Extending Antenna's Hand Distance detection

Jestern,

View TW from this link as all posts show up here.

The distance you need cannot be done with the theremin principle.

In my early research I did something like you want using Ultrasonics. My interest was volume control but it was a bit to sluggish. Pitch is normally a slower process.

I could see Ultrasonics working in your application, can you solder?

There is a spot on the PCB below where the voltage rises and falls due to the distance of an object. Use this variation to control a generated pitch. Stick an antenna on top to make it look theremin like, no one will know.

Schematic PDF          Product

This might be a better approach for a kids theremin in a museum, I will call it the Altermen. (Alternative Theremin)

Hi jestern,

You would be dealing with extremely tiny capacitance changes in an electrically and magnetically noisy environment.  You might be able to make something you can play from ~1m away, but only with larger plate antennas, very stable oscillators, hum filtering, and lots of averaging / low pass filtering - the last of which would probably make the gestural response very sluggish.  You could probably tailor the filtering based on the hand distance to make it response degradation less intrusive / more playable in the normal range, but >1m is definitely in the diminishing returns zone of hand & finger position data acquisition.

The useful maximum playing distance also depends on how you intend to play.  My prototype seems to respond to my pinky finger opening and closing at ~1m, but if I "play" by waving my arm around the response distance is obviously somewhat larger.  Moving my body around gives several meters of response.

[EDIT] And, as Christopher points out, at these larger distances you should probably be looking into alternative technologies to LC capacitance sensing.

Hi Oldtemecula, thanks for your reply, yes I can solder. I can try this. Is it reliable for 22 feet? I can use the input to control other parameters if it is stable... and reaches 22 feet actually.

Dewster I want to use it with dancers so body is more important than pinky.. it seems that your protutype could be interesting for us. Are you willing to tell me how you modify the original design or share your design? Or at least what modification you added?

Thanks for the answer guys :)

jestern, my main stuff is described (ad nauseam) over on my digital Theremin thread.  It's been going for years, and it's kind of all over the place.  Even with large plate antennas, I'm thinking maybe capacitance sensing isn't the best way to go for longer distances. Sonar, modulated LED reflection, or one of those X-Box IR spots + camera (Kinect) interfaces might be more what you're looking for.

An interesting concept.  Are you hoping to sense the bodies of multiple dancers or just one at a time?

Other pitch control techniques might be preferable.  For example, reflected light (infrared or visable) from the dancers could be used for a voltage controlled oscillator.

Multiple plate antennas below the floor of the stage might be used for traditional theremin design. The distance from the antenna of the feet of the dancers would fall into the range of detection.

Others here might have additional ideas.

edit: I see dewster beat me to the answer. Great ideas.  I like sonar. Easy to build and cheap, especially with 40kHz transducers.

jestern said:  Hi Oldtemecula, thanks for your reply, yes I can solder. I can try this. Is it reliable for 22 feet? I can use the input to control other parameters if it is stable... and reaches 22 feet actually.

You keep moving the goal post, you first said 3 or 4 meters and now we are up to 6m or 7m. If you want to monitor dance then I would have ultrasonics pointed down from overhead. Not knowing what  the response triggers or what kind of dance I would also place cameras on the bottom side or underneath looking up.

Christopher

I am slowly making my own theremin, and during my first experimentations, I used passive mixing for the pitch, that with the volume bias voltage on the same circuit. Something like:

fixed oscillator:   ----||----|_______ command grid of the audio preamp
variable oscill.:   ----||----|     |
volume bias:      ---------------|                  

Between the 3 oscillator's outputs and the command grid, all circuits was using passive components like capacitors, coils and resistors, and a diode for the volume bias, resulting into a general and messy mixing of the 3 signals.

The result was a very powerful and loveful sound, but with the volume oscillator influencing the pitch. At the pitch antenna, when the hand is moving away from it, the frequency was going from ultrasonic frequencies down to 0Hz and up again until a few hundred Hz then down again. It was really weird and fun to play with it, even if the instrument was not very usable outside very experimental performances (or as proximity sensor...). The 0Hz point could be set at a distance from a few centimeters to more than 1 meter from the antenna by playing with the frequencies of the oscillators. The range of the second lobe was up to several meters from the antenna. If a few hundred Hz is enough for your performance and you are able to hack the circuits, it can be a way to experiment with.

Now, I have added a preamplifier stage between the output of the pitch mixer and the VCA. The result is a good separation between the pitch and volume circuits, but I cannot go down to 0 Hz anymore because the pitch oscillators are synchronizing on each other at a frequency difference of about 30 Hz. I also get the second lobe at the pitch antenna, but it look like to have a much shorter range than before. So maybe it is also a path for experimentation here. But I didn't investigated that, so I have no clue for now.

Slowly making progress with my theremin. With the volume and VCA as it is now, I can vary the voltage supply on the volume oscillator. For a given setup, a higher voltage supply imply a shorter distance between the lowest and the highest VCA volume, and a lower voltage imply an extended distance. The minimum distance is limited by the power/current the oscillator can handle, and the maximum distance by the ability of the detector to control the VCA, that because less supply voltage imply less oscillator voltage, which imply that, at some point, the couple oscillator-detector will not be able to fully control the VCA.

That is just practical observation. To get something usable to extend the range of the volume antenna to a few meters, which I think should be possible, one should make measurements of the output of the detector at different distances. These measurement would determine at which maximum distance the oscillator begin to react. They will also make possible to create an amplifier stage between the detector and the VCA, stage which have a logarithmic gain or something like that. I also think it will be a main difficulty. You will get very weak oscillator variations at a few meters, which imply you must have very low noise supply, oscillator, detector and logarithmic amplifier stages. For the measurements, you will need a good frequency meter that can be read from a few meters away.

Maybe a better approach would be to use a beat oscillator for the volume. But without trying both, it is impossible to know.

Now than I am done with the making of my theremin, I begun to play it. What I can say now is than its antenna range is good for playing, but will be difficult to extend. The reason for that being the capacitive coupling with the antenna which start from a few picofarads when the hand is near the antenna and decrease fastly with the distance. You can try with a bigger antenna and win a few inches, but not more.

A better solution would be to increase the frequency until you can use some geolocalisation like system. It is such devices on the marquets, they use laser light for that. A single laser coupled with a rotating mirror. The rotation have several axes, which make possible to orient the laser in all wanted directions. The laser is pulsed, and like with a radar, it look at the reflected light and is able to analyse it. These devices provides an interface which can be used to command anything the customer want to command. Such devices are used into the Geneva airport to command the speed of the speedwalks in function of the number of passengers using them. In theory, they can be interfaced to anything.