<=home 

RS Theremin
01.08.12

True Serendipity

This webpage will only be interesting to the certain few that seek understanding!

While experimenting with heterodyning, an unwanted yet interesting phenomenon I don't immediately understand developed. There just might be another method to make use of the heterodyne principles. At the test bench I was not getting the results I set out to accomplish. On a particular experiment I was doing this unwanted modulation method pestered me for two days. Out of frustration I began to wonder if maybe there was something useful here and quickly realized certain unusual events were occurring. The experiment begins using two standard transistor Hartley pitch oscillators very similar to the Theremax except operating at 850 kHz. That is 7 basic components for each oscillator. 

My question is what type of modulation is this? On the oscilloscope I see a 30 kHz carrier frequency taken from the normal theremin audio output jack. In all my theremins including the vacuum tube model I use single diode detection. The 30 kHz frequency passes right through the detector diode as if it did not exist. There is an ever so slight compressing and stretching of the carrier wave with normal pitch field hand movement. This appears to be frequency or phase modulation?

Until I properly demodulate the audio from the carrier the final sound will be a bit of a mystery along with other unusual features. FM demodulation is something I am unfamiliar with.

Below is what the audio wave shape looks like though it may be flawed. The sound card inputs the 30 kHz carrier frequency and may be giving a misrepresentation of the wave shape, but I don't think so. This can sound like a  breathy flute, the wave shape indicates odd harmonics!

30 kHz Signal Capture .wav 5.5 meg
best sample for Audacity spread out		 Sound Capture .mp3 323k
over rainbow for linearity test
This sound sample has noise, it is for demonstration only!
The sound is flute like with odd harmonics.

View Board Layout

Measure your pitch field linearity by fist popping, make one octave jumps with a closed fist to an open hand throughout the pitch field. Over the Rainbow is commonly used because of the octave jump and the collection of notes in-between. In a musically linear pitch field your hand has same response throughout the field. This won't make you a good Thereminist but it will significantly help!

Proper demodulation may solve the audio issues, currently the 30 kHz carrier is 2v p-p fed directly into my sound card.

There are negative artifacts in the sound sample but that's not what I'm addressing. The heterodyne method I am using is so responsive that you actually pass through several Null points reaching toward the aerial. There is one dominant range or field of notes, that is what I tune too. The pitch field still behaves mostly as would be expected.  

This passing through of multiple pitch field shadows heard as background birdies may be the downfall of this approach unless I can use some sort of filter. 

Above Audacity captures the approximate 30 kHz carrier frequency.

Below is the 30khz carrier of the above 5.5 meg .wav file spread out. 
The audio modulation becomes visible within the Audacity display.
.

If this is FM?  I need to find the best method for extracting the audio from this carrier.

Observations:  Some are intriguing, these require further study.

Most significant The perfectly linear pitch field octaves tightened if the pitch field was expanded using the pitch tuner pot and the octave count increased in number, this is backwards? (This needs a re-setup and evaluation someday)

Normally expanding the pitch field evenly stretches the equal octave widths and does not add any more or does it add tiny ones near the antenna?

There is no Null Point, going below the lowest frequency 1 hz  the field just reverses and the pitch rises, this can be seen at the beginning of the 5.5 meg wav sample above spread out in Audacity. Amplitude is unchanged throughout the range.

The pitch field appears to be perfectly linear relative to keyboard note spacing without the need for any special antenna configuration circuit. I am using a dangling 18" alligator clip lead to make the above samples. Playing the musical notes becomes more intuitive rather than muscle memory. This opens up the possibility that current interpretations of why linearity comes about is seriously flawed, certainly my own theory! Maybe research needs to study the combo Electromagnetic Magnetic Field dance occurring between the antenna and "both" the variable and fixed oscillator.

Vibrato in the pitch field is stiff using normal pitch intervals, as if being defiant. This is weird?

The carrier modulation is not the standard amplitude modulation and appears to me as FM or phase modulation.

The triangular audio wave form is consistent throughout the sweep range, no low end pulling or amplitude change.

The 30 kHz carrier frequency can be a sine wave or overdriven through the TL082 OP to a soft square wave without distorting the audio sound.

Multiple theremin pitch field shadow sweeps or birdies are heard throughout the pitch field range, while the dominant one is tuned upon. This may be the downfall of this approach. I still need to do proper demodulation of the audio and filtering.

Antenna sensitivity and response is as good or much better than standard theremin antenna response, farther outside.

A carrier frequency is generated, adjustable between 25 kHz and 35 kHz with the audio signal embedded in some way.

I am curious if thermal drift exists, I did not think of testing for this when set up.

I used an unorthodox oscillator LC coil configuration and  coupling. 

This is an etched pcb, clean no loose wires. 

 Visit Board Layout

Fred Said:  Christopher, I certainly don't think you are "making this up" -

The 5.5 meg wave file for examination is below the the triangle graphic up above. Open this in Audacity where you can view the embedded wave form by expanding the display slowly. Also you will see one hertz.

But I am at a loss to explain it in any way - and I do not understand some of the things you are saying:

>>"On the oscilloscope I see a 30 kHz carrier frequency taken from the normal theremin audio output jack" "The 30 kHz frequency passes right through the detector diode as if it did not exist."  This is accurate and visible at the diode!

Here is where my misunderstanding starts - unless you are a dog, I do not see 30kHz as "audio" (LOL) - and I wonder if "The 30 kHz frequency passes right through the detector detector as if it did not exist."  is actually caused by this 30kHz being removed by the filter which usually follows the detector..  If it was some 'noob' posting this, I would assume that this was the case, but as it is you, I think I must be misunderstanding something.

Being this is the week of revealing research I will put up a graphic of the configuration as they reveal as much as schematics which I rarely draw today.

So, my first question in: What frequencies are the oscillators running at, and therefore what is the actual difference frequency you SHOULD be seeing from the detector? 

>>"The heterodyne method I am using is so responsive that you actually pass through several Null points reaching toward the aerial. There is one dominant range or field of notes, that is what I tune too. The pitch field still behaves mostly as would be expected."  

This "pass through several Null points" leads me to believe that you are seeing the effects of harmonics from the two oscillators interacting with each other - that, perhaps, the strength of these harmonics is sufficient to mislead you about where the 'true' difference frequency really sits.  This is possible but does not explain the phenomenon's occurring.  This unusual oscillation kicks in tuning the LC oscillator coil.

Alternatively, is there some other source of signal which is being unintentionally mixed, NO and producing (probably from its harmonics - and if the unintended signal has a square shape, these will mainly be odd - as will your oscillator harmonics if they are clipped, giving a whole series of "null" points) these "multiple pitch field shadows heard as background birdies".

So, these are the 'conventional' type of explanations I can think of - other "conventional" explanations would include measurement error - you show a "carrier" at 30kHz, which is what I really have a problem with - where is this "carrier" coming from if the oscillators are running at about 850kHz? - Is this "carrier" the result of one oscillator running at (say) 850k, and the other running at 880kHz? ... If this is the case, and your soundcard is sampling at 96kHz, then all hell will break loose and you cannot believe anything you see or hear!  This is the best answer so far!

But all the above probably comes over as disparaging, and this is not my intent - I am just stating the obvious, which I feel sure you must have looked at.. I just have a real problem understanding what you are saying on this page! I show some of the things I don't understand (and this may well be because my brain aint functioning as well as it once did, LOL):

>> "There is no Null Point, going below the lowest frequency 1 hz  the field just reverses and the pitch rises, this can be seen at the beginning of the 5.5 meg wav sample below the triangle graphic above spread out in Audacity. Amplitude is unchanged throughout the range."  This is accurate!

Where is this 5.5M, and how do you view 5.5M on a 96k sampling scope? The 30 kHz signal is fed into the sound card and recorded by Audacity which could be misleading the outcome. 

How are you resolving the frequency down to 1Hz? I cannot do that with a pile of sophisticated equipment (including HP analogue frequency discriminator).  View the 5.5 wav file below the triangle graphic plus you can hear it if you play the 5.5 meg wav file.

The main problem for me is what you mean by "carrier" - is this actually a "carrier" formed from the difference frequency, and are you looking "inside" this carrier for other signals? View the 5.5 meg file and spread it out and the modulation becomes visible, a sample of this is on this webpage.

You show a clean 263.1Hz triangle wave, and you say " The sound card inputs the 30 kHz carrier frequency and may be giving a misrepresentation of the wave shape, but I don't think so. This can sound like a  breathy flute, the wave shape indicates odd harmonics!"

Are you REALLY feeding 30kHz into your your sound card???   Yes!!!!!!   25 kHz to 35 kHz as it is adjustable and the sound card is demodulating the embedded audio signal, not ideal but it works until I find a better method.

What ever is happening creates perfect linearity heard in the mp3 sound byte on this page and the octave intervals "contracted" if the pitch field was expanded which adds more octaves to the pitch field. This in itself makes for a good project for that adventurous youngster out there. . . but they don't exist anymore. We are all becoming old men! LOL

Fred.         Visit Board Layout

The pcb PDF and more info is available at no charge for this experiment, just email me.
This same pc board is used for my enigma universal stand alone volume control or pitch only theremin.

Email: RS Theremin  Christopher

.