Sunday, August 12, 2012

versatile EH small stone

smallstone1 smallstone2 This Electro-Harmonix Small Stone phaser has components from eras of both the second and third Russian-made versions. The PCB layout seems to be the third version, though I suspect the schematic and part values are similar in either.

I've always thought it had a great sound.. but the major drawback is its extreme lack of flexibility. A stock Small Stone has only two controls- a potentiometer to vary the rate and a 2-position "color" switch.

First, I added a CV input- a switched 1/4" jack which interrupts the connection between the LFO's output and the inputs to the four OTA phase shift stages. Clamping diodes and a current-limiting resistor protect the OTAs. This modification allows the phase to be controlled by a synthesizer's LFO, a powered expression pedal, random S/H, an analog sequencer, etc.

In the original design, turning the DPDT "color" switch "on" adds a lot of feedback, increases the depth of the LFO, and also slightly slows its rate. I disconnected the feedback portion of this switch, and installed a pot to vary the amount of feedback- from completely off to near-oscillation.

The other half of the original "color" switch (which affected LFO depth and rate) is now on the top left of the pedal and is labeled "deep/shallow". The switch on the upper right, which is labeled "phaser/vibrato", simply removes the dry input signal from the output, creating a nice mild vibrato effect.

The four original OTA caps were all .0068uf. I replaced C4 and C5 with .022uf and .22uf. A DPDT toggle switches the third and fourth caps (C8 and C9) between the stock value and .00047uf / .0047uf.

My unit did have a very noticeable volume drop when engaged. On web forums, many people suggest replacing R42 (4.7k) with a 10k (or even 12k) resistor, however many also complain about reduced headroom and bad distortion once they've made this modification. R42 affects transistor biasing and will cause distortion if its resistance is greater than 8 or 8.5k. So, a 8.2k resistor would be a much better replacement. In my case, I used a 10k trimpot (set to approximately 8.3k).

Also, that annoying 1/8" DC power jack has been replaced with a standard 2.1mm, negative center jack.

Here are a few sound samples:
- smallstone1.mp3
- smallstone2.mp3
- smallstone3.mp3

All component names are from these reference schematics (via generalguitargadgets.com):
- ggg_smallstj_sc.pdf

And here is the documentation for all of my modifications:
- my_smallstone_mods.pdf

E-mail me if you'd like to have your Small Stone modified similarly!

Friday, April 24, 2009

moog VCF in vintage enclosure

Over the course of the past few years I've owned two Moog Prodigies. I got the first one about six years ago and modified it extensively. Eventually I sold it to buy the second one, which is covered in an older post.

Of all of the Prodigy's features, I've really just missed the warmth of the Moog transistor ladder filter. But I've been so completely satisfied with the EML101's sound and versatility in every other regard that I decided it would be best to simply build a VCF module based on the Moog design.

I've never had much patience for prototyping or for planned circuit board layouts. So, as I've almost always done, I let the schematics guide the arrangement of the parts as I went. It's the most challenging aspect of my electronics work.. but its a good exercise of creative problem-solving.

While checking local thrift shops for an enclosure, I found this old electric shoe-shiner. The 'Regina'-brand "Electric Shoeboy." It originally had an industrial-type motor inside and rotating brushes protruding from each side.

After removing the motor and the shaft, dremmeling the case, and super-gluing/epoxying cut-up aluminum to cover the holes, I had this finished enclosure! (The last image in this post was found on google to show approximately what this unit looked like before modification.)

I based my circuit on yusynth's plans, but changed a few component values for a greater range of control. I also had to scale the keyboard CV input for the EML101's 1.2V/Oct.

Here are a few sound samples:
- vcf1
- vcf2
- vcf3
- vcf4


E-mail me if you'd like me to build a synth module in an enclosure of your choice!

Thursday, March 26, 2009

moog (modular) prodigy

This is a truly unique Moog Prodigy! I purchased it on ebay, broken, about a two years ago. After restoring all of its original functionality and its original appearance, I realized it held a lot of potential for interfacing with other modular equipment, including my EML101.

Luckily, this was the later model with the Sync Input, VCF Cutoff In, Oscillator In, Keyboard In/Out, and S-Trig In/Out.

I added a Gate Input, Gate Output, Oscillator 1 Output, Oscillator 2 Output, Mixer 1 Input, and Mixer 2 Input.

The 'Mixer Inputs' come before the Mixer potentiometers, so (one or two) external signals can be injected, attenuated, and mixed.

The Oscillator Outputs are interesting, too, because they act as effect send/returns. For example, one oscillator can be run through a phaser or vibrato effect, while the other is left dry.. and they both run through the VCF and VCA.

In lieu of drilling and adding a VCF FM control pot, the Oscillator Output jacks make it possible to patch the audio frequency CVs to the VCF Cutoff Input with a simple attenuator in between.

I built an additional unit to serve as the Prodigy's second LFO, Noise Source, and Patching Attenuator.

Due to financial constraints, I had to sell this Prodigy on ebay about a year ago. I vowed to one day build a cloned VCF (so the EML101 could impersonate the Moog sound).

Here are some sound samples (recorded with the additional module for modulation, noise, and/or patching.. and some spring reverb):
- prodigy1
- prodigy2
- prodigy3
- prodigy4
- prodigy5

E-mail me if you'd like to have your prodigy modified similarly!

Thursday, February 26, 2009

modified poly-800 mkII [number two]

This Poly-800 mkII features the same modifications as my previous Poly-800, but it also includes the 2/4-pole filter switch.

The 2-pole VCF output sounds wonderful. If Korg (25 years ago) had me helping to make their big decisions, I would definitely have argued that the filter output must be switchable between 12 and 24dB/oct.. I also wish they'd added this option on my beloved DW-8000.

My control layout and labeling were also improved on this Poly-800. Additionally, the external input can now be used regardless of the setting of any parameter when the 'Noise+EG3' switch is turned "off." This causes the third envelope to have no effect on the external input level. So, to solo the external input, parameters 17 and 27 (Osc. 1 and 2 levels) simply need to be turned down.

When using my EML101 (or the Axxe I'm modifying for a customer) to generate triangular LFOs, random S/H, or even envelopes to control the Poly-800's filter, I realize how lively the Poly-800 can really sound. It takes on a much more antiquated and nuanced character. I think the 800's modulation generator just doesn't sound as smooth as a real analog LFO.

Recently, Kate and I have recorded a lot of material using the Poly-800 for its abilities as a synthesizer, but also as an effect processor for guitars. It appears on some of the new hushes and briefscenes songs.

This Poly-800 mkII is currently up for auction on ebay.com! ..and YES- it does come in that cool original case!

Saturday, December 20, 2008

modified korg poly-800 synth / processor

The Poly-800 mkII includes a wonderful, warm (stereo) digital delay (with modulation!), a great-sounding VCF, three envelopes (one for each of the two oscillators, and one for the noise source/VCF), and an LFO to modulate filter cutoff or pitch.

I've also added several well-designed modifications on this unit. First, it now has a VCF CV Input, which allows the filter cutoff to be modulated by a modular synth, CV pedal, or other signal generator. The input is designed to receive the standard 0-5V. Clamping diodes and a current-limiting resistor ensure that voltages outside of that range will not be harmful to the Poly-800's circuitry.

Second, I added an External Audio Input to the Poly-800's VCF and digital delay. The modulation delay provides great analog-type echoes, flanges, and choruses. Modulated filtering can also yield some very interesting effects.

This input is designed to work perfectly with guitar or other low-level signals. An added circuit provides the right impedance and just enough gain to boost such signal, and a pot allows you to attenuate the input level before the gain stage.

The Noise / External Audio Switch disables the white noise source, so that the external signal can run through the Poly-800 alone. The DPDT switch also holds the external input level at maximum (regardless of the settings of parameters 61-65).

The Filter FM control pot adjusts the depth of Filter Cutoff Modulation from DCO1's audio. This allows for an extreme range of sounds- from growling synth bass to ring modulator-type noises. It should typically be used with Chord Memory in 'mono' mode.

Last, the VCF Cutoff and Resonance pots allow the cutoff frequency and the amount of resonance to be tuned in real-time.

The program parameters (31 and 32) now set the maximum levels. The filter will self-oscillate, and the resonance pot offers its full useful range when parameter 32's value is set to 12.

Here are a few examples of this Poly-800's more unique capabilities:
- FMsyn (FM, cutoff, and resonance settings are being tweaked.)
- SHgtr (Guitar is being processed with a random S/H input to the VCF CV Input.)
- spacegtr (The delay sounds beautiful as a guitar effect.)
- SHsyn (Again, the filter is being modulated by the EML101's random S/H output.)
- wackygtr (The filter is being modulated by the Poly-800's internal modulation generator.)

E-mail me if you'd like to have your Poly-800 modified similarly!

Monday, November 24, 2008

analog MIDI drum machine

This is my analog MIDI drum machine.

The enclosure came from one of our local thrift stores where I was hoping to find a cash box or similar metal enclosure. This recipe card holder happened to be exactly the right size, and its easy-opening lid was something I needed because I didn't want all of the controls to be mounted on the outside of the case.

The 8 knobs on the top of the case control Master Volume, Kick, Clave, Snare, Conga, Low Tom, High Tom, and Hi-Hat levels. The two LEDs indicate power and midi activity.

On the rear of the case I've added only the most basic connections: +15V DC input, audio output, MIDI input, and MIDI thru.

The drum sound circuitry was taken from a broken old Farfisa organ. I removed the entire rhythm accompaniment assembly and later built a power supply for it, turning it into a stand-alone drum machine.

The sounds were great, but the preset patterns and preset drum mix were limiting. I began to investigate the options for MIDI control.

The Farfisa circuit board is the larger one mounted on the box's top. I added the blue trimmers to the solder-side of the board to allow for easy adjustment of resonance on the kick, both toms, clave, and conga.

Originally, the drums' levels were simply mixed by resistors, but I added 10K volume pots for each sound, and replaced the resistors with new values for more matched levels.

I added MIDI control with a lot of help from John at HighlyLiquid.com who sold me a customized MSA MIDI decoder kit.

His chip turns MIDI data into 8 channels of +5V triggers. 7 of the channels were configured to trigger drum sounds, but the 8th momentarily switches on a 'long-decay' effect for the hi-hat.

By tracing the original Farfisa circuits, I realized variable attack and decay settings were set by logic to create a variety of cymbal and shaker sounds.

So I added the two controls inside the case to adjust hi-hat attack time, and hi-hat release time (which only has an effect when hi-hat 'long decay' is triggered by MIDI).

I originally meant to use the IEC mains jack and an internal power supply, but despite my experiments with filtering and shielding I could not eliminate the noise from the power transformer! Eventually, I gave in to a wall-wart power adapter.

The +15V from the adapter is filtered, converted by a regulator into +12V, which is filtered and powers the audio, mixing, and amplifier circuits. The MIDI board uses a 7805 to supply the +5V for trigger pulses.

I had to configure the MSA decoder's firmware by SysEx to respond to the right MIDI notes and with the right pulses and triggers.

The
Kick and both Toms use an output mode which toggles and latches +5V and 0V alternately upon receipt of a "Note On" MIDI message. The Clave and Conga use an output mode which supplies +5V constantly, except when a "Note On" MIDI message triggers a short (.5ms) "off" pulse. The Snare, Hi-Hat and Hi-Hat 'Long Decay' Switch use an output mode which sends a +5V trigger for the duration of a MIDI note.

On these channels the drum sounds can have variable durations, which is very useful and allows for some interesting programming tricks.

I changed a few resistors to increase the gain of the original preamplifier and amplifier circuits, but the output level is still a little weak. Eventually, I'll probably replace them with some simple Op-Amp circuit.

Here are some sound samples (recorded without any effects or processing):
- drumtest1
- drumtest2
- drumtest3
- drumtest4

E-mail me if you'd like to have your analog drum machine modified similarly!

(photos by kate and robert.)