open.theremin

There's an open source theremin design I haven't seen before here:

http://www.gaudi.ch/DigitalSuper/open_theremin.php

You can see the more sophisticated version (i.e. with volume loop) being demonstrated here:

http://www.youtube.com/watch?v=PEJVFd66zxQ

I was wondering if anyone here had encountered it before, or if they would care to express an opinion on the design.

Gordon

The schematics on this website are too small for my eyes and I'm not so experienced in the digital domain. I think that FredM could give more qualified commentaries.

It seems, that I cannot see the entire page.
Circuit board lay-outs and parts lists are hidden.
Perhaps, we have a Capitalist on our collective hands?
Good Luck!
teslatheremin

[i]"I think that FredM could give more qualified commentaries."[/i]

Thank you, Thierry ;-) .. I am trying to avoid spending much time here these days - Want to get Gordon his Enkelaar back, and get a Theremin to you next month.. and I have others who are running out of patience rapidly and justifiably...

What I see / hear on youtube does not fit with what I see on the schematic - The youtube video shows a reasonabley playable Theremin.. The schematic shown should not produce such a Theremin..

Remember.. WHAT YOU SEE / HEAR ON YOU-TUBE IS [b]NOT[/b] THE THEREMIN PRESENTED IN THE SCHEMATIC! - This is not just a matter of a missing volume antenna - it is about linearity and tone as well.. Whatever is being played, it has little to do with whats on the schematic..

Also - is this really "open source" when all the essential details are hidden? Even the schematic given does not include readable component values.. If this board couples to a processing board to make the Theremin shown on You-Tube, will the full schematics and firmware for both be made available? The first indications incline me to thinking that TT might be right!

I could not get the image large enough to fully evaluate the schematic - However..

1.) UPDATED:

Pitch reference oscillator is not based on a 4046 as I previously thought - it is based on a 4060 oscillator / divider.. The oscillator is fixed frequency based on crystal or ceramic resonator, I cannot read the frequencies.

2.) The variable (pitch) oscillator seems to be a simple RC oscillator using a 4046 PLL VCO, and using the CV input to control tuning.. This COULD be used to provide linearity correction (one of the methods I use) by feeding a correction voltage back after processing the frequencies.. but I do not see any connections to facilitate this from an add-on board. In fact, I see no way that linearity correction can be applied without modifying the circuit.

3.) Reference and variable oscillators are combined ("heterodyned") digitally using an exclusive nor gate (XNOR) whos output is fed to an active low-pass filter / audio signal amplifier.

Conclusion:

There is nothing new, original, or noteworthy about the bits of the design I can see - Linearity will be utterly abysmal, and this 'design' is expensive and complex for what it achieves.. If one is happy to use CMOS digital IC's to build a Theremin, you will get equally good results using a hex inverter at <1/6th the price of a 4046.

FROM WHAT I CAN SEE (and the diagram is hard to read) one would be far better going for one of Art Harrison's low cost designs than opting for this one.

The ONLY (marginal)'clever' bit I can see in this circuit is the use of the VCO CV input as a tuning mechanism.. But, one could have implemented this using the CV input of a TS555C timer which is smaller, cheaper, and more stable (The 4046 VCO is designed for use in a closed loop system, and in this circuit it is being used open-loop, so I doubt that stability will be good - The TS555C is sure to be as good or better)

This looks like yet another piece of Theremin timewasting.. Oh yes - I am guilty of having published a "Theremin" circuit of equal "merit" many years ago, when I first started exploring these instruments, and before I had any sense of the difference between something which gave a changing pitch as a function of proximity, and a Theremin which is a musical instrument to be crafted -

OK. I won't worry too much about this then.

By the way, here's the circuit diagram (http://www.mechatronicart.ch/diymakeaway/wp-content/uploads/2009/11/OpenTheremin_Schema.pdf) as a pdf.

And here's a video (http://www.mechatronicart.ch/diymakeaway/wp-content/uploads/2009/11/OpenTheremin.mov) of it without the volume control etc.

I found these here (http://www.mechatronicart.ch/diymakeaway/?page_id=473).

Thanks Gordon, that makes life much easier!

my opinion remains unchanged - Art Harrisons Minimum Theremin (http://www.theremin.us/100/minimum.html#SCHEMATIC) is a lower cost, less complex design, which will match or surpass the performance of the "open.Theremin".

Even the only +ve feature I thought the open.theremin had, it doesnt have! I thought tuning was implemented by using the 4046 VCO control voltage - it isnt.. Tuning is implemented by changing the VCO TC resistance - this is a high frequency bit of the circuit which is connected to a potentiometer - always a bad practice.. One only does this if one must [b][Edit: The following is wrong]->[/b][i](as with Art Harrisons circuit, where there is no simple way of implementing voltage control)[/i][b]The following is correct ->[/b][i] The Art Harrison Minimum Theremin cleverly applies voltage control to provide a tuning potentiometer which does not carry HF signals - this CV is applied to the supply pin of one of the inverter packages - another preset potentiometer does course tuning in the HF signal path.. This is a clever way to give voltage control.[/i]

Using a VCO makes sense, because this allows one to controll the frequency from a voltage - it allows one to have a potentiometer which is not carrying high frequency signals, and therefore can be wired as one pleases. It makes no sense to have a VCO and not use the CV.. one may as well use simple gates as Art Harrison (and many others) have done - or to use a TS555C timer (which has the facility to add a tuning potentiometer which does not carry HF signals).

The open.theremin is wasteful - the 4046 and 4060 are under-used overkill.. The crystal removes one source of drift, but direct use of a 455kHz ceramic resonator and a single CMOS gate would give a stable reference frequency - and 455kHz is directly usable.. But, to be honest, in a circuit like this fixing one possible source of drift while ignoring the much more major 'problem' component (the potentiometer) is pointless. A whole package is used to get one XNOR gate, and an op-amp is wasted.

The above is all fine for a quick one-off lash up - But is not worth putting into production.. there are much better, simpler and cheaper circuits available.

Hello ThereminWorld

Thank you for your contributions to the Open.Theremin project.

More Information is available here: Open.Theremin (http://www.gaudi.ch/OpenTheremin/)

Hi Gaudi,

Real neat PCB's !

You mention noise (50Hz) being a problem with RC circuits.. This caused me a problem some years ago using CapSense (an RC MCU based system) - There are some solutions.. The easiest is having a HPF on the antenna circuit.. The best is having a LPF to extract and invert signal with frequency below about 20kHz.. Then add this to the signal you get from the antenna.. This way you cancel out the LF stuff.

But, truth is, its not really worth bothering with RC.. simple LC's are much more reliable, have the advantage of higher antenna voltage, and are easily coupled to linearization inductor/s.

Thanks. Good to know. Based on your comments I thought about building a module based on the MinimumTheremin. LC is just much cleaner - just have some problems with temperature drift on the cheep SMD coils. Andrei Smirnov suggested to use two SMD coils next to each other for reference and variable oscillators... should be nice. Also thought about using the AD7746 to build a completely digital theremin - if possible. Gaudi.

[i]"Also thought about using the AD7746 to build a completely digital theremin - if possible." Gaudi [/i]

Forget the AD7746 or any similar "solution" Sure, they have great resolution (up to 24 bit) and capacitance sensing range is ideal for Theremins.. BUT.. look at the time required to capture at that resolution!

Update rate: 10 Hz to 90 Hz (90Hz = 11ms = 15 bit resolution [just good enough] - but even this is way too slow) for 17 bit (one really needs >=16 bit) Conversion time is 62 ms .

Ok - I dont use SMD.. Getting good inductors (particularly adjustable ones) has been a challenge for me (oh, there are plenty - but few with required thermal coefficient) but even so, LC oscillators have proved to be better.. best being a XTAL reference oscillator (using a XCO crystal which can be tuned slightly with a Varactor, and the frequency then divided down digitally) and a good LC osc for the VFO.

Fred.

--> Added -->

When it comes to low-end Theremins, please don’t think I am in any way endorsing the Harrison designs – I do not like RC oscillators, and I think that use of CMOS gates to produce Theremin RC oscillators is a bad choice.. I merely highlighted the method Art Harrison used for tuning..

If one wants to go the RC route, the TS555C timer IC is far better – tuning can be implemented directly on the IC’s CV input, and the IC’s Q output can be taken directly via a resistor to its THR and TR inputs which are connected to the antenna (forming a capacitance that controls the output frequency) – The simplest possible Theremin oscillator, having fixed amplitude high Z triangle wave out, and a low Z square wave out which can be taken to an XOR.

BUT – Even the above is abysmal compared to the performance one gets from one of Smirnov’s single transistor LC oscillators.