Basics of EW-Pro.. #2 (and related issues)

"CV Tracking does not operate well below 100Hz, so tracking of bass notes would be difficult" /



I said: [i] "I suspect that the filters / modifiers must therefore be locked to some fixed CV when the pitch drops below 100Hz.. meaning that tones produced in the bass register would be processed through fixed, and not dynamic processing" [/i]

Gordon asked: [i]
"Unless the CV is generated before the frequency is divided?"[/i]

Yes - generating the CV before division would solve the problem (Or after multiplication - I discussed this idea in a prior thread CV Generation (http://www.thereminworld.com/forum.asp?cmd=p&T=4157&F=3)).. However, If this was being done in the E-Pro, one would expect the E-Pro's CV output to be stable when it was playing audio below 100 Hz..

My understanding it that the CV output has the same problem that the EW+ pitch to CV output has - this was the reason we discussed a voltage shifting circuit in a prior thread ..

IF it is true that the EW-Pro CV output does not work well below 100Hz, then the implication is that the actual final audio frequency is used for pitch to CV conversion - which would be extremely puzzling - with a higher frequency source signal to convert, it does not make sense to use the lower Frq signal for pitch-CV.. If, however, the CV is stable below 100Hz, then everything makes perfect sense.

Two things would not make sense to me:
1.)Having a seperate pitch to CV for the internals, and (an inferior) one for the CV output..
2.)Not using a higher frequency for pitch-CV if such a frequency was available.

There are, no doubt, other mysteries under the bonnet..

Coalport wrote: [i]When the "garage band" sound became fashionable in Rock & Roll in the 1990's, groups would spend a fortune in the finest and most expensive recording studios in the world, attempting to make those studios sound like some kid's basement rec room. Sophisticated FX chains were devised to make digital recording sound "dirty" and "home made".[/i]

Ah, now I can guess the reason for this particular insanity. If they did it the obvious way and just bought some cheap gear, they would get a visit from the record company accountant, and he would say, "Hey guys, you underspent your annual budget by a hundred thousand dollars. Well done. Now I'm going to reduce your budget for next year by a hundred thousand dollars, because you clearly don't need it."

Fred wrote: [i]Or after multiplication[/i]

You keep mentioning that and it rather puzzled me. Division I can see, but multiplication? So I did a bit of Googling.

Good golly, phase locked loops are a whole discipline all by themselves! The sort of thing people write great big expensive books about, chock full of equations. Basically it's a circuit that will home in on a frequency by successive approximation, and I guess a fair chunk of the theory is about choosing the right one for a given application, and getting it to home in jolly quickly and then maintain its lock with a fierce determination come what may. Yes?

Frequency multipliers existed already a long time before the PLL and cheap CMOS circuits came in use. There were circuits which multiplied the input frequency by 2,3, or even 5 with one triode stage and a resonant circuit. And such circuits could be cascaded...

@Fred: Stop the guesswork and review your EPro analysis. There is no link between any CV and the Waveshaping/Filtering section. You opened Pandora's Box, so do it in an exhaustive manner.

No withdrawal! ;-)

[i]"The doors to the Holy of Holies of the E'Pro theremin have been ripped asunder and the contents of the tabernacle have been revealed for Philistine eyes to gaze upon!" - Coalport[/i]

ROFLMAO!!

The E'Pro is a goddamn musical instrument! Like all musical instruments it is constructed from available components, uses known science, has no mystical or supernatural components or energies.

Chobbs quoted Arthur C. Clarke [i]"Any sufficiently advanced technology is indistinguishable from magic."[/i]here.. (http://www.thereminworld.com/forum.asp?cmd=p&T=4544&F=3)

However, the "problem" is that the E-Pro is NOT a "sufficiently advanced technology" to merit this "magic" status. It is not "magic" - More than anything else IMHO it is [b]BOLD and BRAVE[/b] (or at least its Creator was).

There are MANY who knew the secret - in fact, any and every engineer or technician who has ever needed or been interested enough to look at the E-Pro pitch board would have seen the secret..

The trouble (to me) comes not because they did not disclose the information to the "Philistine's" as much as the fact that the "Philistine's" were being misled by these self-appointed "keepers of the secrets".

Yes, most (if not all) of these "keepers" were doing this in the belief that they were serving a greater calling -

The major thing which pushed me to "rip asunder" The doors to the "Holy of Holies" was the 'disclosure' that this "Holy of Holies" contained a sample playback mechanism of some kind.. This, to me, was utter blasphemy! ;-) I would have been more gutted if I had discovered this to be true, than discovering that an XOR was on the tabernacle.

Taking the "tabernacle" analogy further - the idea that it was "empty" is, I thin, wrong.. It was not empty - it just didnt contain what we expected to see.. There were no gold or jewels, there was 'just' grain and water.. There was no new analogue method of heterodyning which allowed ranges to be scaled (as "implied" by the block diagram in its manual)- there was a digital range selection and digital heterodyning..

But - IF one gets past the shock, one can see that the ideas underpinning these 'contents' and the fact that from these basics the Epro was created, this could be worth more than finding gold and diamonds.. Because it FREES us explore technology we would not even have looked at before.

I do think that earlier disclosure may have had an adverse effect on the E-Pro acceptance - but it has now become accepted as superior Theremin (IMHO possibly the best Theremin ever built, bar the Ether Vox - although I have never had my hands on an Ether Vox.. I am taking this opinion from someone I trust - I get my chance to work on an Ether Vox next month).

I feel that this disclosure now can only benefit Theremin development.

I have had some rather unpleasant emails - non from any TW member as far as I can tell.. One, which I think was probably a joke, had a section which went along the lines (all capitalised - I paraphrase)..[i] "I bought 4 EW-PROs as an investment, and planned to sell them boxed and unplayed,for a large profit, when I went into retirement.. You have put my plans at risk![/i]

I emailed him back, asking what price he would accept for them - and cheekily suggested that if he was not planning to retire soon, he could always sell his Epro's now on Ebay, and invest the money in Fundamental Designs.. ;-)

That depends what kind of magic you're talking about. Remember that stage magic is 90% audacity, and clearly the ePro is [i]sufficiently[/i] advanced; people love it.

Note to Fred's mystery emailer. Don't be silly. You can't [i]put[/i] an investment at risk - investments are already a risk. Ever heard this phrase before? "Warning: The value of your investment may go down as well as up."

[i]"phase locked loops are a whole discipline all by themselves! The sort of thing people write great big expensive books about, chock full of equations. Basically it's a circuit that will home in on a frequency by successive approximation, and I guess a fair chunk of the theory is about choosing the right one for a given application, and getting it to home in jolly quickly and then maintain its lock with a fierce determination come what may. Yes?" - Gordon [/i]

YES!

Dont let the complexity frighten you - they are actually really simple!

Oh damn! - I can only say that, for frequencies of the kind seen in Theremins (pre heterodyning) where the frequency deviation from the mean is a tiny percentage, PLLs are as easy to use as dividers.. And to get frequency multiplication one uses dividers.. So, take range switching - take the division logic in say an Epro, and use this in a PLL scheme, then - instead of say a /8 one will get a *8..

Which is why I came to putting E-Pro style dividers and range switching logic into a chip - not to make an E-Pro, but to multiply rather than divide.. an "upside down" E-Pro if you like ;-)

PLLs consist of a phase comparator, which outputs a pulse train (or in the case of the 4046 PC 2 a much better 3 state pulse train - high, open circuit, and low).. The phase comparator takes an input frequency and a comparison frequency, and the output pulse train is taken to an integrator.. The voltage from this integrator drives a voltage controlled oscillator.. If the output from this oscillator is fed back as the comparison frequency, then the output of the VCO will track the input frequency, because the integrator will settle on the control voltage to maintain lock.

With the type 2 phase comparator "successive approximation" does not really apply - voltage from the integrator will rise or fall (depending on whether input signal lags or leads the comparator / VCO frequency) and once at the correct value, is effectively held at a sample/hold level which is updated when phases drift even a tiny fraction..

The "magic" occurs when one feeds the VCO into a divider and takes the dividers output back to the phase comparator.. When in lock, the frequency out of the VCO is the frequency of the input signal * the dividers value (as in /16 will give *16)

The VCO can be any suitable VCO.. My VCO consists of a pair of analogue LC oscillators being mixed (analogue heterodyning) - one of these LC oscillators being fixed frequency, the other having its frequency varied from the phase comparators voltage by using variable capacitance diodes instead of an antenna.

Yes - A phase comparator's output (via integration) can directly control ANY theremin by taking control of its antenna capacitance, and can lock this Theremin to *almost any audio input frequency..

Alas, (the frustration this caused me I cannot express - I never saw it at the time I had the idea, but I should have) it is only *"almost".. This idea has exactly the same problem as pitch-CV converters have.. Frequencies below about 250Hz do not update the phase comparator fast enough.

I have overcome the above problem - but I am not planning to tell anyone how! (or at least, not yet! ;)

[i]"@Fred: Stop the guesswork and review your EPro analysis. There is no link between any CV and the Waveshaping/Filtering section. You opened Pandora's Box, so do it in an exhaustive manner. No withdrawal! ;-)" - Thierry [/i]

Sorry Thierry - its not about "withdrawl".. but I simply have no access to an E-Pro anymore, and no intention to acquire one again, and not enough time or inclination to try to deduce the full circuit from photographs.

It is entirely up to those who have knowledge and understanding of the E-Pro to share this, if they choose - or to retain it, if they choose.. its up to them.

What I disclosed was only a tiny piece of the story - the only piece I knew, and even this knowledge is most basic. I know you believe I was absolutely wrong to disclose it - I know you probably regard me as a traitor, and I know you will not understand when I say that (due to my personality, or lack thereof) I had no option but to share what I knew..

I have taken this action - and I am responsible therefore for any concequences of this action.. If my action was "immoral" or "unethical" or "unwise" etc, it is down to me.

Now, however, there is nothing to hold you, or anyone else, back from sharing matters related to the E-Pro with the Theremin community (and particularly the technical Theremin community) - guilt free.

The Epro manual (http://files.centrmus.com.hk/Moog/Etherwave%20Pro/EwavePro-web.pdf) (page 17) has a block diagram which clearly shows the CV being routed to waveshaping, filters and CV out.. Are you in fact saying that this block diagram is untrue?

[i]The Epro manual (page 17) has a block diagram which clearly shows the CV being routed to waveshaping, filters and CV out.. Are you in fact saying that this block diagram is untrue?[/i]

No it isn't untrue. My comment above was too short, unreflected and imprecise because I was thinking in EStandard categories where there is no extra filtering section behind the waveshaping circuit.

It is true that the pitch CV signal is not fed into the [b]waveshaping[/b] circuit which is only influenced by the waveform and brightness control voltages from either the pots or the preset resistors depending on the position of the timbre selector.

But it is fed indirectly into the [b]filtering[/b] circuit and will add (more or less, depending on the timbre selector) to the filter CV which comes either from the filter pot or from one of the fixed resistors (still depending on the timbre selector).

Thierry Said: [i] "It is true that the pitch CV signal is not fed into the [/i][b]waveshaping[/b][i] circuit which is only influenced by the waveform and brightness control voltages from either the pots or the preset resistors depending on the position of the timbre selector.

But it is fed indirectly into the [/i][b]filtering[/b][i] circuit and will add (more or less, depending on the timbre selector) to the filter CV which comes either from the filter pot or from one of the fixed resistors (still depending on the timbre selector)."[/i]

Thank you for that clarification, Thierry.

From this, I provisionally deduce that the CV is not of [b]primary[/b] importance in in the charactaristics of the output tone..

This, however, still leaves a few questions open..

(1) - If the pitch-CV conversion is derived from the frequency which is actually output by the Theremin, then there must be a mechanism (?) which prevents the 'quantizing' effects when audio is at low frequency, from modulating any VC filtering applied to the (bass) audio and messing it up..(?).


(2) As there are higher frequencies available from which the CV could be generated, why was this not done? - If (as I believe from prior threads) the CV goes unstable at bass frequencies, it seems obvious to tap a higher frequency and use this - Even if the FPO and VPO were directly and seperately digitally heterodyned to produce the top difference frequency, and this pitch was converted to CV, the range function could still be employed - all that would need to be done would be to scale the CV (simple control of the gain and addition / subtraction of voltages) to match the scaling applied to the frequencies by their division. Did they just not think about this?

OH! Duh! I think I see it now..
At the bass setting, the audio (difference) frequency is the result of the difference frequencies between the VPO and FPO - There is no higher frequency that can be derived (!?)(?)

No - This doesnt quite add up to the way I thought the ranging operated - but I suppose it is possible...


Fred.