This could be a fun thread to keep going, on these sluggish summer days, because it draws input from individuals coming from different technical perspectives.
[i]I'm simply pointing out that sometimes it's a good idea to stand back for a moment and view the instrument for what it is - a touchy, VLF radio receiver, rather than a musical enigma wrapped in hype and intrigue. :)
And how does one fully mature a stream of electrons?[/i]
Jim, the authentic theremin “voice” I pursue develops in the electron stream within tube/valves; it grows, is born and then disciplined with a follow up circuit. I am not talking about age old guitarist tube amp arguments much less the even/odd harmonic debate. I don’t think you have explored any meaningful theremin discoveries or you would have glimpsed theremin enigma wrapped in hype and intrigue. This may be why you decided to avoid the natural theremin voice… you can’t find it! )-'
Edit: I can't play a tune but that does not stop my research!
My Hybrid Sound (http://www.oldtemecula.com/theremin/sounds/smooth4.mp3).mp3 160k
Hi Thierry -
First, I misunderstood the original posting.. On re-reading it I realized that additive synthesis (as in, being able to add individual harmonics in a controlled manner) was not what was being asked for.. What was being asked for was a lot simpler.
As for analogue frequency multiplication, I have found the costs of this approach was a lot higher than the PLL approach - for what I wanted to do - but for the application here, one probably doesnt need multipliers at all -
Let me explain what I was aiming for.. I wanted my theremin to have additive mixing of the first 16 harmonics, and was also looking at making this 'engine' compatible with CV input (ie, make it a VCO) so that I could also sell it to the synth market - there were two 'simple' approaches - one was to generate VFO and REF signals with complex waveforms and to filter these through a bank of BP filters so that the harmonics were independently available, then to scale these outputs for consistant levels, then to adjust each level to 'taste' and mix (not heterodyne) them, then take these mixed waveforms (one fron the VFO bank, one from the REF bank) to a 4Q multiplier for heterodyning.
I managed to build a prototype board with control of 6 harmonics, but the problems were enormous - everything - filters, VCA's, mixers were all dealing with frequencies ranging from 250kHz up to 1.5MHz .. pushing up to 16 harmonics would have taken the top harmonic frequency to 4MHz - A low frequency for digital circuits, but a bummer when one wants clean audio from the output and one is dealing with analogue circuits at these frequencies.
I then went the "digital" route.. When I say "digital" this is not entirely correct.. "Mixed signal" is more accurate - PLL's contain both 'digital' and analogue components... What is the square wave output from a Theremax? is it digital or analogue? - Anyway, the circuit was greatly simplified and a lot more stable, sounded far cleaner (in fact,completely clean), and was a lot cheaper..
In the end, however, the complexity of the user interface was mainly what made me abandon this idea.. even for 6 harmonics, there were (minimum) 6 controls to set harmonic levels - but to make it useful, one needed frequency dependent control over each harmonic so that the spectrum could change dynamically as the frequency changed - this added another 6 controls.. Then, for full versatility, one really needs volume dependant control - another 6 controls... Minimum 12 controls, optimum 18 controls for just 6 harmonics!
To make a usable product I needed to go for a digital user interface with presets so that a "timbre" could be created and stored - This involved a lot of development.. I had this on the 'back burner' - I put it aside while I got on with my (2) theremin product I planned to launch at HO2011, with an optimistic hope that I might have enough spare time (LOL) to build a prototype in time for HO2011.
As they say, the rest is history.. I am only now able to sit at my desk and 'work' for a few hours, and have absolutely no idea what the future holds for my boxes of 'junk' or my reams of designs.. Apart from which I am broke.
[i]"analogical brain"[/i] LOL! Love it! - Mixed signal perhaps? ;-) Oh,and congratulations on your well deserved position at TW!
In Memory: It was not to long ago I would have conversations with theremin enthusiasts like Fred Nachbaur and Maxie Baars. People like them were a rare kind “sharing” with webpages the theremin knowledge they had gained.
This is what I like about FredM and his continued development of the Element 14 database or Thierry’s selfless circuit analysis.
It is no secret that I do not have formal electronic training, rather a mishmash of disconnected knowledge gathered from years of gadget building. I have an original theremin design in the works, some of you may have peeked? What is difficult is shaping the project to be very builder friendly. Two questions to all engineers:
I use 4 - 3” x 6” and one 3” x 3” pc board, I don’t know an economical approach for getting these made commercially. (Could be $1000 per set, don't know?) I now etch these for myself with my own generated Gerber files displayed in GCPrevue creating iron on PDF’s. This keeps the total theremin build under $200, but what approach is economical for other home builders to use?
My pitch section uses two tube/valves while my RF volume control is simple heterodyning, using two bipolar transistors…I think it is wise to use two transistors that come from the same batch# in order to match characteristics to minimize thermal drift issues. In a properly biased bipolar transistor oscillator circuit Beta should not have much affect… but with theremin sensitivity you never know. Should the two transistors also have a matching Beta/hfe characteristic and even matching inner electrode capacitance numbers?
You haven't to match the oscillator's transistors by beta or hfe, it is more important to keep them at the same temperature, so that if they already drift, the drift at least together. A very fine solution for this problem can be found in the tVox tour theremin: There is a copper shield between the oscillators in order to reduce coupling and the transistors are fixed from both sides onto this copper shield with thermal compound.
Most PCB suppliers have on-line quoting - you can enter the board details, quantity etc, and an automatic quote is generated.. Do not be put off by the appearance of some of these which lead one to believe that you are actually placing an order - you can go through the whole process, get the data you want, then terminate.
For reasonable quantities, I have found Chinese suppliers are lowest cost, and quality is usually acceptable. I used Quick-teck (http://www.quick-teck.co.uk/offer/offer1.php) for my HO2010 boards and they were good quality (some slight smudging of the component overlay print - but still fully acceptable) and extremely low price. I have also used many other suppliers and not had problems with either quality or delivery. Ball-park, for 10 double sided 100mm x 100mm standard boards on 10WD delivery, I expect to pay about £100, for 100 or more boards I would not pay more than £2.50 per board - Economies of scale are enormous when it comes to PCB's!
Where things get really expensive is when one wants fast turn-around.. and here it is worth making your own PCB's. I use P+P Blue (http://www.techniks.com/) transfer paper which one can print a lazer image onto and transfer this to the PCB - one needs to invest in a GBC laminator if one does a lot of boards and require consistant results (I use a modified laminator where I have slowed the transport motor electronically) but you can use an iron for occasional boards. Once the image has been transferred, one can etch the board.
Double-sided boards can be difficult with the above method - I use 1mm thich board for the solder side, and a seperate 0.2mm flexible board for the component side - etch both, and then glue them together using epoxy so that the pads line up (when the boards are etched, it is easy to line them up over a light source) - Once the glue has dried, the board can be drilled (drilling is the most time consuming aspect) - Also, the holes will not be plated, so you need to design your board carefully to allow you to place thru links where needed, and so that connections (between upper and lower boards) do not need to be made at innaccessible components.
With regard to transistor matching - I have found that using transistor array IC's solves all the problems - the transistors are on a common substrate, so thermal difference is tiny. The CA3083 has 5 NPN transistors (two of which are closely matched) and is cheap.. You can use the matched pair for one oscillator, and any other two for the other oscillator, and there is one spare which you can use to monitor the substrate temperature and act as a heater to keep the substrate at a temperature higher than the circuits maximum ambient temperature if one wants to be extreme.
One issue of recent concern was how practical designing around US grounded 117v power is, not realizing there is a market of international cost effective voltage adapters. This adapter product looks good to me Adapter (http://www.voltagedepot.com/Voltage-Converter-Step-Up-Step-Down-100-Watts-Tans-p/st-100.htm) At the most my design uses less than 25 watts @ 117v and must have three prong earth ground access!
My setup allows me to switch between 12.6v or 6.3v filiments. This should make tube/valve choices readily available for some time.
My design approach is modular so the first pcb could be made by a high school student for a theremin project on a 3" x 3" pcb that plays over the AM Radio and then in the future add a second circuit board for a pitch only theremin with a line-level output into a power amplifier.
What's special is these same two pc board layouts together with all the parts installed becomes the ideal stand alone volume control for all those pitch only theremins out there for less than $50 usd.
The volume control above is used in conjunction with my two tube/valve pitch oscillator boards along with my pre-amp board. The output is line-level and adjustable with an independent pitch preview output. All of these pc boards have been designed and demonstrated on my excellent working theremin model.
There is still a lot for me to organize, I am retired and work on this daily, and as FredM and I know there are thousands of endless details that must come together for a final project, not to mention 10’s of thousands of dollars spent looking for the best solutions. Then along comes that sacrilegious kid who yells out, “I want to build a theremin but I want to build it cheap!”.
I will not order commercial pcb’s until I have verified my work with that special theremin engineer captured in that state of euphoria we all begin with, if he builds it I can get feedback for documentation. I would even etch his boards if he drilled his own holes. It would be an advantage if I found him local, in So. Cal.
Once I have a theremin model finished I would like to get a master player to at least record me a quality .mp3 sound byte and give me an evaluation of this theremins behavior.
My friend Touchless has mentioned you before and how beautiful you and your singing are. I have never been east of Texas so New Hampshire would be a nice adventure. I worked for a company (20yrs) called Nashua which might get me some bonus points once I cross the state line. The title "Master" is a powerful word, I think it is bestowed upon someone like knighthood with recognition from your peers.
It does have a nice ring though, but what about Thereminista Virtuosa Amey? (-‘
In my new theremin design the pitch oscillator section powers all the boards by taking advantage of the 12.6v filament section and using a LM7809 for voltage regulation. My stand alone volume control receives power along with the audio signal through a ¼” stereo cable. I have the ground on the sleeve, the audio on the ring and +V on the tip which avoids shorting issues or accidental mono plugs.
I also put a LM7809 on the volume board for when it is powered independently by a 12v battery and proper ground.
Data Sheet (http://www.fairchildsemi.com/ds/LM/LM7805.pdf)
Normally the pitch section regulates the 9v power arriving through the stereo cable and bypasses or places power on the otherside of the 7809 on the volume control PCB; the extra regulator is not needed for normal setups.
This creates a condition with the now unused 7809 on the volume control board where pin #1 floats, pin #2 is tied to ground and pin #3 has +9 volts trying to pull current in the wrong direction?
The question: Could this be harmful to the 7809 or would it behave like a diode with its polarity reversed? Data sheets confuse me. I could set it up and try to smoke one!
What is the minimum ma current needed for proper regulation using a 1 Amp LM7809?
Thanks for your feedback