Armstrong, Hartley, Colpitts, Clapp, Wallin...

I'm starting this thread to discuss (probably) new kind of oscillator.
Separate thread will help not to miss interesting idea lost in tons of other posts.

Simplified schematics: LC tank + resistor + differential amplifier (comparator)

Is current sensing differential amplifier LC oscillator known?

Last invention of LC oscillator mentioned on Wiki is dated 70 years ago.

Armstrong, 1912
Hartley, 1915
Colpitts, 1918
Vackar, 1945
Clapp, 1948

I couldn't google anything similar to oscillator proposed by Eric (dewster on TW).

The only current sensing solutions I found are related to power consumption measurements.

If it's new kind of oscillator, it looks like a real invention.

Wallin, 2020

First description of Eric's idea can be found in this post of "Teensy 4.0 600MHz ARM Cortex M-7 MCU - ideal for digital MCU based theremin?" thread.

If you are going to have a high speed op-amp in there, you might try differentially sensing the coil drive current, which is 180 degrees out of phase with the drive voltage (the following was just hacked together a while ago with CMOS and could be improved, not sure the BPF is good or necessary, and of course the sensing R can be a Q killer)If you are going to have a high speed op-amp in there, you might try differentially sensing the coil drive current, which is 180 degrees out of phase with the drive voltage (the following was just hacked together a while ago with CMOS and could be improved, not sure the BPF is good or necessary, and of course the sensing R can be a Q killer)
-- dewster

Correction:

I actually came up with that circuit in this post on 12/10/2013.
-- dewster

Post with BJT implementation test report in "Let's Design and Build a (simple) Analog Theremin!" thread.

I believe it's awesome idea.

Stable 400Vpp swing on antenna oscillator is probably the best one for theremins.

I wonder if it could be possible to see in future wiki article https://en.wikipedia.org/wiki/Wallin_oscillator ?

At the time I am working on the measurement set for coils SRF investigations. The main requirement is connecting the coil just via "bottom end". Istead of differential input amp and current sensing resistor I use a current transformer (a ferrite ring coil). Works well, but the dewster's circuit is possible better. Needs try.

The LT1711 is too expensive.

"Stable 400Vpp swing on antenna oscillator"

Not a problem. Directly depends on the pump current.

A too high voltage swing at the antenna might become a problem when several Thereminists want to play together because the higher the voltage the higher the interferences seen by neighbored instruments. 

An Etherwave Pro drives the pitch antenna circuit already with 42Vpp which is then multiplied by ~10 in the series resonant circuit formed by the linearization coils and the pitch antenna. When an EPro works in a room, it’s almost impossible to play an Etherwave Standard/Plus in the same room, due to strong ghost tones and noise. A tVox tour theremin side by side with an EPro will simply be mute because the radiation of the EPro’s pitch antenna will overdrive the volume phase discriminator of the tVox.

That’s why George Pavlov, the designer of the tVox and the late Fred Mundell always stated that the effective voltage at either antenna should be bigger than 40Vpp for noise immunity and smaller than 90Vpp to reduce interferences between instruments. Two tVox tour Theremins can operate side by side when set 2 to 3 meters apart.

Thank you for the recognition Vadim!

"First description of Eric's idea can be found in this post of "Teensy 4.0 600MHz ARM Cortex M-7 MCU - ideal for digital MCU based theremin?" thread."  - Buggins

I actually came up with that circuit in this post on 12/10/2013.

"That’s why George Pavlov, the designer of the tVox and the late Fred Mundell always stated that the effective voltage at either antenna should be bigger than 40Vpp for noise immunity and smaller than 90Vpp to reduce interferences between instruments. Two tVox tour Theremins can operate side by side when set 2 to 3 meters apart."  - Thierry

Interesting.  Of course, much depends on the relative frequency of the interference as well.  I've got D-Lev P2 and P3 guts within about a meter of each other (different coil values for each axis), and P2 is entirely playable.  Plate voltages are around 400V p-p.  On P2 I do see a bit of interference on the volume axis (operating at 455kHz) from the pitch side (operating at 900kHz).  This doesn't seem to affect playability, but it is visible on the scope as a "fuzz" and via single trigger as a "wobble" - the pitch plate waveform is obviously pushing/pulling the volume plate waveform around.  Some of this might be chalked up to the pitch plate having a somewhat higher Vp-p swing, some may be attributable to the volume axis operating below that of the pitch axis?

For all of the difficulty in implementation, I still think the LC DPLL approach is inherently superior to an analog LC oscillator.  Phase noise above the loop BW and below the operating frequency gets rolled off at a first order rate (6dB/octave, 20dB/decade), and there aren't start-up or harmonic locking issues if you design the AFE correctly and somewhat limit the low end of the NCO operating range.  This means, unlike an analog LC oscillator, the instantaneous phase has almost no influence over the current or next cycle (or the one after that, etc.).  The delayed antenna waveform is rock steady. 

With the P3 guts, I can set the pitch plate right on top of the volume plate face-to-face, do an ACAL, and almost play the thing - pretty remarkable.

At the time I am working on the measurement set for coils SRF investigations. The main requirement is connecting the coil just via "bottom end". Istead of differential input amp and current sensing resistor I use a current transformer (a ferrite ring coil). Works well, but the dewster's circuit is possible better. Needs try.
-- ILYA

I've tried to simulate current sensing oscillator with transformer and BJTs. Gives about 100Vpp on antenna.

Inductive decoupling of sensing removes big common mode changes of resistor current sensing, and therefore simplifies amplifier design.

LTspice model file link.

The LT1711 is too expensive.
-- ILYA

LT1711 is the only comparator I managed to get working in LTSpice oscillator simulation.
There are a lot of comparators with the same pinout. We can try to test them in hardware.
But fast devices with rail to rail signal and high common mode range are a bit expensive.

Oscillator is important part of theremin, why not spend more on it? Anyway, comparator price is low comparing to MCU/FPGA.
Similar to audio DAC/ADC or codec IC.

A too high voltage swing at the antenna might become a problem when several Thereminists want to play together because the higher the voltage the higher the interferences seen by neighbored instruments.

An Etherwave Pro drives the pitch antenna circuit already with 42Vpp which is then multiplied by ~10 in the series resonant circuit formed by the linearization coils and the pitch antenna. When an EPro works in a room, it’s almost impossible to play an Etherwave Standard/Plus in the same room, due to strong ghost tones and noise. A tVox tour theremin side by side with an EPro will simply be mute because the radiation of the EPro’s pitch antenna will overdrive the volume phase discriminator of the tVox.

That’s why George Pavlov, the designer of the tVox and the late Fred Mundell always stated that the effective voltage at either antenna should be bigger than 40Vpp for noise immunity and smaller than 90Vpp to reduce interferences between instruments. Two tVox tour Theremins can operate side by side when set 2 to 3 meters apart.
-- Thierry

I saw this effect with EW and Theremini. Theremini was even controlled by EW pitch from 5 meters

When voltage swing is too high, it's easy to reduce it - e.g. by increasing of drive resistor, or adding capacitance from antenna to ground.

I think, one oscillator affects other's oscillation if voltage swing is significantly different. Two identical theremins will not affect each other a lot.
Thereminist orchestra should use similar instruments

"I saw this effect with EW and Theremini. Theremini was even controlled by EW pitch from 5 meters"  - Buggins

This I totally believe.  The Theremini oscillators are total wusses.

"When voltage swing is too high, it's easy to reduce it - e.g. by increasing of drive resistor..."

Q killer, smears out resonance.

"...or adding capacitance from antenna to ground."

Sensitivity killer.  I think instead it would be best to somehow limit the drive voltage, or intentionally drive it slightly off phase.

I see you're using a current mirror for the collectors and constant current source for the emitters - I was actually thinking of trying that this morning.  Might be able to use a small phase lead RC to improve feedback timing.

Back when I was messing around with these things I noticed it was really easy to kill overall Q by loading the sense winding.

Buggins, I havent yet realized completely the new advantage of this circuit. Is it the low temperature drift? The frequency is determined by C1 - L1. If used an air coil the result should be excellent.

Buggins, I havent yet realized completely the new advantage of this circuit. Is it the low temperature drift? The frequency is determined by C1 - L1. If used an air coil the result should be excellent.
-- JPascal

The main reason why I started this topic that I believe this is new kind of LC oscillator, not listed on Wiki.

Except differential amplifier, the only possible sources of instability are inductor and antenna.
Antenna and inductor will present in theremin oscillator anyway.
So I hope, this approach leads to most stable theremin oscillator ever made.

Let's wait for comparision of its performance in real hardware.
I've ordered PCBs for LT1711 based oscillator, waiting for delivery.

Easy way to reduce antenna swing w/o reducing sensitivity to hand movements.

LTSpice model link.

Amplitude of signal driving R_sense is reduced to Vcc*R2/R1..Vcc*R1/R2
R_sense value can be used to change tank feed current and antenna voltage swing together with R1/R2 divider.

This change eliminates comparator requirement to support rail-to-rail inputs. LT1711 can be replaced with some cheaper device. Using of expensive 2ns comparator is an overkill.

Voltage on comparator output, R_sense input and output:

Now signal to compare is far from rails, and voltage difference is voltage difference on comparator inputs reaches 0.25V - easy to detect even with not best comparator.

Current through R_sense:

Tank feed current is now only 2.3mA vs 7mA in version w/o dividers and with small R_sense.

There are spikes when comparator changes its output. Spike height can be changed by changing of R_sense value. Smaller R_sense gives smaller spikes.

Antenna swing is now ~70Vpp instead of 400Vpp:

Even this version with reduced antenna swing continues oscillating with C_hand=200pF and R_loss=47K

Trying to find cheaper replacement for LT1711.
So far, only LT1713 is working in simulation - it's a bit slower and a bit cheaper version of LT1711.
LT1394 is working in simulation, although its voltage should be 5V, not 3.3V according to datasheet.

Improved BJT current sensing oscillator is posted by dewster in "analog theremin" thread.

Simplified version of it (LTSpice model):