2007-09-22
Super-Regenerative Tone Dipper
This dipper is the result of many experiments in improvement over the original tone dipper, super-regenerative receivers and LC oscillators in general. It is extremely simple and easy to build and offers fantastic sensitivity, easily observing resonance in a coil several inches away. Without adjustment it offers absorption wave meter service and the read-out is a frequency modulated audio tone for "eyes-up" use.
The heart of the circuit is your typical super-regenerative detector, although I've arranged for the quench frequency to be in the audible range. That's it really. That is all there is to the unit, except for a simple audio amplifier to drive a piezo to a reasonable volume (even without it the tone was audible just connecting the piezo element across the source circuit). In the diagram below pins 4 and 5 of the DIN plug are swapped around, the supply pin (2) should be shorted to the side that has the 1 nF decoupling capacitor.
As I've noted before, the quench frequency of a self-quenched super-regenerative detector is modulated by just about everything in the circuit, including the instantaneous tank parameters. Essentially the time it takes the circuit to ramp up its negative resistance and begin oscillating varies with the tank loss resistance. (And the "forcing" noise/signal it is driven with at the particular "sampling" moment. The mathematics of super-regeneration are discussed at some length in this fine article by Dr Eddie Insam.) Using this natural property by putting the quench in the audio range we can directly observe the losses/signals experienced by the tank as an audio tone. The tone will change when there is more signal/noise sampled over the ramp, and also when there is more loss coupled into the tank by mutual inductance. This is how the device can be both a wave meter and a dipper at the same time with no adjustment.
I got the general idea when playing with the micropower FM receiver circuit on the bench. I took a lot of time fiddling with the source circuit while observing the quench waveform. During this experimentation I noticed that the quench would change frequency drastically while tuning across a strong FM station, or while tuning my signal generator across the detector frequency. At the time I put this in the notebook as a possible way to implement a sensitive wave meter, or perhaps a squelch for a super-regenerative detector in a walkie-talkie like the Fredbox (although a "quieting" squelch is likely easier to build). Further experiments with the completed receiver noted an interesting "suck-out" effect when I tuned an LC circuit loosely coupled to the detector across a station while listening to it (one of the issues I experienced building the Fredbox too). At times it would pull the quench right down in the audible range, or collapse the oscillation all together. The novelty of dipping a few turns of wire plugged into my "C" jig using an FM radio station as a frequency standard soon wore off and I decided to build a proper instrument which used the same principle.
I rebuilt the Tone Dipper, replacing its circuit board and upgrading its power supply with an additional AAA battery. Because of the slightly different topology (drain resistor to give better tone shift) I had to add a power switch and could no longer rely on the coil plug as a power switch. A calibrated dial was constructed, using the dipper in wave meter mode and the bench signal generator and counter to provide data points. The RF radiation by the dipper itself is vanishingly small, not only because it is operating at such low power levels, but also because it is actually oscillating only extremely briefly. During its regeneration ramp it is chirped somewhat, making the selectivity fairly poor, and the large parametric variations in device capacitance (especially associated with the oscillator topology) make this worse than it would be with a more conventional common-gate/base oscillator. However, the device is a dipper, not a precision signal generator, and a little bit of line broadening is actually helpful in this service.
The detector is sensitive enough to hear off-air signals. As you tune across the FM broadcast band you can hear little squawks and general "roughness" in the quench note, this is the off-air signals pulling around the quench frequency. Similarly it hears TV signals, and even 49 metre broadcasts when dipping an antenna. This may frustrate you, but the tone shift is different for "added" signals vrs "resonances"... It can be a touch confusing actually, you may experience odd sounds when there are interfering signals (or resonances inside the dipper itself - watch out for these during construction; big ones can stop the unit oscillating), but your brain will easily work out what is going on. One enhancement might be to make the audio amplifier more linear and offer a switched source capacitance (or vary the source resistance with a pot) so the quench can be elevated to supersonic frequencies, allowing the user to listen to any modulation at frequency in question. I experimented with this idea a bit but decided not to add it to mine as I was running out of space and voice or music through a "peaky" piezo sounds horrible.
Operation
Here are some videos displaying the operation of the Dipper:
- Dipping an RFC's Self-Resonant Frequency (990 KiB)
- Dipping an Unknown Coil with a known Capacitance (1.9 MiB)
- Dipping by Tuning Dipper and Tuning Unknown Tank (4.9 MiB)
Notes
It would likely be possible to make an even more sensitive version with a supersonic quench and a heterodyne oscillator. This would not be very useful as a dipper itself, I suspect nulling it would drive you mad, however it might make a great resonance detector at a fixed frequency for detecting resonant "tags" coupled into the circuit. I do not believe this is how current shop security systems work, but something similar might be made this way. Thermal drift would be a problem if low frequency response would be needed. As a pickup system for a Theremin (or similar gestural interface system) it perhaps has more promise, but requires the performer to ware a resonator on the appendage to be tracked. Multiple resonators at different frequencies would allow tracking in multiple dimensions, but coupling varies with coil alignment, which is both a feature or a liability depending on the application.
The circuit could be used as a proximity sensor for tagged objects. I had some fun swinging and throwing an LC resonator near the unit sitting upright on the bench. I'm not sure how it would go if the search oscillator coil was the size of a room and the tag was tiny. Not well I'd imagine, too much absorption from other objects and little energy density to illuminate the tag. (Is it actually illuminating the tag? I'd say yes. Electrically it is coupling to it via the coil pair's mutual inductance, but the tag's losses and thermal noise are only sensed by the complete oscillator system when energy is applied to generate the "negative resistance" to cancel the combined losses. This process must obviously be causal, otherwise one could use the system to send information faster than the speed of light by modulating the loss in the tag coil... One wonders how far apart the coils can be with respect to the quench frequency and the exact mechanisms involved. Are virtual photons from each coil's charges the information carriers?)
If it could be made "room sized" while still retaining its advantages of low power, you could modulate the tag (say with a button to close the LC circuit) to allow remote control applications with a completely passive control unit. (I strongly suspect Wacom tablet stylus work on this principle for sensing the nib and button states.) A homodyne proximity circuit with a digitally modulated tag might be better in this application, like a smart-card reader or a proximity fuze. Finally a use for the "tunable hum" problem so common to direct-conversion receivers!
I don't know how novel this idea is. It can't be new, I am sure many radio experimenters have observed the effect and even built dipper-like devices to make use of it. However, a Google search reveals no other similar circuits. The only reference I found mentioning dipping and super-regeneration in the same device is this: Super-regenerative UHF-Dip-meter (DL7HG). It appears to be a separately quenched receiver, not a dip meter (despite the title). I would be very interested in finding out more about just what was the intention of this circuit. The coupling loop does make it appear to be a piece of test equipment, but perhaps it was designed more as a wavemeter for listening to AM modulated radiation? Why one would need the extreme sensitivity of a super-regenerative receiver in this application isn't immediately obvious.
4 comments.