Proximity controller: DONE!!!!!!!!!!!

[Zero the hero]

Good news, I uploaded the Invisible Touch proximity controller on my web site.
Built and tested, it works well. Maybe it needs more testing.
I successfully modded the DOD FX-25 Autowah in order to create a sort of a Z.Vex Wah Probe.
Still reading? Click on the oranges.....

[troubledtom]

cool!!!!!!!!!!!!!!!!!!!!!!!!!!!!!  good job!!!!
         tom

[bioroids]

Very cool man, thanks for sharing it!

Miguel

[nelson]

 This is going to make an excellent expression pedal.


Great Job Alex!

[The Tone God]

Here are some of my observations on that type of circuit.

That sucker I bet is a PITA to tune. I would even suggest putting the trim as a full size pot control available to the open space to tweak it when needed. This implementation will be tempature sensitive. If you keep it just as a bedroom thing then you could get away with it but if your gigging then there could be issues.

The design of the antenna will dictate alot of the behaviour and sensitivity. Since this is a variation of the differential clock theremin you will have to select the clock's R/Cs to fit the antenna.

I bet it only works with "raw" touch meaning only bare or lightly covered touching. I don't think it will work for anything heavy or insulative like footware. Do to the design's limitations I don't think it ever will.

It still a good first kick at the can. It is a similar although much simpler impilmentation of the Tran's ProxiOzzi thing.

Andrew

[Zero the hero]

Thanks everyone!
ToneGod, what kind of think should I do for emproving this project?

There was a lot of talking about making a proximity controller and, as far as I know, this is the first and only complete project I've ever seen on this forum (on the net?). I believe there were more people interested in this kind of stuff.

Does someone have schems or ideas to share?

[Marcos - Munky]

Like Andrew said, Transmogrifox built his ProxiOzzi and posted the schematic online some time ago. Don't know if it's still online. Do a little search in the forum to see if the schematic is still available. And talk to Transmogrifox, he have great knowledge and I think you and he can get a great thing together.

[The Tone God]

ToneGod, what kind of think should I do for emproving this project?

I tried a few different differential clock circuits. Ones that did what you did with the last gate, ones with PLL, even ones with frequency-to-voltage converters. The problem I found with these circuits is not the operation or theory these circuits work on but the human body itself in particular our capacitive range which is very small. It is hard to sense that small and physically narrow range and turn it into a wide varrying output.

You can try using a larger sensing surface which may require recalibration but will of course mean that it will be more sensative to the enviroment and other noise. You can try slowing down the clocks but then you risk leaking into the audio. You could try to improve the user's connection to ground by making them stand on grounding mat. Try higher voltages to boast the transmission. I don't know much else that could be done as I abandoned that type designed later on.

There was a lot of talking about making a proximity controller and, as far as I know, this is the first and only complete project I've ever seen on this forum (on the net?). I believe there were more people interested in this kind of stuff.

Does someone have schems or ideas to share?

You are not the first but you are right now as far as I know the only complete project availible on this forum. There was Trans' thing but if I remembered he used two geocities sites to handle each part of the circuit. The site that had the output portion is broken so no one that I have seen has built it. From what I gather he was putting the output of his clock circuit into a PLL which would be tunned the sensing clock and would output the resulting difference. A similar method to what has already been suggested here. I belive it suffer from some of the same problems.

As I said its a good shot and is workable for DIYers to play with in their homes. Hopefully it spark some more to play with this and perhaps come up with a better solution.

Andrew

[soggybag]

The Proxi Ozzi schematic that was posted here was missing part. Ther is a note on it that says something about connecting the output to a PLL and use it as a frequncy to voltage converter to generate a control voltage.

I keep think the whole thing could be simpler. If you have these two ocsillators producing a pulse wave out couldn't this be run into a cap to produce a control voltage with fewer parts?

It seems like like you could get something happening with a single chip, like a 4069, and a few more parts.

[Paul Perry (Frostwave)]

If you have these two ocsillators producing a pulse wave out couldn't this be run into a cap to produce a control voltage with fewer parts?

yes, if the "pulse wave out" was sufficiently well shaped to drive CMOS logic reliably. The theory of the "analog tachometer" is, you trigger a 555 one shot to give a uniform pulse for each cycle of the input signal. Then the output pulses are integrated (RC filter). More pulses = higher voltage.

[E-Boy]

I'm also working on something like this (I'm starting a little guitar effects hobby/business).  But my actual question is: 'Is it actually even worth the trouble?'  I'm trying to go the other than mechanical pot route and I don't see any advantage to the player besides the cool factor.
The regular Crybaby method lets you stop the sweep more reliably and feels good underfoot.

Any other ideas for wah pedal control?
I've got:
1.) Mechanically rotate a pot.
2.) Mechanically move a slide potentiometer.
3.) Hall Effect.
4.) Pressure sensitive pad. Patch of Shades or Jaques
5.) Optical
6.) Variable capacitance

[Zero the hero]

Well, yes, the coolness factor plays the greates role in this kind of project.

Anyway, this controller comes handy if you have no wah shell or other sensor-based controller.

[E-Boy]

Okay, it makes excellent sense from a diy construction point of view.  I was just wondering if you saw any other benefits for doing it this way.

[zachary vex]

8^)  well done.

i'd make one recommendation, from personal experience... regulate the voltage to the circuit (bring it down a few volts) so it's stable at least with respect to the battery voltage dying.  linear control circuitry generally has trouble with battery voltage... it tends to drift pretty severely over time.  just a thought... there's a 7.2V zener in my old probe circuit to stabilize it, but nowadays i would use the Seiko regulators i use in many of my other products.

 

[Jaicen_solo]

I had the same thoughts actually. I figured that it would be fine to use a regulator to drop the voltage all the way to 5v, that way the rest of the effect will usually have given up by the time the RF osc dies. That said, if a Zener will do the job I guess it's a simpler option!

[The Tone God]

I probably should have mentioned that I regulate my circuit's voltage heavily too. Not just to maximize the stability over battery life but to also keep noise out of the clock circuit(s). Noise can play real havoc with the circuit and put out bad output.

Andrew

[Zero the hero]

8^)  well done.

i'd make one recommendation, from personal experience... regulate the voltage to the circuit (bring it down a few volts) so it's stable at least with respect to the battery voltage dying.  linear control circuitry generally has trouble with battery voltage... it tends to drift pretty severely over time.  just a thought... there's a 7.2V zener in my old probe circuit to stabilize it, but nowadays i would use the Seiko regulators i use in many of my other products.

 

Zach, THANKS for these suggestions!!!
I was really waiting for your response!

[Jaicen_solo]

Sorry to resurrect an old thread, but I just want to let everyone know that i've just built up a proximity controller using the layout here: http://topopiccione.atspace.com/PJ16TPCInvisibleTouch.html

It does indeed work quite nicely, with a fairly acceptable range. However (who knew THAT was coming ),
there are a few issues as expected and predicted in this thread. Firstly bear in mind that i've got it unboxed (no ground plane) and i'm using a fairly small antennae (about 2"x4") as that's all the board I had spare.
Ok, basic operation is indeed exactly as zero has said. Basically, I connected an LED to the collector of the trannie, via a 11k resistor. Without the resistor, it's just on-off so that's one area to look at. I'm thinking the large 3.3uF cap could be increased to improve fine tracking. It's fairly easy to get it tuned, but keeping it there is another matter.
I'll post more in a few minutes, got some things to do first!

*j

[Zero the hero]

Good job, Jaicen!   
R4 and C5 set the envelope duration, and a larger cap value will increase the decay time.
I didn't knew that removing R4 would lead to a flip-flop operation... by linking the CVOUT to a 4053/4066 or a relais wwe could do a nice proximity bypassing circuit... but it's just another story.
I'll stay tuned for more!

[Jaicen_solo]

Well, proximity switches are pretty easy to realise, it's variable voltages depending on proximity that's hard to do.
I've spent a bit more time tweaking and found that even leaky Ge trannies will work, so that's a possible use for all those you think are scrap
As it is, i'm quite pleased with this thing. I'm going to add a rectified power supply for sure, but it seems fairly stable for now.
I'll hook it up to one of my effects tomorrow and see how it works as an actual controller.
I think for sure that with a rectified low voltage power supply in a shielded box, this thing has great potential.
Thanks Zero!

Edit: FWIW, if you swap out the transistor for a PNP you can reverse the operation of the controller. With NPN, the LED will go dark when you touch the plate, with PNP it will light up. I'm thinking it would be useful to buffer the CV, then split it to both types of transistor, allowing both up and down sweeps simultaneously (does that makes sense??). With a stereo chorus/flange/phase effect, that could be awesome!
I think it would be useful to make R4 variable, if only with a trimpot, as it can affect the range of sensitivity, and improve tracking at the closest points.