Idea for Theremin Staccato Pedal

Looking at what industry uses for similar applications, Strain-gauge technology is by far the most prevalent.. Loads at <£20 ... I presume the strain gauges are fed directly to a MCU or dedicated chip..

If one could get the parts at low cost, this could be the best route (no reason for latency to be long, but I wouldnt think these designs would bother with making it quick! ;-)..

In all likelyhood though, one coul buy 10 scales for the price of one set of components - if you could even buy them.. And in all likelyhood, you couldnt get whats needed from the circuit even if you did attempt to build a pedal from a modified scale..

£12 @ Amazon

The Salter Black & Chrome Slimline Kitchen Scale has an easy to read display with 20 mm high digits and metric / imperial conversion. The easy to clean slimline platform is stainless steel and small enough to store away neatly. Features include aquatronic liquid measure function and add & weigh feature. This scale measures 160 x 180 x 2 mm. powered by lithium CR2032 battery which is included. Salter are so sure that this scale will not develop a fault they have give it a 15 year manufactures guarantee.

 PS - Jason and Thierry.... This is not intended to be SPAM!  ;-) Honestly, Im not getting paid.......

"So is this the simplest approach" - Gordon

I think that the simplest mechanically would be a few FSRs sandwitched between two panels - perhaps panels being bonded together with silicon sealant..

So say one sensor near each corner to distribute any localised force..

For experimental stage, all 8 wires from the sensors run to the 'box' - this allows sensors to be wired in series, paralell or however.. (and would also alow future modifications - for example, panning depending on where one applies the force - or anything derived from the X + Y force vectors)

You end up with a changing resistance proportional to force..

Huge number of way to convert this to current or voltage, but essentially you want it configured  (using opamps or whatever) to give say 0V with no force and 5V with the maximum desired force, and to give the inverse (0V at maximum force, 5V at zero force)

And there you essentially have the CV side done - feed either of these voltages to a VCA (or have 2 VCA's working anti-phase to each other) and your done.

Details you need to sort out is matching the output voltage to whatever voltage or current required for your VCA.. This is just a matter of trimming initial values etc. You probably need a user potentiometer or two for adjustment and audio levels or whatever, and probably a switch for +Ve or -Ve operation.

Apart from the mechanics, its all simple stuff.

If you could construcr an assembly to actuate a single FSR, then the process is virtually the same - just some trimming and perhaps a bit more gain to deal with the changed resistance.

Fred.

Hahaha! Theremin with bathroom scales attached. "I squeak your weight!"

And yet another sensor you could use - probably quite a good one for this application..

Linear (analogue) Hall sensor with magnet "actuator"..

Same sort of construction as I showed for the FSR, the hall sensor replacing the FSR and a permanent magnet above it which never physically comes in contact with it.

The beauty of this scheme is that its output not force actuated, so one would define the "feel" of the pedal in terms of displacement and the mecahanical force required to achieve this displacement - as in, the compressibility of the spacing material or spring or whatever.

These sensors are used on the Continuum, they cone in all sorts of variants.. This one looks good and is <£5 ..

http://docs-europe.electrocomponents.com/webdocs/1074/0900766b81074495.pdf (programmable)

http://www.analog.com/static/imported-files/data_sheets/AD22151.pdf ( Standard - probably best for this application - Unipolar mode with thermal compensation )

Fred.

COMPLETELY OT:

(I have just realized that I can place hall sensors under an existing keyboard and easily construct a sort-of Continuum type instrument - particularly easy for monophonic.. the output being proporttional to reletive displacement on adjacent keys.. so you could go between semitones by playing two adjacent keys..

This is really exciting to me.. it opens the possibility of an Odes type instrument that could provide both poly output and "ribbon signal" output - the "ribbon signal" could track some selected 'mode' - for example "last key played" or "highest key played" or "lowest key played - and the poly output (if included - initially it could just be a mono instrument)  could ignore the "ribbon related" keys..

This to me would be a nearly ideal instrument - could drive mono synths and voltage controlled heterodyning theremin voices, could get perfect pitch and microtonal intervals absolitely under the players control without any difficulty..

Yeah - this is what I want for me! ... Next visit to my loft I will dig up some old EP or other keyboard not worth repairing, then when I start, buy 100 hall sensors and magnets (will be cheaper than buying 50! ;-) , interface these to a FPGA or PSoC...Shouldnt take more than a couple of years to have a instrument outputting polyphonic MIDI and CV "ribbon"... ;-)

Dewster, a question of yours I didn't address. My guess is that Taubman's second pedal is a dummy, to raise his other foot to the same height as the one on the active pedal, for comfort or balance.

OK, time to summarise my thoughts.

Staccato - is that the right word? The proposed pedal has a wider versatility than just facilitating staccato. Perhaps better to adopt electronic instrument terminology and call it a Theremin Envelope Pedal (Envelope in the sense of an ADSR envelope modulating signal amplitude.) Or Theremin Articulation Pedal - T.A.P., or, as it has potential application other than with theremins, Flexible Articulation Pedal - F.A.P. - "The Fapper"?

Question. Of how much benefit is the Decay phase of the envelope? Modern synthesisers mostly provide it, presumably with good reason. How much would including it raise the cost and complexity of the device?

A requirement. It should be as thin as possible, so that from a playing point of view it should be as similar to tapping the floor with your foot as possible. 

The requirement for thinness and other factors precludes proximity sensing by capacitive, optical and ultrasonic methods, which is a pity, as this would provide maximum information from the player. Of contact sensing, the possibilities are a simple switch, velocity sensing (as with e.g. piano style keys) or pressure sensing (as with e.g. a continuum keyboard.) Consideration should be given to musical advantages and disadvantages of each approach, and to cost and complexity of construction.

A requirement. The ADSR logic should be reversible. That is to say, it should operate so that either when the foot is not in contact with the pedal that the audio should pass through the pedal without reduction in volume, this arrangement facilitating very brief interruptions to the sound with a rapid tap of the pedal, or that when the foot is not in contact with the pedal the audio should be muted by the pedal, this arrangement facilitating very brief staccato notes with a rapid tap of the pedal. The switch to reverse the logic of the pedal should be a standard stompbox foot operated switch. 

A requirement. Parameters of the ADSR envelope that cannot be inferred from a pressure or velocity sensor should be user configurable with potentiometers. In the case of a simple switch pedal, all the usual parameters should be user configurable. Duration of the Sustain phase would be determined by the length of time that the foot is on the pedal (or off the pedal, depending on the ADSR logic.) 

A requirement. It should be sensitive enough to respond to very delicate taps.

A requirement. It should be robust enough to withstand being stamped on by an enthusiastic player.

Future development (or "pro" model.) Integration with other electronic musical devices. Making and breaking foot contact with the foot pedal represents, in MIDI terminology a note-on and note-off event (or vice versa.) This information could be used by other devices to further shape a note e.g. by triggering an envelope to vary distortion of the audio waveform. This could be achieved with an analogue Gate signal, or any of various digital conventions e.g. MIDI or MIDI or OSC over USB or WiFi or bluetooth. similarly, the entire envelope could be transmitted over CV etc.

It can scarcely be denied that the supreme goal of all theory is to make the irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum of experience.

Albert Einstein.

Hello again Gordon - Sorry I turned up 2 hours late for the design meeting ;-)

No, im not going to go into a long one .. (not by my standards anyway ;-)

To me, it all starts at the sensor side - once this is "defined" EVERYTHING you are asking for / proposing can be achieved, and probably achieved quite simply.. So lets look at the the sensors:

[I will present each case with the apropriate MIDI possibilities just for clarity - these dont need to be implemented.]

I will also give a guess rating on a scale 1 to 10 (Mechanical and Electronic Complexity, 10 being most difficult or expensive ; Reliability and "Quality" {suitability for specified requirements} 10 is best) ... Everything below is IMO/MOO and largely guesswork.

Also note that any scheme below can be used to trigger an Envelope Generator [EG].. Forget about "Active down" or "Active up" - this is a simple logic (or in the case of an analogue output sensor, signal) inversion function - No problem.. If one wants, a simple comparator could be added to any analogue scheme to provide a trigger or multiple trigger points. 

1.) {M=1 E=1 R=6 Q=*1*?} The most minimal is a simple switch thats activated by putting pressure on it.. This conveys no data other than on/off.. [MIDI Note On message with velocity at 64, running status with "note off" defined as "On" with velocity = 0, or a independent Note Off message] There is no "analogue" data of any kind, nothing other than on/off

2.) {M=5 E=5 R=5 Q=*2*?} A change-over switch where the duration that the actuator terminal is not in contact (traveling between the two change-over poles) is timed, can provide "velocity" data in both directions  [MIDI Note On message with velocity determined by velocity ;-),MIDI Note Off message with velocity determined by velocity , Swap the status {Note on <-> Note off} to implement an "inversion"]  "Analogue" data is limited to the velocity at which the pedal is pounded and released .

IMO the likely problem with this scheme is the mechanical aspects of obtaining a switch of low enough profile and sufficient ruggedness... I think this scheme would be the least useful.

--------------------------------

The following are analogue output, for MIDI implementation I would propose comparators to detect "foot down" and "foot up" trigger points and transmit these with derived velocity as Note On / Note off messages, and a CC message continuously streaming the sensors positional data. In a Non-Midi setup, these analogue and comparator signals could be used directly or in combination with an asocciated EG to drive whatever processing (VCA, Filter / Wha or whatever).

3.) {M=3 E=3 R=9 Q=9} Linear hall sensor/s.

Linear hall sensors have a reliability advantage of having no physical contact and nothing to wear out through repeated flexing or stressing - reliability will probably come down to the quality of the chosen actuator magnets.

4.) {M=3 E=6 R=8 Q=9} Strain gauges

Electronically, strain gauges are by far the most precise and least subject to drift if implemented with precision electronics.. My feeling is that this level of precision is probably not required for this application, and the extra cost and complexity probably isnt justified.

5.) {M=1 E=2 R=7 Q=9} Force sensing resistors.

I would probably choose hall sensors for a high-end product, but its so close with FSR's that im not sure .. If cost of FSRs (Higher than Halls) is balanced against higher cost of the mechanics for Halls, and they came out the same or close, I would certainly choose halls.. If FSR's turned out much cheaper (when one factors in the mechanical and magnet costs) then I would go with FSR's.

----------------------------------------

The Q rating of 9 is kind of arbitrary for the above - they can all do everything required and more, they all probably deserve 10*.. Fred.

Of those options, I think Hall Sensors. I like the reliability.

Fred, shall we pause here for a while for other people to comment. Not everyone is on TW every day. :-)

Also - scratch the ADSR talk in my summary above - that made sense for a switch. For a pressure measuring device the envelope is whatever your foot does once it is in contact with the pedal, so no need for knobs for it. Perhaps, rather than one large spring, several small ones so that the player can incrementally adjust the amount of force required to move the pedal to his or her preference by removing or adding springs.

Yeah -

Sorry folks.. I treat this place like a design office at times, and like im dealing with potential clients rather than chatting with friends..

[final comment on the last .. think electronically adjustable, non mechanical "spring" - think loudspeaker cone with voice-coil wired to rheostat.. A big idea of mine back in the 80's to electronically adjust keyboard weighting.. it wasnt any good for keyboards though.. needed to be powered and took too much of that as each key needed one - a solenoid per key :-]

Nothing wrong with doing this in a business-like manner.

I'm as bad as you are for rushing off ahead of everyone. A bit of slow and steady is good too. ;-)

(Ah, right. Electronic springs. Noted.)

Question..

Is battery powering important / essential ?