Improving simple VCF for guitar use

[Mick Bailey]

I've got an old mono synthesizer (a MFOS WP-20 built some decades ago) that I want to re-use for guitar use. It works quite well - much better than I expected. The very simple VCF is rather limited for guitar use, though it sounds pretty good when the synth is triggered from a sequencer. The main issue is that the resonance is fairly tame and I'd like to add a resonance control - if possible getting to the point of self-oscillation. Can anyone suggest how this could be implemented?

[Mark Hammer]

Normally, in that kind of bandpass filter, one raises the gain, and consequent resonance, using the feedback resistance. In this instance, it would be what is shown as R29 (again, noting that it is 100K not 100R). Note that doing so shifts the resonant frequency range downwards. So in order to achieve greater resonance, but retain the spectral range of sweep, you might want to reduce the value of C13/C14 (e.g., 15nf), or incorporate a range switch. But I think you can wait until you hear the result of dickering with R29. Try using a 250k-500k pot in series with R29 for starters.

I'm fond of simple all-in-one synth noise-makers, as well as everything the late great Ray Wilson gifted us with, so it will not be any surprise that I have a WP-20 sitting in the basement. It works, but I haven't used it in a while. I would offer to do the suggested mod on my own, but too many other things to get to first. But certainly let us know if these suggestions are worthwhile.

[Mick Bailey]

I'll certainly give that a try later on this evening if I get chance. I've had some really good sounds out of the setup today - very convincing, musical analogue synth noises. With improvements to the VCF it will be even better. I've fed the WP20 from the output of a pedal I built that has selectable 1/2/3 octaves down, fundamental and one octave up. It will hold a note for quite some time and long enough for the AR/AD envelope generator to work over longer settings. I've set up a comparator to generate a trigger off the guitar signal. My intention is to rehouse everything into a guitar synth package but omit the noise generator and VCO as I won't need these, plus make some other changes to give adjustable modulation to the VCF from the LFO.

Another thing I'm stuck on right now is getting the envelope generator to re-trigger when a note is struck but before the AR/AD cycle has completed. As it stands to get a good bowed sound with slow attack I have to wait until the cycle has completed, otherwise the new note falls within the cycle of the already triggered envelope. At the moment I'm experimenting with discharging the 10uf timing capacitor using a BJT fed from the trigger pulse, but that isn't working as expected and has no effect.

Edit: Got the re-triggering working - I had a bad ground connection.

[Mick Bailey]

A 470K pot in series with R29 works really well. It does shift the resonant peak down, but I like the growl it gives and want to keep that feature. Switchable caps is a good suggestion so that will be the next step.

[J M Fahey]

I see a very simple/crude bridged T filter, so very low Q, so "weak".
Feedback will improve that a bit, but nothing spectacular.

Works, sort of, with harmonic RICH synthesizer waveforms (squarewave - triangle - ramp - pulse) but not much on "natural" instruments such as guitars.

I suggest you check pedal schematics designed for Guitar-Bass use from the get go, such as Mu-Tron, Dr Quack, etc.

Morley wahs also used a simple RC filter (modulated with LDRs), but they tweaked it to perfection, just do what they did.

[Mick Bailey]

I've made quite a few changes within the scope of the original circuit (which is already committed to PCB) and it now resonates nicely and will even feed back. The signal is from a multi-octave pedal that outputs a square wave as well as a pseudo-sine, so is rich enough for the filter to act upon. It does work much better than it did with a straight guitar signal. I realise its a poor design - my main modular synth has proper VCFs - a Steiner-Parker as well as MiniMoog filters (that both also track 1v/oct), but I wanted to use a redundant circuit that was lying around.

[Mark Hammer]

The king of guitar filtering is going to be 4-pole lowpass. I have an old Sequential Circuits Pro-One. It has an external audio input External audio is not connected to either the internal VCOs or the envelope generators, but it does yield an envelope-controlled filter sweep. And let me tell you, that filter sound is delicious; much tastier than a simple bandpass, like the WP-20 uses (as does the old Doctor Q and so many other autowah pedals).

I have a schematic for a 4-pole ladder-type autowah from an old Japanese effects projects book. I tried building it once, but buggered up the PCB, so I need to try it again...properly.

[Mick Bailey]

Both the Minimoog and Steiner-Parker filters are also 4-pole - the nice thing about the latter is that it's switchable between low pass, band pass and high pass. It also has a very pronounced, aggressive resonance. Both can be easily driven into distorting, which gives added complexity and richness to guitar signals. My modular can be patched to take an audio input and with a trigger generator can use the envelope generator, VCA, sub-octave generator, S/H and ringmod. I suppose the WP-20 like comparing a DIY cigar box guitar with a Les Paul. It's a shame the Curtis CEM3379 is no longer made, as that may have been a good basis for a simplified pedal - apart from a 4-pole VCF it has a VCA in a single IC.

I've seen quite a few circuits over the years from a Japanese FX book - maybe from two different ones. Are there any PDFs out there? I've searched over the years but never found the complete book.

[Mick Bailey]

For comparison I went back to my modular and patched my trigger generator to the synth's envelope generator and the output of this to the VCF then to the VCA. Then patched an LFO to the VCF for a bit of modulation. The results are far, far beyond what the WP20 can achieve - which is what you'd expect. I also discovered that the Steiner-Parker is a much better filter for guitar than the MiniMoog.

I would have persevered with the WP20, as I had tuned the VCA to a much better sweet-spot for guitar (along with good resonance) and was getting some convincing sounds from it. The biggest problem though is the envelope generator is not re-triggerable without a significant bump due to the timing capacitor being quickly discharged via a BJT with each trigger. I've tried a smaller cap (from 10uf down to 4.7uf) and slowing down the discharge, but when I get to the point where I can eliminate the bump the envelope generator becomes ineffective for swells.

[Mark Hammer]

I've seen quite a few circuits over the years from a Japanese FX book - maybe from two different ones. Are there any PDFs out there? I've searched over the years but never found the complete book.

I was asked by my source not to post scans, since apparently folks are pretty strict about copyright and such in Japan. They may be less strict, 40-some years on, than he perceives, but he was kind enough to loan me the books to scan, so I'd rather not betray him. Some of the circuits in the books are straight clones of commercially-available products (at the time), while others are a little more advanced. This archive has lots, though not all the stuff I was allowed to scan. https://www.experimentalistsanonymou...tronics%20Book