VFO / Reference / Volume frequency relationships (?)

"Now, back to our reason for being here in the first place... theremining!" - Jason (on spam thread)

Yeah - Great idea!

This is an open question - Any input / thoughts / advice most welcome..

Ok - With an analogue theremin, the reference oscillator and variable oscillator frequencies are bound together in that, at the "null distance" they are the same.

Volume oscillator frequencies (for volume circuits using a fixed frequency) are usually different (higher) frequency than the reference and VFO frequencies.

Digital theremin designs and other designs where the reference and VFO frequencies need not go to "null" but can be spaced apart to provide a "offset" frequency seem to sometimes have other rules..

I have used the reference oscillator  directly as the volume oscillator (on my designs the reference frequency is fixed and tuning is performed by adjusting the VFO frequency - my design would certainly not work for a theremin where tuning is done on the reference oscillator, as it is with the EW)

Question:

What are the "standard" ratios (if there are any) between the volume and other oscillator frequencies? Is there any scientific / mathematical basis (other than that one doesn't want the volume oscillator to beat with the reference or VFO frequencies to produce audio difference [ghost] tones, or any harmonics of these frequencies to produce ghost tones) for the selections ?

I want to get some boards made - My lab is dysfunctional (and I have virtually no access to it) so my prototyping has used only basic portable test kit - For this reason I am a bit worried about committing to a route that seems to work and seems real easy, but which no-one else seems to use - this usually means there must be a good reason!

But I cannot see a problem, nor can I see a theoretical problem... My volume antenna is radiating the reference frequency, so even if it mixes into the pitch antenna, no ghosts will be produced.. The only effect I THINK I can observe is a tiny degree of oscillator 'synchronization' that occurs if I am badly grounded and difference frequency is low (<100Hz)

But any ideas on what might go wrong would be most helpful - and any theoretical basis for chosen frequencies would also be helpful.

Fred.

BTW - Both oscillators are series LC with 60V P-P, operating ~500kHz. The volume 'antenna' loads a series LC BPF of almost identical design to the oscillators. Detection is performed by phase comparison of the drive vs antenna signals. Volume antenna amplitude is ~60V P-P (about 6db drop when fully loaded) but I am not using amplitude I am comparing phase at zero crossing.

Drive voltages for both oscillators and BPF are 5V, sine current in BPF inductor is ~10 to 20mA P-P. I have an identical BPF with fixed capacitor connected to the VFO, and (with phase detection) use this to produce a linearizing control signal for the VFO.

I dont trust myself at the moment - but will publish the design IF when im sure im not doing anything stupid and have proved it fully works..

"What are the "standard" ratios (if there are any) between the volume and other oscillator frequencies? Is there any scientific / mathematical basis (other than that one doesn't want the volume oscillator to beat with the reference or VFO frequencies to produce audio difference [ghost] tones, or any harmonics of these frequencies to produce ghost tones) for the selections ?"  - FredM

Thierry could answer this best I think.  I've only performed one quickie interference experiment with HV LC oscillators (Clapp FET with tens of volts p-p at the antenna).  There I found two oscillators running around 1.7MHz and placed about 400mm apart were OK looking on the scope if they were more than 6% or so different in frequency.  At 100mm this interference zone increased to +/-20% or so. 

In operation the VFO can vary quite a bit in frequency, and this I believe is why the volume section is generally run at a higher frequency than the pitch: it is normal to touch the volume antenna during play, which brings its VFO way down in frequency, but at the same time the volume is being muted, so the interference issue becomes moot.

"Both oscillators are series LC with 60V P-P, operating ~500kHz. The volume 'antenna' loads a series LC BPF of almost identical design to the oscillators. Detection is performed by phase comparison of the drive vs antenna signals. Volume antenna amplitude is ~60V P-P (about 6db drop when fully loaded) but I am not using amplitude I am comparing phase at zero crossing."

If I'm reading you right, the pitch antenna frequency is then is slightly below the volume frequency antenna frequency, but any interference is kind of already there in the first place via heterodyne mixing and other forms of interference due to oscillator physical proximity on the PWB, ground sharing, etc.  I might worry about this a bit if using a parallel tank where the drive is kind of wussy, but you seem to be using low impedance series drive (relatively high inductor currents) where it might not matter as much.

"I want to get some boards made..."

I envy your can-do spirit!  Whenever it seems to be nearing hardware time I find myself procrastinating like a mofo.  Did you ever find a good, inexpensive small run board fab?

"If I'm reading you right, the pitch antenna frequency is then is slightly below the volume frequency antenna frequency, but any interference is kind of already there in the first place via heterodyne mixing and other forms of interference due to oscillator physical proximity on the PWB, ground sharing, etc.  I might worry about this a bit if using a parallel tank where the drive is kind of wussy, but you seem to be using low impedance series drive (relatively high inductor currents) where it might not matter as much." - Dewster

Yeah - Pitch oscillator frequency is (except at null) always below reference frequency, and the reference frequency is the volume frequency.

I dont see any problem with volume interfering with the pitch oscillator (other than syncing the VFO to the reference and limiting the bass end if there was extreme coupling) - I was a little worried that the pitch (VFO) coupling into the volume antenna might produce inter-modulation, but haven't seen signs of this (this is probably my biggest concern).

I really dont want to add a separate tunable oscillator for volume - Prototypes and on-offs I have used the above topology on have all worked and seem problem free.. But they were a lot less critical / important than my present projects.. This one (if things go to plan) is destined for high volume (by theremin standards) production.

I need to keep cost of modules low, but cannot risk problems (its "ok" to manage problems and sort them out if the quantity is low, say <100.. but above that there's no room for risk) so I am also looking at producing a volume frequency which is an integer multiple of the reference frequency - I have PLL's available, and would only need swap the logic low cost dividers for a slightly more expensive PSoC to get any integer multiple..

"I envy your can-do spirit! "

LOL ;-) ... This is pure desperation! - I go to court on the 15th and 25th, representing myself because I cant afford a lawyer let alone barrister - Its likely my home will be sold to pay my wife's legal fees.. I literally have one week before my fate is decided, if it doesn't go my way I will be homeless by December (unless I can generate some income).. Sometimes I wonder which is more pointless, fiddling while Rome burns, or trying to extinguish the flames with a bucket of water... I have one interested party (regarding manufacture of theremins and theremin-related instruments and modules) - Odds on it coming to anything are probably the same as if I spent the money on lottery tickets, but what have I got to lose? -  All I can do is try to salvage my life so that I am in a position to support my children when they need me.

Envy is misplaced - ;-) .. You wouldn't want the "motivation" that's driving me, I assure you!

"Did you ever find a good, inexpensive small run board fab?"

I did, but they stopped trading - Chinese manufacturers are a bit of a lottery, but usually ok if you dont place the order too close to their new year! - leave a good month or two clear - I was let down badly being promised the boards from a Canadian fab, but it turned out to be Chinese.. Turns out that folks in the know order well early, because all non-Chinese fabs are over-loaded during this period (and more expensive) so loads of orders go in for a month or two before the holiday, and those who haven't been wise overload non-Chinese fabs during this period.

I now use is a local Chinese agent who uses whatever fabs are available (anywhere in the world, but mostly China and India) at the best prices he can get - prices (to me) are not the lowest, but I can rely on quality and delivery.

Small run is the problem - my boards tend to be about 10cm x 8cm, and < 10 boards just makes no sense economically - but 10 costs too much if one ends up changing the design and only using two.. So I usually etch a couple of boards first at home (have constructed a toner-transfer "laminator" and have a small "fab" for knocking prototype boards together quickly) - I dont have access to my "fab" at present, which is why I am nervous about committing to boards I have only tested on pad-board with a DMM and portable 'scope.

- And why I am inquiring about frequencies / relationships etc.. 

Fred.

Been looking at boards again -

Prices have come down a lot, I think I will go for This at £36 for 25 boards (9.75cm x 22.8cm 3.84" x 9"  DS) its the cheapest i've seen PCB's (this is inclusive of postage to UK).. Can get 10 off for £33 but £3 for 15 boards? gotta have them! (222 sq cm / board, 10 boards = £3.30 / board, 25 boards = £1.44 / board)

There's Botech who I have used before (or at least used the same fab via my contact) who made good quality boards, 10cm x 10cm x 10 boards for $30 + $20 postage to UK and USA... Works out a $5 ( £2.90) per board (100 sq cm ea)

Ebay prices tend to be about as low as one can get for small quantity - But you can sometimes get better prices from Ebay sellers - I got a 20% reduction on my 20 H1 boards because the Chinese agent was interested in theremins and I invited him to come to HO-2010 and photograph the boards in action - (which he never did).

Fred.

Fred, thanks so much for those links!  Though by posting them you're actually doing me a disservice - by limiting my dawdling excuses! ;-)