Enhancing The EtherWave Sound

"I did download it but the grayness & graphics seemed clunky, I am sure it works for complicated stuff like precise digital but I am only doing basic analog circuits where ohms law and 5% accuracy of a slide-rule gives me answers."  - Christopher

The FPGA tool I use can do digital a lot better so that's where I do almost all of my digital sims.  LTSpice will work for mixed analog and digital, but I only use it for analog stuff.  And it's not that analog sims are more accurate than reality - if anything they are only a rough approximation, so you have to take their results with a grain of salt.

"I think relying on modeling would have kept the method hidden from me as I would never have had hands on much less made anything physical."

There's a lot of serendipity to be had tinkering around on the bench, though I can muck around in a sim and quickly try a bunch of things in an hour or so that would take me days on the bench, so sometimes luck strikes there first.  For me it's a best-of-both-worlds kind of thing; when I get the sim and reality to jibe then I understand the reality a lot better.  Once you get past the learning curve pain, sims can be a lot of fun!

======

Your waveform looks a lot like clipped 33% PWM.  I bet there's an easier way to get it (sans opto).  The two transistor mixer I posted to the analog thread can make somewhat similar waveforms.  Just a matter of mixing to something rather sine-like, then offset & clip (doable with offset and overdrive hitting the limits).

dewster said: "Your waveform looks a lot like clipped 33% PWM.  I bet there's an easier way to get it (sans opto).  The two transistor mixer I posted to the analog thread can make somewhat similar waveforms.  Just a matter of mixing to something rather sine-like, then offset & clip (doable with offset and overdrive hitting the limits)."

Being you have my schematic you can see what I am mostly doing. Since that drawing I now also hammer the audio with the square wave output from a low power 555 to add buzz.  PWM not quite right but the audio transformer creates the harmonic skew on the top & bottom of the wave shape.

I would like to hear your simpler approach, bread-boarding takes effort, when was the last theremin sound sample you posted?

If it were obvious how to improve sound easily and cheaply, many people would have improved the theremin sound long ago. It is a matter of taste in what is good. I had emails from people when the first EtherWave Pro came out saying they were not happy with the sound, what created the over priced cult following? Only later to learn even the volume control had issues.

My passionate singing lady.wav  EtherWave Standard direct to sound card, no reverb or any processing, no puny whistles.

This previous sample was more reedy or violin like sample.wav  Deliberately punching the low end for bass lovers.

I now believe the modern EtherWave Standard, that already has a good volume response, can develop into the classical theremin of choice. With my harmonic approach Gene Segal might want to resurrect the Wavefront Classic Model.

I may have to ship my first modded theremin to a Master to keep... so they can have time to do something creative with it.

Edit: I purchased a new set of amplified computer speakers and my sound was more like vocal chords outside the throat. You must use speakers that use acoustics to develop the sound, bigger is better or the results are like a violin string stretched between two sticks. My samples are only audio wave-shapes until the first time allowed to physically move air and resonate.

Christopher

Playing around with PWM generation in Adobe Audition 3.0, I can get 6th, 12th, 18th, etc. harmonic suppression by setting the duty cycle to 16.667%.  This maybe makes sense somehow because 100 / 6 = 16.667?  Anyway, I believe that's one of the mechanisms operating here.

What is the waveform coming out of the mixer and what is its peak to peak amplitude?

I have a story to tell, on November 3, 1993 an aberration appeared in my room. I was presented with old writings of electronics from the days gone by. I was also given a pair of special glasses that allowed me to understand what had been written down. My electronics & theremin journey would begin long before my first visit to Theremin World. 

The first words of the writing were almost funny which said: “Like you, tubes are lazy compared to solid state, that is the magic.”

Not mentioned previously in the oldtemecula research is the two buffer transistors are weakly driven along with the detector diode barely conducting. The diode is forward biased with a 1 meg resistor and the signal level is adjusted down to 100 milli-volts p-p.   This is very important to mimic the tube signal characteristic.

There are several areas that probably can not be expressed in modeling.  Serendipity is a good word dewster. (-‘

The image below is the waveform at the audio side of  the single diode mixer, adjusted to 100 milli-volts p-p level. RF makes the image fuzzy.

The image below is taken from the line level output which feeds the sound card, it is 1 volt p-p. The freq for both is around 300 Hz.

A wise-man once told me these words:

"I like the old technology associated with tubes, although they waste a lot of energy in the process of doing what they do.  I have collected a bunch of tubes, components, sockets, testers, transformers, chassis, etc, thru the years and wanted to use some of it to experiment with your hybrid tube design and help you capture that angel's voice that eludes you.. "

My old oscilloscope alternates between traces, can't capture the image in one photo.

To be continued….

Christopher

"My old oscilloscope alternates between traces, can't capture the image in one photo."  - Christopher

Buy one of these for $400 (I would in a heartbeat but my old dual trace keeps putting along):

http://www.amazon.com/Rigol-DS1054Z-Digital-Oscilloscopes-Bandwidth/dp/B012938E76/

(there are hacks on the web to make it 100MHz and enable all of the features)

Dewster said: “Buy one of these for $400 (I would in a heartbeat but my old dual trace keeps putting along):”

My creativity comes from a place of humility, if I owned that scope I would feel too bitch’en to accomplish anything.

For that price I rather mod a new EtherWave Standard and give it away to a budding artist.

Above I mentioned line-level, but actually it is my mod board output, a potentiometer is used to set the proper level to feed back into EWS VCA. Using my mod I would change the EWS waveform knob and keep the brightness control that can add a reedy flavor.

*  *  *  reedy-sound.wav   * * *

Over all my approach has seven adjustments that can influence the wave-shape (sound). I watch a simple freeware harmonic display on my computer to get where I want to go. Less third harmonic is More.

The end home installer does not need a scope, just that $18 DVM from Amazon that has a frequency counter.

I visited another local Meetup inventor’s type of group last week. I always feel embarrassed to say my hobby is about the theremin which most don't know what it is or have only experienced trash. Then it happened, the head of the group (70 yrs old) pulls out an arduino, plugs in a six inch wire and demonstrates to the group what a theremin is, the sound a mousy square wave.  I remained silent after that, how could I compete.  Honestly, I thought I was going to cry.

Christopher

dewster, now you have me curious about those under $100 USB oscilloscopes to use mainly for screen captures if the resolution is acceptable. Do you like any of them? My old two-channel scope works fine for me, though I have always wanted the built in frequency counter.

As I experiment the magic appears more and more dependent upon the sluggish signal pushed across two NPN buffer transistors barely conducting at 1-ma along with the distortion introduced by the mixing diode barely conducting, without using a forward bias resistor. This creates the fatter earthy sound using only solid-state. This sound is muddy without the extra circuitry which brightens the sound with controlled even harmonics.

This project consists of two small circuit boards, both fit comfortably inside the EWS box.

I am working on a webpage for next week that I will post the schematic and PCB layout before I get any circuit boards made. Someone might suggest something I did not think of. dewster already contributed some  good stuff in this area. The bare boards PCB will be given away at no cost to any of you if you can prove past construction skills. You will also receive a one click parts order Mouser list.

My only claim is I am as sharp as a marble with a keen sense of intuition. I found it very encouraging when someone from Lev Sergeyevich Termen’s hometown contacted me this week through my Google Forum Webpage about this EWS modification.

These words below came to me early on in my research:

"Mounds of human heads are wandering into the distance. 

I dwindle among them.

Nobody sees me. But in books much loved and in children's toys

I shall rise from the dead to say the sun is shining."

Osip Mandelstam of Russia 1891-1938

------------------------------------------

Unless something changes I am in fear the classic theremin will pass away to time with our generation. 

For $96 this EWS board and the above $50 upgrade can give you an instrument better than Clara Rockmore's original Lev Sergeyevich built theremin with the classic sound.

Christopher

"...now you have me curious about those under $100 USB oscilloscopes to use mainly for screen captures if the resolution is acceptable. Do you like any of them? My old two-channel scope works fine for me, though I have always wanted the built in frequency counter."  - Christopher

Every time I look into them they seem to promise much more than they actually deliver.  A/D conversion + LCD do not a scope make; a good scope is much more than the sum of its parts.  $400 for a 4 channel 100MHz scope is just incredible, I'd be afraid to pay less actually.  The input amplifier circuitry has to be up to snuff, and the engineering pulling it all together takes some doing, so you can't short change it without seriously hurting the final product.  You want as much bandwidth as you can get so that you don't need to take input rise time and the like into consideration when viewing relatively slow signals.  A scope is your eyes into what is going on and as such should be the biggest investment on your bench.  Triggering modes and deep memory can show you things you would never be able to capture on an analog scope.  A convenient FFT is really handy, as is multiple capture averaging.

I'll make another pitch for simulation.  To me it's rather like an artist sketching before doing a painting.  He/she can try a bunch of stuff out on the sketch pad before committing to paints.  Tens or hundreds of sketches might be generated between paintings.  That's my experience with sims anyway, where the sim is the framework for my understanding.  There's a reason they jump into sims almost immediately with formal EE education, they can amplify your understanding enormously, and they are a tool you'll likely use for the rest of your life.  Simulation is where I increasingly start these days, and once I refine it and really see what is going on I
transition to hardware, where it's more of a minor investigation into the how an why of any differences exist between sim and reality, whereupon I often change the sim (add parasitics, etc.) in order to reflect reality better.  In digital design, sims ("testbenches") are actually used formally impart detailed knowledge about the design, and as proof-of-concept.

I must say I think what you're doing, improving the EW voice via an add-on board, is a worthwhile project.

“I must say I think what you're doing, improving the EW voice via an add-on board, is a worthwhile project.”

Dewster, that is an excellent writing. 

I can not disagree with anything you said as getting to where my wave shape board is today required many re-designs and testing over the years before I understood what creates the classic theremin sound that I like. Modeling may have proven more effective and quicker than serendipity.  (-'

You stated you could get the wave shape or sound I like using a simpler circuit approach, I still challenge you to demonstrate or at least provide the image of the layout taken from a Modeling Screen. The one or two builders left might like to explore that.

My theremin wave shape breaks down into two aspects, the low end (diode) and the upper harmonic side. (Transformer)

The diode seems so important now it might be wise to try other diodes to test their characteristics. I always found transistors with higher voltage ratings created a better sound.

If the diode is forward biased (5v/1M) the signal level goes up x10 and I am left with the normal unwanted healthy audio sine wave.

My journey from transistors to vacuum tubes back to a combination of both (hybrid) to just sticking with the EWS to keep it simple ended up being dependent on that day you shared your idea of using an emitter-follower. Buffering was often mentioned, I knew I needed it, but you were the first to show the concept scratched in pencil on a piece of paper, for me a light went on. Things began to make better sense.

The classic theremin sound is a clean single trace wave shape with the correct bends and bumps in the proper places, it is really that simple. The goal is to eliminate the third harmonic, I have verified this to be a fact. Coalport uses expensive equipment that might be doing this along with a good acoustic method of recording.

I do not know how musicians enhance sound, if it were as simple as using an equalizer to remove the third harmonic why does no one talk about it or better yet demonstrate it. 

I am going to try and record the wave shape as I always do but also have the acoustically developed sound on the other channel. Some computer speaker systems do not handle a raw wave-shape effectively so some might miss the point.

Maybe I am chasing a ghost; which I guess is better than one chasing me?

Christopher

"I do not know how musicians enhance sound, if it were as simple as using an equalizer to remove the third harmonic why does no one talk about it or better yet demonstrate it."  - Christopher

Equalization / filtering will only alter the amplitudes of fixed frequencies.  Squashing things like the third harmonic of a variable pitch waveform takes wave shaping.