine exactly what goes where. To make matters even more complicated, while Ken often includes a large black and white line drawing of the board in the documentation the CGS49 features only a blurry coloured image which is more difficult to read. Still, this didn’t make the build too difficult.
Where things got tricky were in the wiring of the Frequency potentiometer and Spread potentiometer. In the CGS documentation it states;
“It has been suggested that connecting the frequency pot between +VE and -VE instead of +VE and GND as shown will allow the filter to be used at lower frequencies. If doing this, you may find it beneficial to change the wiper resistor up to 270k, or add an extra resistor of 150k in series with the wire from the wiper to the PCB.”
From what I read and learned for myself this is not merely a suggestion and must be done in order to get the module to function correctly. I found that wiring the lug 1 of the frequency pots to the -15v ferrite bead worked well for this purpose. Make sure though to wire to the side of the bead further from the power source so that the current has passed through the bead before passing to the potentiometers. I chose to solder my wires to the bottom of the board at the solder point of the bead. The connection was very secure and it looks more tidy. BE VERY CAREFUL not to let any exposed wire from these runs be left or they could touch some other component and zap it.
When wiring the additional 150K to the wiper connection of the potentiometer I suggest you wire the resistor in line with the wire and encase the connection in heat shrink. If you wire the resistor at the potentiometer’s lug you run the risk of the resistor bending over and touching the -15v connection. You could also just substitute a 250K (or 270K) potentiometer for the 100K one indicated in the documentation.
The Spread potentiometer wiring is also a bit tricky. Two wires run from lug 2 (the centre lug) of the potentiometer. One runs to the CVSN connection pad on Filter 1. This connection must have a 100K resistor wired in line. Again I suggest wiring the resistor into the line and then heat shrinking it rather than connecting the resistor to the pad. The second wire must connect to the junction point of the Filter 2 TL072’s pin #3, the 2K2 resistor connected to it and the 100K resistor connected to them both. This is actually easier than it sounds. You can wire the lead to the side of the 2K2 resistor which connects to pin 3 of the TL072. Again, I wired this to the solder connection on the underside of the board. Make sure to expose only the necessary amount of wire from the shielding to prevent shorting.
One other thing to note is that there are ten (10), yes, ten jumpers to be soldered to the board. If you miss one as I initially did you will have a board that does not function correctly. In my diagram of the PCB I have labeled them as J1 through J10.
Sounds a bit difficult, yes? It really isn’t but hunting down the information and then testing the options can be. That’s part of the reason I’m writing this article and I hope it provides some help to other builders of the CGS49.
During my build, unfortunately, I must have done something bad to my #1 board. Things worked then didn’t work. Then they partially worked. I swapped chips, I replaced caps and I tested parts. I also created several diagrams which I will share here. The CGS boards list the part values on the board rather than identifiers such as R1, R2… R42, etc. This makes for easy building but difficult debugging when you try to reference a part to others for support. I created a reference scheme for the CGS49 and then associated the parts with their location on the schematics. Here are the diagrams;
NOTE: You can right click the images above and download them to obtain higher resolution copies
After attempting to debug the first module for over a week I decided to shelve it and proceed to wire the second module I built. This module worked perfectly the first time further leading to believe that I blew up a component (or components) on the first board while playing mucky-muck with the -15v lines. I’m going to go back for one more look but in the meantime I’ve ordered another PCB.
EDIT: It was later discovered that using cheap, CA3080 chips obtained over eBay from china were the culprit all along. Avoid using cheap knock-offs. Pay a few bucks extra, get product that works and save valuable time.
I installed the CGS49 module in a 2U MU format panel I designed. The design utilizes 20mm knobs which are not my favourite for MU modulars and I might try to re-work my design to make enough space for the 25mm knobs I prefer. In any case the filter has a very unique sound and really adds a lot to the system. I’ve created a short video outlining its capabilities.
Ken Stone’s Cat Girl Synthesizer modules are definitely not builds for beginners. They are, however, fantastic additions to any modular synthesizer and I highly recommend them!
EDIT: Here’s the panel design with the 25mm knobs