Ken Stone’s Dual CMOS filters, a.k.a. “Twin Wasp” VCF.

Here’s a link to Ken’s description of the board.

Continuing the squeeze-it-into-1U-with-small-knobs approach, I designed this Front Panel Express panel for the Twin Wasp.  Key to the design is a miniature rotary switch, Mouser part #633-MRK-112A.  It’s single-pole, twelve-position, configurable to limit the number of positions to your needs.  At over twelve bucks a pop, it’s pricey.  But at least the knob is included.  The switch mounts in a quarter-inch diameter hole and even could fit in a slightly smaller hole, the bushing is that tiny.

Compared with the Bridechamber 3U Panel, the only features I had to give up were the inverted CV inputs and the spread CV, which is an external modification.  And mine is 1/3 the size!  I have created similar 1U panel designs for a dual CGS Steiner VCF and a dual CGS Serge 1973 VCF.  But I probably will not make those, because I have so many filters as it is.

The construction pic follows.  You need nimble, steady fingers to lay-solder the five wires to the itty bitty PC-mount legs of the rotary switch.  This filter uses four of the obsolete CA3080 OTA chips, the heart of so many 1970′s synthesizer circuits.  I was happy to finally use some of my stash of 50 that I’ve been hoarding for a few years.

I made one circuit change after taking the photo here. The FREQ pot sets the initial frequency. Ken Stone’s wiring diagram showed the CCW lug going to ground. But it really needs to go to -15V. And the summing resistor should be 270K or 300K, instead of the 100K on the schematic. I just added a 200K resistor in series with the wire going to the center lug on the pot.

I have to learn how to use this filter now!

Quantity:  2 in 1U

A pair of Flight of Harmony Plague Bearers, built into this small-format 5U panel:
Dual Plague Bearer

These are sold in a variety of packages.  See the Flight of Harmony Plague Bearer page for details and to download the manual.  I bought two barebones packs.  They will customize the order as you like with power connector, pots, connectors.  I didn’t order the pots, because I use panel-mounted Alphas.  What does it do? From the manufacturer:

It is designed to infect, corrupt and pervert a signal beyond recognition. It contains a Voltage-Controlled Resonant Bandpass filter that is designed with the goals of maximum signal alteration and maximum parameter controllability. Each filter has controls for varying the high and low corner frequencies of the passband, as well as variable gain and an input attenuator. Each section of the Plague Bearer can be used as a filter, an oscillator, a noise generator, or even as a resonance oscillator! Depending on the settings, the filter can be a lowpass, highpass, single-bandpass, or multiple-bandpass. Adjusting the resonance point can cause the filter to self-oscillate in many ways. By adjusting the controls to just below the point of oscillation, and then applying a pulse, square, or other abrupt-edged input signal waveform will “ring” the filter. A simple “click” on the input can give a percussive output sound – from bass drum to bell to a harsh metallic clang; Crank the controls up and this beast will scream like a banshee! The oscillation can be damped or continuous. Chain a couple in series and get accumulating feedback – the build-up can be slow, fast or instantaneous. White Noise is easy to do – you don’t even need an input signal! Turn the input all the way down, and then slowly turn the gain up. You can adjust the “color” of the noise with the frequency controls. Daisy-chaining multiple filters can give a “comb” or multi-phase noise. About phase-shift characteristics: A single filter can give a STRONG chorus effect if so desired, useful as a sub-oscillator or just to “fatten” a sound.

Construction

I butted the two PC boards together and mounted them to a Bridechamber 3-jack bracket.  It’s a nice, long bracket that left plenty of room for the jacks in front.  The P.B. boards each have four mounting holes, two on one end and two on one side.  I had to drill the bracket to match these.  The wiring uses headers on the board for all connections.  I like this, because if there were to be a problem, I could easily return a board for repair without unsoldering wires.

Although this model of the P.B. has a CV Gain input, I had to omit that in order to fit two of them into the 1U panel.  This is a little unfortunate, because the gain modulation produces interesting timbre changes.  Here’s a photo:

You can barely see the tiny surface-mounted chip!  These small boards are also low-cost.  Notice the two MOTM/Blacet/Oakley style 4-pin power connectors.  I made up a special power cable that jumpers them both.

Usage

I made a patch that uses them in the filter position in a nominally standard patch.  It’s called ‘Plague Birds’.  Look for it in the music section.

Quantity: 2

Description:

Mattson Mini-Modular Envelope Generator

Mattson Mini-Modular VCA

I combined the Mattson Envelope Generator and VCA modules into one panel.  Here is George Mattson’s description of the Envelope Generator:

ADSR transient generator with independent control of the Attack, Decay, Sustain and Release stages of the envelope.

Output is a varying control voltage from 0-5V.

Repeat switch. Allows for automatic repeat of the envelope sequence.

Repeat Level. Allows for selection of the envelope output level where the repeat function will re-trigger.

Man gate pushbutton allows for manually triggering the envelope.

And his description of the Voltage Controlled Amplifier:

Audio In jack for injecting audio signals into the amplifier.

Manual level allows for setting the master gain cell level manually. This allows any audio signal present to “bleed” through the amplifier output. This control comes in handy for checking the audio signal or using for continuous “drones”.

Master CV In Jack for inputting the main gain control voltage signal. Primarily, an Envelope Generator is patched into this jack, but other CVs can be used here for variable gain control of the VCA output. This input responds to a 0-5VDC control voltage which controls the master gain. No output will be present at 0V unless the manual level adjustment has been set above minimum.

Mod in jack allows for modulation sources to be input into the VCA.

Mod amt control allows for variable control of the amount the CV effects the VCA. As the control is rotated clockwise, the modulation goes from amplitude modulation through ring modulation then, back through amplitude modulation. This modulation affects the audio input signal, not the master gain.

Output level control allows for overall output level adjustment.

Main output jack provided for sending the VCA output to another module, an amplifier or headphones. The jack is a TRS jack and will source both sides of a set of headphones in dual mono.

I made the following minor modifications:

• Instead of an output level control, I have a CV input attenuator pot.  This is nothing more than putting a pot in front of the master CV input jack.
• Soldered a resistor on the back of the VCA board, acting like the output level pot turned all the way up.
• The VCA output is a normal jack, not a stereo jack.
• The output of the EG is patched to the switch lug of the VCA CV input jack as the normal path.

Construction

The Front Panel Express Panel uses the MOTM-jack grid design I’ve standardized on for my own miniature MOTM format.
Download custom FPE panel

As with my other Mattson conversion projects, I used a Bridechamber 3-jack bracket, combined with an adapter that I fabricated out of thin aluminum stock.  The adapter is needed because the Mattson PC board mounting holes are a bit too widely spaced to mount on the Bridechamber bracket.  In this case I also needed a bit more depth to get clearance for the jacks in front.

I used the convenient wire kits from Mattson.  Notice the power cable between the two boards.  George made them shorter than usual, just for this application.  The MOTM 4-pin MTA power connector on the rear supplies power to the whole module.

Operation

Both the EG and the VCA sport special features.  The repeat function on the EG is controlled by the Repeat switch, which disables the external Gate and manual push-button inputs and re-triggers the envelope at a point adjustable by the repeat level pot.  A quick trial run shows that audio rate re-triggering is possible by setting a short release and increasing the repeat level.  The EG can be an oscillator!  With repeat off, the Gate input and manual button both function, with a logical OR relationship.

The VCA is a high-quality AC-coupled, linear design with initial offset and CV attenuation (with my panel change). The cool, unusual feature is the “modulation” input for audio-rate ring modulation effects.

The EG and VCA are completely separate and independent modules, other than the connection of the EG out to the CV in to save a patch cord for the most-common usage scenario. This fits my goal of putting more functionality in a small panel space perfectly!  Yes, I built TWO of these.

From Ken Stone’s CGS website:

This is a licensed adaptation of the classic Serge DTG/DUSG

The Serge VCS module released in conjunction with from bananalogue.com is an extremely versatile control voltage generator and audio source. In the early 1970s, Serge Tcherepnin developed the Positive Slew and Negative Slew modules for the original Serge synthesizer. In time these merged into the classic Dual Universal Slope Generator. The Bananalogue VCS and CGS75 are an adaptation of Serge’s original circuit with a few new features.

The VCS is a unity gain voltage follower. The rising and falling slopes are independently and jointly voltage controllable over a wide range.

VC Transient Envelope Generator

A pulse at the trigger input will start the envelope, or a gate input will sustain the level and the envelope will fall when the gate goes low. Rise and fall are independently and jointly voltage controllable, with variable linear and exponential wave shapes.

VC Portamento

Voltage is slewed according to the rise and fall times.

VC LFO

When the cycle switch is thrown, the trigger input is connected internally to the end trigger output, creating a VC clock with variable waveform and independent rise and fall times.

VC Oscillator

While not as wide ranged, or accurate as a dedicated oscillator module, the VCS is still an excellent audio source. The Exp CV input is scaled approximately to the 1v/oct standard. The Output wave can be swept from triangle to saw with linear and non-linear waveforms. End Out also produces a pulse waveform.

VC Non-Linear Audio Processor (Low-Pass Gate)

If an audio rate signal is slewed, the module responds like a VCF, and a rough VCA. The signal is low-pass filtered down to silence, similar to a low-pass gate.

Envelope Follower

Positive and negative peak detection envelope follower.

VC Pulse Delay

Trigger input starts the envelope and a trigger will be produced again at the End Out when the envelope completes its cycle.

Sub-Harmonic Generator

If a series of triggers are applied to the VCS faster than the total rise and fall times, the module will divide the incoming signal by a whole number. In the audio range the output will be the sub-harmonic series.

I used the Bridechamber panel and a 2-pot Bridechamber bracket which fits the board out of the box.  I am especially happy to have these, because I am lucky enough to have used a Serge in the seventies and experienced the Negative and Positive Slew modules and the Dual Univeral Slope Generator.

I took the trouble to gain-match the NPN and PNP transistors in the lag processor, using a simple technique I found in Electronotes.  The 2N3904 pair are matched with each other and likewise for the 2N3906 pair.  Don’t know how important this is, but it couldn’t hurt.

Here’s a photo of my VCS assembly.  I ran a #20 stranded wire out from the power connector to the ground bus on the jacks.  I used twisted pair for most of the inputs and output, grounding the black wire back only at the jack, since there are but two grounds available on the board and those are dedicated to pots.

Description:  Mattson Mini-Modular Voltage Controlled Mixer

Here is a MOTM-ized Mattson VC Mixer!  From the Mattson website:

The VC Mixer consists of 4 channels.

Each channel has an associated gain control and control voltage input.

There are two outputs; Normal and inverted.

The control inputs respond to a 0-5V control voltage input. 0V is fully attenuated and 5V will produce a 100% output level.

The Level control attenuates the channel level.

The control inputs are cascaded. 5V is applied internally to the Control input for the channels and cascaded from channel 1 to channel 4. This allows the module to be used as a 4-channel mixer with no control voltage input to the Control in jacks.

When a Control input is plugged into a channel Control in jack, it interrupts the 5V from that channel through channel 4 allowing the gain for that channel, and all subsequent channels to be controlled by the new control voltage. New control voltages applied to the subsequent channels will interrupt the preceding channel control voltage for that channel and be applied to the remaining channels.

The trim pots on the circuit board are labeled with the associated CV gain adjustment and are set for a 1:1 input to output ratio. They can be trimmed to provide additional gain to utilize the mixer as an amplifier, if so desired. The gain will clip the original signal and can be set to overdrive the signal into a square wave peaking at the positive and negative power supply levels.
The mixer will work with both audio and CV inputs.

Construction

I used a Front Panel Express Panel of my own design, based on the MOTM jack grid like other recent panels I’ve been making.  This grid allows fitting more pots into a smaller area.

Bracket parts

The Mattson PC board is just a tad too wide to fit on a Bridechamber 3-jack bracket.  I made an adapter bracket out of thin aluminum stock I got at Lowes.  The adapter bracket mounts onto the Bridechamber bracket and then the board to the adapter bracket.  This photo of the bracket shows the bottom side that attaches to the Bridehamber bracket.  I matched the holes up to existing holes in the BC bracket, drilling out to 1/8-inch to work with some old PC motherboard standoffs I had.

Wiring up a Mattson DIY board is easy if you purchase the corresponding wiring kit.  All I had to do was to trim the wires and connect them to the panel parts.

Description: Mattson Mini-Modular Quad Buffer

I took Mattson’s DIY Quad Buffer board and added four attenuator pots to turn it into a quad buffered attenuator.  I call it a distributor, because each channel has two independently buffered outputs.  The board supports cascading the channels by providing a third buffered output for the first three channels that can be connected to the third lug on a switching jack.  I designated the four channels A-D, and decided to cascade A into B and C into D.  The pot on the first channel impacts both.  In other words, the A pot affects the A outputs and the B outputs.  Likewise the C pot affects the C and D outputs.  The B and D pots only affect their respective outputs.  There is a 1% voltage drop on the cascaded channels, due to the use of 100K pots and the 1K output impedance of the buffer.  (Note that when wired as intended the Mattson module doesn’t have any voltage drops along the cascade, because the input impedance of the buffers without pots is very high.)

What’s it for?  Quite a few modules in my system have control voltage or signal inputs without attenuation.  My other buffered attenuators (Oakley Multi-Mix, and Blacet Mixer) are bipolar and don’t have a good zero setting.  The Blacet mixers do have a center detent pot and a trimmer to null the center, but a small signal can still penetrate.  My own DIY Mixer-Comparator has only one channel. So I really wanted more zero-off buffers.

Construction

I used a Front Panel Express Panel design of my own, again using the MOTM jack grid like I did for the CGS Super Psycho LFO and Slope Detectors.  This is a compact design that essentially allows fitting more pots onto a panel.

The tricky bit is mounting a Mattson PC board onto a Bridechamber 3-Jack bracket.  The bracket isn’t quite wide enough.  I had to make an adapter bracket.   See my post on the Mattson VC Mixer for all the details.

Wiring up a DIY Mattson module is relatively easy.  Mattson DIY boards are assembled and tested.  I purchased the corresponding wiring kit with all necessary wiring pre-fitted to connectors.  All I had to do was to trim the wires to fit.

Quantity: 1 in a 1U-wide panel

From Ken Stone’s website:

This module is a much expanded version of the Psycho LFO, featuring six free-running oscillators, each variable between LFO and audio ranges, two of which can be switched to have triangular wave outputs. Each oscillator can be switched between low and high ranges, as well as off, and also has a rate LED, to allow visual determination of the frequency at which it is running. As per the original, there are also level and glide controls.

This module can be used to create a range of pseudo-random modulation voltages that can be used to generate unusual sequences, control VCFs or or VCAs etc. When run at higher frequencies, it can be used as a series of audio drones, or to generate complex sounds by frequency modulation a VCO.

Panel:

This is the second module I’ve built using the 8×2 MOTM jack spacing and small pots and knobs. You can download this Front Panel Express Panel.  This format works perfectly for the Psycho LFO, which I had never built because I did not wish to devote 2U of panel space to it, which would be required if standard 1-inch diameter MOTM knobs were used.

Modifications:

There wasn’t room on the panel for Triangle/Square switches for the two oscillators with triangles, and because the triangles impart an undesirable DC offset when switched off, I chose to omit the triangle option and make all six oscillators use square waves. I substituted 2K2 resistors for the specified 1K for the LED current. I could have used 5K6, because the LEDs came from an original Blacet Miniwave. They’re too bright! I’ll probably change them.

Circuit comment:

How many modes of operation does the Super Psycho LFO have?  It has six identical oscillators, each of which can be Off, Low Frequency or High Frequency.  Do the math:  There are 21 different combinations, I think, between one Low Frequency oscillator on to one High Frequency oscillator on.

Here’s the construction. The LEDs are not glued, but are held in place by the wiring harness.

The empty spot on the lower right of the PC board is where the IC for the triangle waves is omitted.  You can  see the two jumpers hard-wiring the square wave option.

Here’s how it looks in the cabinet with the Blacet Miniwave on the left and the CGS Slope Detectors in the same panel format to the right, next to the MOTM-650.

The slope detector is an event-driven gate/trigger generating device. It monitors a control voltage, and responds according to what that voltage is doing. If the voltage is rising, the slope detector gives a “gate” output (approx. 5V) on its “rising” output. Likewise, if the voltage is falling, the slope detector gives an output on its “falling” output. The duration of this gate signal depends on the incoming CV and the setting of the sensitivity pot. When the CV is remaining constant, the “steady” output activates.

These gate signals can be used to fire envelope generators dependent on the action of the CV, for example opening a filter when a melody is rising in pitch, and closing it when the melody is falling in pitch. It can also be used to generate gate signals from LFOs, envelope generators, to advance sequencers etc.

The sensitivity pot allows the slope detector to work with either relatively fast events, like the edges of a square wave pulse train, or very slow events, such as the sine output of an LFO being used to drive filter/phaser sweeps.

This is a dual module, housed in a Front Panel Express Panel of my own design. It’s the first of four modules planned in this style, which has sixteen equally spaced positions (8 x 2) where pots, jacks, switches, or LEDs can be placed. The vertical spacing between components is one inch and the horizontal 0.85 inches, which is the same as MOTM jack spacing. The use of 16mm Alpha pots and 3/4-inch Alco knobs allows the placement of pots anywhere. This is the design insight and also the main compromise. The Slope Detector has one pot, three LEDs and four jacks, perfectly fitting two of them into a 1U panel.  The red LED comes in a chrome bezel, Lumex PN: SSI-LXR4815ID.  SSI-LXR4815YD and SSI-LXR4815GD are yellow and green. I use green to indicate positive voltages, so I chose it to indicate Rising.

Modifications:  I changed the voltage divider resistors on the outputs to be 3K3/1K8 instead of the original 1K8/1K to reduce current draw.  I use these values for all of my digital outputs.  RLED is 2K2 ohms.

The 3-jack bracket from Bridechamber allows mounting the PC boards adjacent to each other.  (Two additional holes need to be drilled.)

Here it is in the cabinet, next to a standard MOTM format module. The most unusual thing is the LEDs located in what’s normally the jack region. I think the smaller knob fits in well.

This module is derivative of the Buchla 292c Lopass Gate.  The PC board and assembly instructions were made by Thomas White, described on his site on the Lopass Gate page.  The circuit is similar to the Cyndustries Quad Low Pass Gates, that I also built, and which is also based on the Buchla 292c.  I have always loved the sound of the LPG.  Although I had four of these in the Cyndustries module, I wanted to build Thomas White’s design for several reasons.  The QLPG has but one pot for each gate, which corresponds to the Offset pot on the Resonant LPG.  It sets the initial frequency.  The RLPG adds an attenuator pot for the control voltage input, plus a second CV input, and features a resonance booster with pot.  With two RLPG modules I have two lopass gates in a 4U space with more panel features.  I also decided to use VTL5C4 Vactrols, which have a slower response time than the VTL5C3 used in the QLPG.

Construction Notes

The Alpha pot part numbers in Thomas White’s part list are for the PCB-mount versions.  Unfortunately those are the wrong part numbers, because they are audio taper, and the circuit calls for linear taper pots.  I was building with the 1U BrideChamber panel and bracket, so I ordered linear Alpha pots, mounted them on the panel (using #5 O-rings on the bushings) and wired over to the board.

I found an interesting knob for the rotary switch, Mouser #506-PKGP-70B-1/4, which though a bit pricey at $3.75 USD each is really well suited.  It is a smaller diameter than the standard MOTM know, but has a pointer indicator that fits the fingers nicely.

Since I built the 1U panel with one signal input, I used the direct audio input option on the PC board.  I made use of the control voltage mixer to mix the CV and Control inputs.

How does it sound?  Very much like the QLPG, naturally, since it is based on the same Don Buchla design.  Of course the resonance (filter mode only) can be turned way up!  Do take the time to view Thomas White’s video demonstrations.

This module is a sawtooth-based VCO capable of frequency modulation (FM) extending past zero frequency into the negative-frequency regime. Thru-zero FM provides a much wider and richer variety of sounds than “ordinary” (positive frequency only) FM.

The negative-frequency version of a waveform is simply a time-reversed replica of the original waveform.  When a VCO is modulated through zero frequency, the waveform slows down to a stop and then speeds back up in the reverse direction.

The Teezer’s output waveforms also include triangle and sine waves, for producing sounds with fewer high harmonics, as typically used for bell sounds, train whistles, and so on. The unit also features a variable synchronization control that can be adjusted over a range of settings from hard sync to a fairly loose soft sync.  The module can also serve as a highly accurate and stable “ordinary” VCO (i.e., without the deep FM), with upramp, downramp, triangle and sine output waveforms.

A really cool, inexpensive through-zero VCO!  The Bridechamber panel is nice quality.   Its quirky layout with the offset pots and jacks is non-conformist, in keeping with the module itself!

Construction comments

I got the panel, PC board, and a kit of some special parts from Bridechamber.  High quality pots are used throughout.  As shown in the photo, I was able to mount the PC board to a Stooge 4-pot short bracket I had lying around.  The PC board was not designed for the bracket.  I could use only four of the six mounting holes in the board and had to drill holes in the bracket to match.  There was just enough room.  The power connector sits on the back of the board, cantilevered over the end of the panel.   I wish Mr. Fritz had provided more ground lands on the PC board; there is only one included in each of the two header-style connectors.  I used twisted pair for all the IO wiring and glommed all the grounds together onto a solid wire soldered into the one hole.

Circuit comments

I did the 15 volt power supply modification, as described in the instructions, using zener diodes in place of ferrite beads.  Five separate internal power busses are used!  Prior to soldering in the integrated circuits I connected power and verified all the supply voltages were as expected.  Good thing, because I had installed one zener diode backwards!  That was easy to fix.  As recommended, I had carefully matched the zener diodes D12 and D13 to within 0.01 volt.  Despite this, the triangle wave had a glitch that could not be trimmed out.  I modified the circuit to add a small resistor and got an improved waveform.  Still there is a small glitch that is audible as a harmonic.  You can hear this in the MP3 below.

Performance comments

This VCO tracks very well over at least six octaves.  I didn’t notice any problems.  I compared the thru-zero FM with the Cyndustries ZerOscillator (ZO).  The two VCOs do FM quite differently.  The Teezer has a simple-to-understand Initial Frequency pot that is a manual input to the linear FM.  In the mid position, the frequency is theoretically zero.  To the right is positive frequency (and a rising sawtooth) and to the left is negative frequency (falling sawtooth).  The Coarse Tune has an extremely wide range with audio frequencies being to the right of center.  Very low frequencies are possible for LFO use.

Through Zero shootout

I found that comparable effects are obtained from the ZO with the Range in HIGH and Bias switch in the LOW position.  For this test recording, I used a MOTM-300 sine wave as the modulation source.  The ZO and Teezer were set to the same initial frequency and the triangle outputs of each VCO were sent to a mixer.  You first hear the ZO unmodulated triangle, followed by a manual twisting of the LIN FM pot up and back.  This fades out and fade in to the unmodulated Teezer.  Here you notice more harmonics on the triangle than the ZO.  The LIN FM pot is twisted up and back.  This is a log pot on the Teezer (vs a linear pot on the ZO), so my jerky fingers are more noticeable.  You hear a slight beat also, which could be tuned out with the fine tuning.  I haven’t yet understood how to make the best use of the relationship between the initial and coarse frequency pots.