Here is a creative way to generate a low-power, isolated auxiliary supply "for free": it uses the stray field from a conventional, 50/60Hz transformer.
If you can't beat it, use it !
The collecting device is simply a coil having an open magnetic circuit.
In this case, I have recycled a coil from a proximity sensor or a card-reader, I do not remember exactly.
When pushed against the laminations of a transformer, it has an inductance of ~1 Henry.
To improve the collection efficiency, it is made to resonate at the third harmonic with a parallel capacitor of 1µF, and the voltage is rectified through a voltage quadrupler:
The resulting voltage is sufficient to bias a zener, light a white LED and provide the supply of a LCD DPM (an ordinary one, requiring 1~2mA, not a micro-power):
The positioning of the coil on the transformer is quite critical:
Here, it is placed at the junction of the I's and the E's, where the magnetic leaks are maximal.
I made no optimization effort, and the transformer is of good quality: it is made by Siemens.
The scheme could be used to generate a bias voltage, or to feed a millivoltmeter, as in this example. The fact that the generated voltage is completely floating makes it particularly useful: it could spare a small auxiliary transformer
If you can't beat it, use it !
The collecting device is simply a coil having an open magnetic circuit.
In this case, I have recycled a coil from a proximity sensor or a card-reader, I do not remember exactly.
When pushed against the laminations of a transformer, it has an inductance of ~1 Henry.
To improve the collection efficiency, it is made to resonate at the third harmonic with a parallel capacitor of 1µF, and the voltage is rectified through a voltage quadrupler:
The resulting voltage is sufficient to bias a zener, light a white LED and provide the supply of a LCD DPM (an ordinary one, requiring 1~2mA, not a micro-power):
The positioning of the coil on the transformer is quite critical:
Here, it is placed at the junction of the I's and the E's, where the magnetic leaks are maximal.
I made no optimization effort, and the transformer is of good quality: it is made by Siemens.
The scheme could be used to generate a bias voltage, or to feed a millivoltmeter, as in this example. The fact that the generated voltage is completely floating makes it particularly useful: it could spare a small auxiliary transformer
The enamelled wires exiting directly from the winding are rather delicate, and I wanted to spare them the strain/bending of repeated manipulations which is why I glued the connecting wires directly to the top of the ferrite half-core with cyanoacrylate glue, thus no magics involved there, just plain practical considerations.
I have tested the no-load voltage: it establishes itself at 27V, thus the scheme could easily be adapted to the generation of a negative grid bias for example, and it is not the end of the story: the E-caps I used are 10V rated, but at 27V output they see ~13.5V, which means they probably start to leak.
Anyway, it would be easy to adapt the number of turns of the coil, or the number of multiplying stages to get other voltages.
Also, the geometry of the ferrite cup is probably not optimal: I think that the magnetic collection would be more effective without the outer cylinder part. Just the inner mandrel and the end plate. Some older types of potcores adopted this type of construction
I thought the power company could catch you for doing this up under the high tension lines, if enough of a load to be noticeable. Open circuit voltages probably could be enough to shock the bejeezus out of you….
Can you imagine the energy that setup could harvest if placed near an EV (Tesla) charging station?
Various people have made attempts to harvest the energy emanating from HV power lines:
https://makezine.com/article/craft/1000-fluorescent-lights-p/
https://www.epanorama.net/newepa/2020/11/24/under-the-110-kv-power-line/
This circuit is the electrostatic counterpart of the electric bed bug:
https://www.diyaudio.com/community/...-stray-currents-harvester.335610/post-5736940
I just realized that I shot myself in the foot: if you look at the picture in the first post, you will notice that the alligator clip feeding the DPM is connected to the wrong side of the final diode.
Consequences? The millivoltmeter sees a heavily undulated supply, and the local bypass of the module interferes with the voltage multiplier, a less than ideal situation.
Nonetheless, it did work in that condition: I didn't cheat when I took the picture. This means that the scheme isquite robust
