I've up-dated my website to add a circuit diagram, some comment on it, and the text from the corresponding LTSpice .asc file.
formatting link
The circuit produces a balanced pair of 1V rms 17kHz sine waves. The LTSpice simulation suggests the third harmonic content is about 60db below the fundamental, and the fifth harmonic is even smaller.
It's not a particularly simple circuit, and would take quite a bit of fiddling to get it to work at a significantly different frequency - all of it fairly obvious.
It exploits an asymmetrical current mirror to provide the gain that sustains the oscillation, and tweaks the - linear - gain in the mirror to stabilise the output swing.
It's not a Wien bridge, but a phase shift oscillator. There's no obvious amplitude control, so it presumably relies on clipping against the rails, which introduces is own distortion.
formatting link
Wien bridges all need some kind of gain control element, and if you want clean sine wave with well-controlled amplitude (which I did when I invented the circuit) the options are limited.
That's actually a phase-shift oscillator. They typically have a lot of distortion. They can be fiddled to have marginal loop gain, to minimize distortion at the cost of reliability. They are hard to tune!
I might Spice my weird resonant phase-shift oscillator idea this weekend.
--
John Larkin Highland Technology, Inc
Science teaches us to doubt.
Claude Bernard
What actual design problem is the circuit trying to solve that this topology is alleged to be superior to than the others?. Accuracy? Distortion? Supply current? Supply rejection? Start-up time? Phase noise?
At this frequency, they are way simpler and cheaper methods to get a low distortion sine wave.
My read on the inductors, is... they're windings on a single core. One inductor. Startup time is better than a (for instance) noise-dependent crystal oscillator, and phase noise will beat an RC (like Wein bridge) design.
Don't know why anyone sees a phase-shift oscillator here, it's just using some power-supply regulation of currents, to keep the transistors all out of saturation.
On Sunday, August 16, 2020 at 1:27:15 AM UTC+10, Kevin Aylward wrote:
and
low
of
Don't be silly. That's six windings in one transformer
You should know enough about Spice to be able to decode "K1 L1 L2 L3 L4 L5 L6 0.99 "
From my web-site
formatting link
r1.htm
"This circuit was originally developed as a retrofit to excite a linear va riable differential transformer used to measure the progressively increasin g mass of a single crystal of gallium arsenide (GaAs) being grown in the Me tals Research GaAs Liquid-Encapsulated Czochralski (LEC) crystal puller."
I needed a very stable amplitude for the sine wave excitation - and the LVD T happened to have a spare winding (originally intended to drive a demodula tor) which I demodulated (with a copy of the signal demodulator circuit) t o monitor the excitation voltage I was feeding into the LVDT. I wanted a fa irly clean sine wave - the switching demodulator was sensitive to odd harmo nic content - and I had to get the whole circuit onto a rather small printe d circuit board which sat in a rather confined space above the crystal pul ler that happened to rotate with the crystal. All the signals went in and o ut through silver brushes - not my choice. The whole thing was designed as retrofit. The operators liked it a whole lot better than the original circu it, but mainly because I'd taken out the original uA741 on the output - wh ich suffered from pop-corn noise - and replaced it with a next generation p art - possibly the OP07 - which actually had a low frequency input noise sp ecification. The effect was that the radio-frequency heating of the GaAs cr ucible stayed constant at about 40% of full scale, when before it had used to pulse-width modulate with a roughly 90 second period. I liked to claim that this improved the quality of the GaAs single crystals we pulled, but t here's absolutely no evidence that this was true. My next project used Giga bit Logic's GaAs parts, so I would claim - purely for comic effect - that I'd taken on the project to get myself better GaAs parts. The reality was t hat the guy who had done the electronic design for the original version of the crystal puller - some ten years earlier - was great at putting together stuff that worked - and kept on working - but rather slap-dash, and I spen t a year getting rid of his more desperate improvisations.
The 741 I got rid off had been driving a long cable - long enough to persua de it to oscillate - and my predecessor had "cured" the oscillation by load ing the output with 100nF to ground (making the amplitude too low to be vis ible). I put in the usual fix for a capacitative load when I reworked the c ircuit.
Name one. A Wien Bridge wouldn't have been any simpler, and would have use d more power to do the job.
The phase-shift oscillator was Anthony Stewart's proposed alternative - and he called it a Wien Bridge. Actually a Wein Bridge, for added extra comic effect.
The regulation does keep the transistors out of saturation, but in the original circuit the aim was to get a very well-defined - and stable output amplitude.
Monitoring the voltage at the centre-tap isn't a particularly good way of doing this, but it does keep the proof-of-principle circuit relatively simple.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.