Idea for Theremin Staccato Pedal

"that we could be carrying some assumptions over from the initial proposal, which prompted me to challenge every part of the "yellow pearl" specification" - GprdonC

Hmmm..

You are right - this does need to be done every time any change to any part of the concept happens.

And, thinking about the rrequirements now, for a single (or changeover) switch, and looking at Capsense with shielding on the bottom and its continuous base-line tracking..

I actually wonder if a capacitive foot facing sensor, actually sensing the foot (rather than some "actuator") might even be possible.

But whatever - actuator or foot, the difference would be firmware if using PSoC CapSense.

Oh - Before I adopted Cypress PSoC I used Quantum Research (a Southampton University Spin-off company) QProx sensor IC's - At that time they were way ahead of everyone else, but expensive.

ATMEL bought Quantum shortly after I became a CYPros, and incorporated into some of its MCUs.. Its been a long time since I looked at whats available (I dont actually use simple short-range capacitive sensing on anything) but QPROX was at least as good as CapSense years ago, and probably better - so the ATMEL MCU's may be a better choice than PSoC (for this application at least - where one doesn't need the PLD's) .

Fred.

Just to confuse matters more ;-) .. I do actually, somewhere, have some QPROX sensor ICs - both switch and linear variants.. These are 8 pin DIL simple parts (I had completely forgotten about) with an in built DSP - I used them in a grip sensor for joystick manipulators.. They were used to detect loss of operator control at a nuclear waste processing plant - one never gets feedback on such projects unless they fail - and anyway I was just an employee at the time ;-)

These parts cost me about £12 each back then but are sitting in an antistatic bag somewhere - you interested? - No good for production as now replaced, but could be great for a one-off.. They are eventually destined for disposal via Ebay.

It may be worth searching QPROX - but on the other hand, probably not.. Having the sensor and MCU (whether PSoC or ATMEL) combined seems by far the best approach AFAICS.

(just had a look at some Atmel parts - they seem to do everything needed, including QPROX, some limited programmable logic, timers etc, ADC AND 2 channels of 12 bit DAC! - Hey, I might just jump ship!! - [I wont, I need PLD more than DAC -] ;-)

But I suspect there are more Atmel proficient developers in the UK than PSoC developers!

Atmel Xmega Data sheet

ADDED ->

Just done some reading of the Atmel Qtouch library .. Active shielding is not implemented, and must be added externally, in its simplest this is a logic buffer (somewhat less effective) otherwise a fast RR op-amp. The primary method they use to do "shielding" is to add another sensor to detect the 'background' capacitance and use this for correction... Other than that though, QPROX is still, by the looks of it, a bit better than CapSense.. It always was, but now that its free with Atmel MCU's I understand the war between Cypress and Atmel...

Back in the early days, Cypress put some QPROX IP into CapSense - I was "with" Quantum Research then, and they started (or threatened) legal action against Cypress (a brave thing for a small UK manufacturer to do to a large USA company) but Cypress conceded immediately and altered their API's.. A load of clients who had gone over to PSoC from QPROX returned to QPROX as Capsense was far inferior - But then Cypress greatly improved Capsense, Quantum realized that clients wanted QPROX in a MCU and not as an expensive extra part, so teamed with Atmel...  And in the end, the Americans now own the lot!

:-(

Fred, I can't find where you said it, but you reminded me that the velocity sensing keyboard solution was for further development or version 2.0 or something along those lines. I did find the bit where you said it was your preference, and there should at least be hooks left for it.

Yes, I absolutely agree that it should remain an option. I feel that the "volume is proportional to distance from sensor" solution is more theremin-like in its bare bones simplicity, whereas an envelope generator is a more synthesiser-like solution that may make the device more appealing to a wider range of musicians. So yes, if it ever comes to marketing a product, then it should have knobs and switches a-plenty and lots of interesting options. But first let's see what the bare-bones can do. 

" actually wonder if a capacitive foot facing sensor, actually sensing the foot (rather than some "actuator") might even be possible."

Which would bring us full circle. :-) 

I had a read of the Amtel proximity design guide (http://www.atmel.com/images/doc10760.pdf) and it certainly gives me the impression that a floor or table standing shielded plate sensor would work just fine.

And with a range of up to 250mm too. Which actually isn't too bad for a regular theremin volume loop. If it can have a large sensitive range as well as a very small one that it will work just as well as a hand operated "volume-only theremin" - something I have thought would be a good idea for quite a while. 

It also occurred to me that a bare-bones device could have auto-calibration like the Open Theremin - press a button, put your foot in the right position for full volume and silence in time with the beeps. It should be smart enough to allow either full volume or silence to be nearer the sensor. 

"And with a range of up to 250mm too. Which actually isn't too bad for a regular theremin volume loop. " - Gordon

Yeah - I came to that conclusion as well ;-)

I am a bit wedded to higher voltages on the antennas - remember this low voltage C-sensing using PSoC is where I started, and where I made my first really expensive mistakes.. Ok, the early CapSense from Cypress was little more than an RC with timer, and its got a lot better in these 7 years.. The Qprox sensors were better back then, but too expensive and required too much other stuff to integrate them into Epsilon.

And yes - even the early CapSense could do the job - IF one applied extensive averaging... But I wanted latency to be low, so that option was out..

At this time I havent even really studied either the Cypress or Atmel data on their CapSense / Qtouch, just looked over it briefly.. And really doubt that I will use CapSense - If I was making a pedal in an enclosed capacitive containment then I certainly would use CapSense (or at least try to ;-) and if I was just going for shallow switch "mode" proximity sensing I may try it - but for anything requiring slightly further range (say > 100mm) I feel extremely cautious.

I have, however, absolutely no doubt that "theremins" and theremin related products (volume controllers, pitch controllers, 3d E-field interfaces etc) using some of the many out-of-the-box capacitive sensing "solutions" will appear. Already a "theremin" has been published by Cypress.. IMO, its not good for "us" -

<deleted section>

Certainly I would strongly caution against jumping into Capsense or Qtouch for any distance sensing application - It all looks wonderful on paper, and even ok at first on the bench, and in ones enthusiasm (if you are like me - and being an aquarian we might have similarities ;-) one may fail to look for flies in the ointment until you have made a great big pot of the stuff... Oh, I could be wrong! - It may have improved to a level where it is usable and fast enough - but I would be highly surprised if this was the case.

So before you go Capsense, give me a call ;-) ... I intend to publish my stuff, but might just give myself a year or two head start before I do - But would be happy to help friends.

Fred.

Fred Said: "Lol, someone else who thinks "bubbles" ;-) .. Ok, 1MHz, 2" sensing ... Thats REAL simple!"

Congratulations to Fred when his staccato design is demonstrated for the first time.

There is no way I could compete with the complexity this staccato thread has taken on. For those that are at a hobbyist level in electronic design I invite you to visit my approach to the Staccato Pedal which I demonstrated four months ago. Like most of my WebPages this project was retired due to lack of modern day makers. Most kids today think anything can be built using an Arduino, where did they get that mindset? Interestingly Radio Shacks carries them. I shall remain old school with the little colorful pieces and live in my bubble full of solder smoke.

Visit my pitch-popper webpage, several analog methods of aggressive theremin volume control inspired by TW.

Christopher

Aside - found another theremin variant with separate controls for dynamics and articulation, this time both in a hand held controller.

Oy! Enough of that.

Back to the pedal - a floor or table standing plate with a large capacitive sensing range is possible in principle, but the lever and magnets model is doable with existing technology. I choose lever and magnets. I don't need the large range for this application, and I do like having magnets involved. Magnets are neat. Not the most sensible of reasons, but there you go.

And I had another thought - having a moving plate means we could attach a sensor to the top of it, at the opposite end to the hinge - possibilities might include inclinometers and accelerometers - or build a little platform over the end and mount a sensor above it - optical or ultrasonic, for instance. 

OK.

Agsin, thank you for your massive contribution to this project, Fred, and everyone who has helped me get to this point. I hope you all got something out of it. For my part, I now have a very clear idea of what I want, sufficient to give unambiguous and specific requirements for both the electronic and the mechanical aspects of the pedal, and I have learned a good deal on the way.

A break from posting seems in order while I order a selection pack of neodymium magnets online and do some experimenting to see if I can make them work as I envisage, and, if successful, draw a design for the pedal and a spec for the electronics. 

That looks great Fred. Thank you.

On the subject of potentiometers - I was wondering about those linear pots that get used on test rigs to detect motion - they look like little pistons but can't find prices for them - they seem to be a trade-only thing. Worth pursuing or dead end?

And a bit of an update. I'm mostly continuing this on the dreaded Facebook for a while as I can reach EEs and designers on my friends list who don't habituate Theremin World there. 

Here are a couple of videos I have posted - you don't need toy sign in to Facebook to view them - first one demonstrates magnets in action and second is a a rough sketch of what the pedal might look like...

https://www.facebook.com/video.php?v=10152388969976319&l=8384082133649363803

https://www.facebook.com/video.php?v=10152394238306319&l=8855288541763050929

Screwed up above posting - pasted below:

Hi Gordon,

In designing my PSoC4 stuff, I needed some low cost DIL D/A converters, and found these which look ideal for your application (as a DCA)

<<I have edited /changed this post>>

UPDATE 2 >>

Much simpler solution - Maxim Log digital potentiometer, 65*1db steps, and available in DIL.

http://www.maximintegrated.com/en/products/analog/data-converters/digital-potentiometers/DS1801.html

Maxim / Dallas DS1801 Digitally Controlled Potentiometer Application Note.

The above schematic looks like everything you need for the DCA and even gives a spare audio channel free if you wanted it!

I love the ZCEN feature! this just clinches it for me .. and the fact that the input can go up to 700kHz actually makes it suitable for HF theremin signals!

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

In reply to your last posting:

"That looks great Fred. Thank you. On the subject of potentiometers - I was wondering about those linear pots that get used on test rigs to detect motion - they look like little pistons but can't find prices for them - they seem to be a trade-only thing. Worth pursuing or dead end?"

Hi Gordon,

Those videos look great! - I am quite surprised by the magnetic action in the video.

As for the 'pistons' - I think these are mostly linear displacement transducers based on inductance, and are staggeringly expensive! (last time I needed one was for a Langmuir Blodgett trough back in the '80's and it clocked in at £250) - there may be cheaper or even resistive ones now, but I know nothing about these.

I dont visit any face-book or similar sites.. Never realized the extent of theremin conversation going on outside TW (and LevNet)..

And I have instructed myself to stay away. I was in the process of trying to find my log-in details and join the conversation... But if I did that, my life would be gone, over, I would never get to do anything real -

Hit me like a brick- I am on the edge of being a junkie.. TW is the methadone, just keeping me on tickover..

-I think the internet is as dangerous as "the needle" for some people, and I am one of them.

I know what you mean. I'm a hopeless addict, and I spend far too long tweaking things that don't really need tweaking. To save you from temptation I have embedded it...

I thought the pistons might be pricy, but gosh! That much!

The magnets are kind of cool when there's no stopping mechanism. :-)