EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: vranghel on January 13, 2023, 10:24:17 am
-
Hello everyone!!
I've been playing around with this circuit
(https://www.basic4mcu.com/data/file/k5/3660040649_CGuT5Xz6_MagBot2520Pendulum.png)
A more similar version, which replaces the solar cell with a battery, see attachment
I built a very 'successful' pendulum using a 3.5 ohm, 0.8 mH coil and adjusting the resistors until i've got the best result:
https://gfycat.com/idealagedbunting (https://gfycat.com/idealagedbunting)
It uses a small solar panel to charge a small Li-Po battery that runs the circuit for weeks at end even any light.
I'm trying to a build another version with a more powerful 'kick'. I got some small coils (designed for RFID) from aliexpress https://ae01.alicdn.com/kf/S7d2d357031bf45c99b54f7e3b88e7f702/125KHz-ID-T5577-EM4305-134-2KHz-COB-coil-tags-10pcs-Lot-RFID-passive-ID-chip-and.png_.webp (https://ae01.alicdn.com/kf/S7d2d357031bf45c99b54f7e3b88e7f702/125KHz-ID-T5577-EM4305-134-2KHz-COB-coil-tags-10pcs-Lot-RFID-passive-ID-chip-and.png_.webp)
These are 122 ohm and 4.6 mH.
The problem is I can't get the circuit to work, no matter how I adjust the two resistors. The magnet is always attracted to the coil. I believe the resistance of the coil is too large and the circuit fails to function. However, I do not know how to modify it to work with the new coil.
Any help would be highly appreciated!
-
If the idea is that the magnet is pushed away by the coil, and it's currently attracted by it, try either flipping the coil over or swapping the connections (not both). You also need to avoid ferromagnetic materials in the coil (looks OK in your photo) or the magnetic attraction to those will dominate.
-
I've read that in order for the neodymium magnet not to be degaussed (demagnetized) over time, the coil should PULL on the magnet.
Both circuits are identical, and the only variable is the coil. I've tried the new circuit with a spare coil (i have 2, one used in the working circuit and one spare) and works well.
Just with the new coil (higher resistance and inductance) the circuit fails to work no matter what the two resistors values.
-
The polarity of the coil and magnet relative to each others matter. In one direction the magnet would give the pendulum a push, in the other it will tend to stop it.
In the circuit with the polarized capacitor the polarity looks the wrong way around.
The neodymium magnets are not that sensitive to demanitization - the coil is way too weak for this. If worried, add a millimeter of steel behind the magnet (still part of the pendulum) and this would already reduce the demagnetizing field more than what the tiny coil could produce.
P.s. : Stacking 2 or 3 magnets like in the video already reduces the demagnetizing field a lot and makes the magnets much more stable. Demanetization was an issue with the old Alnico magenets and for this reason they only came in a rather elongated form or horse shoe shape. Sticking two together makes them like a much more stable, longer magent.
P.S. 2 : with a battery supply it may be a good idea to have a series resistor on the input side / emitter of the PNP to reduce the overall gain. The solar cell is more like a current surce and thus high impedance. Without it the circuit can just latch to the current on case and stay there.
-
This design works reasonably well provided the battery voltage remains constant.
http://www.4tubes.com/3-BOOKS/BOOKS-LITERATURE/ENGLISH/Magazines/Wireless-World/Cardboard%20Clock%20-%20Michael%20McLoughlin.pdf (http://www.4tubes.com/3-BOOKS/BOOKS-LITERATURE/ENGLISH/Magazines/Wireless-World/Cardboard%20Clock%20-%20Michael%20McLoughlin.pdf)
If you want the pendulum to keep time you need to reduce all losses to a minimum.
-
The polarity of the coil and magnet relative to each others matter. In one direction the magnet would give the pendulum a push, in the other it will tend to stop it.
In the circuit with the polarized capacitor the polarity looks the wrong way around.
According to the circuit description, when the magnet approaches the coil it's being attracted to the coil, and once it reaches the top of the coil, the coil turns off and the magent coasts away.
The same principle applies when the coil PUSHES AWAY the magnet.
Again, my problem is very specific: the circuit works with a 3.5 ohm, 0.8 mH coil (see gif), but it does NOT work when I replace the coil with one that is 122 ohm and 4.6 mH. I assume it's due to the two resistors which control the voltage and the current that goes through the coil.
With the larger coil, it is only attracted to the magnet continuously, and the coil is not turned off so the magnet cannot coast away.
This is shown by the LED, which should blink when the coil swings above it (due to back EMF). With the larger coil, the LED is either always on, or with adjusted resistors always off, and it never turns from on to off and therefore blink the LED.
EDIT:
In the circuit with the polarized capacitor the polarity looks the wrong way around.
Correct, in this schematic the electrolytic cap is indeed wrong. I followed an alternative circuit diagram where it was wired correctly.
-
In the circuit as shown in the inital post the LED would only light up if the circuit oscillates and there is a good chance for the ciruit to become an oscillator.
The resistor to set the operation point needs to adjusted pretty well. Chances are a good point would be where the LED is just at the edge of lighting up.
It can also help to have the capacitors to avoid the circuit to get stuck in a DC way and get the most gain just at the right frequency range.
The larger resistance / inductance of the coil changes to gain operating point also quite a bit. The tendency would be to also need larger resistors, e.g. in the 2 Mohms range. Chances are the 2nd coil would be lower power or need a higher supply which comes with other problems on its own. The larger coil may just be a bad choice and fewer turns with a thicker wire could be needed.
-
In the circuit as shown in the inital post the LED would only light up if the circuit oscillates and there is a good chance for the ciruit to become an oscillator.
The resistor to set the operation point needs to adjusted pretty well. Chances are a good point would be where the LED is just at the edge of lighting up.
It can also help to have the capacitors to avoid the circuit to get stuck in a DC way and get the most gain just at the right frequency range.
The larger resistance / inductance of the coil changes to gain operating point also quite a bit. The tendency would be to also need larger resistors, e.g. in the 2 Mohms range. Chances are the 2nd coil would be lower power or need a higher supply which comes with other problems on its own. The larger coil may just be a bad choice and fewer turns with a thicker wire could be needed.
I played with resistor values up to 9 Mohms, and could not get it to work. The power supply was set to 5v, and current limited to 40mA.
Are there no other adjustments that I can make to the circuit to make it work with the larger coil? It's very difficult to find appropriate coils for this project. The 2 I already have have been scavenged from an old 3.5 in floppy. Not many of those still around.
-
For the coil winding one yourself would be a real option.
For the higher resistance coil a higher voltage makes absolute sense. This would also requite R2 to be much larder than R1.
For the circuit it may make sense to have a large capacitor (e.g. 1-100 µF range) in series to R1 (not sure, but could be wirth a try). This would break the positve feedback.
The led constantly lid indicates that the circuit is oscillating - the supply could be part ot this.
-
For the coil winding one yourself would be a real option.
For the higher resistance coil a higher voltage makes absolute sense. This would also requite R2 to be much larder than R1.
For the circuit it may make sense to have a large capacitor (e.g. 1-100 µF range) in series to R1 (not sure, but could be wirth a try). This would break the positve feedback.
The led constantly lid indicates that the circuit is oscillating - the supply could be part ot this.
Thanks Kleinstein, I will give it a try and report back
-
I built a very successful solar pendulum using this coil:
[attach=1]
What's important in the operation of this device is not the size of the kick from the coil but the timing. Small kicks at the right time will accumulate and increase the amplitude of the swing. I think a big part of obtaining high amplitudes is making sure that the magnet travels in a consistent path over the coil -- e.g. without any rotation or twisting.
I have a bunch of these coils and I tried stacking two in series to get a larger kick but I didn't see any appreciable difference.
That said, I've seen videos of solar pendulum designs that use larger coils. For instance, this one:
24/7 Solar Pendulum -- sMartcreations2010
https://youtu.be/vIJV7nWO0Y4 (https://youtu.be/vIJV7nWO0Y4)
According to the video it uses a bobin coil of about 2000 turns of 0.1mm wire. The video shows a schematic at the beginning. Another video (https://youtu.be/4YXWwKeo1k0) shows the dimensions of the bobbin are 15mm with a 6mm center hole and the coil has a resistance of 190 ohms.
Update: The coil I used is about 6 layers deep and around 9 turns on each layer. Another coil option is to scavenge one from a solar flip-flap flower.
-
Does this kind of circuit keep the pendulum swinging in the same plane? If not, it seems you would experience the Foucault pendulum effect, with the plane of oscillation rotating at a rate based on your latitude as the Earth turns. But maybe this is how they keep Foucault pendulums going without affecting their plane rotation.
-
If nicely build and the coil in the dead center the circuit should also work with a Foucault type pendulum that rotates the plane of scilation relative to the setup and less relative to the stars.
-
In the Youtube video linked by ledtester, I don't see anything at the top which limits the plane. And the bobbin coil is round. So it seems the pendulum would soon crash into the posts. What am I missing?
-
In the Youtube video linked by ledtester, I don't see anything at the top which limits the plane. And the bobbin coil is round. So it seems the pendulum would soon crash into the posts. What am I missing?
The pendulum in the video has 2 support wires, so one fixed plane. It is visible relatively early in the video, though not very well.
-
This is the typical way the solar pendulum is built:
Solar Pendulum -- accomplishedorg
https://youtu.be/yZb2-OSmxik
so it is restricted to swinging in a plane.
-
I built a very successful solar pendulum using this coil:
(Attachment Link)
What's important in the operation of this device is not the size of the kick from the coil but the timing. Small kicks at the right time will accumulate and increase the amplitude of the swing. I think a big part of obtaining high amplitudes is making sure that the magnet travels in a consistent path over the coil -- e.g. without any rotation or twisting.
I have a bunch of these coils and I tried stacking two in series to get a larger kick but I didn't see any appreciable difference.
That said, I've seen videos of solar pendulum designs that use larger coils. For instance, this one:
24/7 Solar Pendulum -- sMartcreations2010
https://youtu.be/vIJV7nWO0Y4 (https://youtu.be/vIJV7nWO0Y4)
According to the video it uses a bobin coil of about 2000 turns of 0.1mm wire. The video shows a schematic at the beginning. Another video (https://youtu.be/4YXWwKeo1k0) shows the dimensions of the bobbin are 15mm with a 6mm center hole and the coil has a resistance of 190 ohms.
Update: The coil I used is about 6 layers deep and around 9 turns on each layer. Another coil option is to scavenge one from a solar flip-flap flower.
The coil used in the pendulum that works flawlessly looks similarly to the one you showed.
Does this kind of circuit keep the pendulum swinging in the same plane? If not, it seems you would experience the Foucault pendulum effect, with the plane of oscillation rotating at a rate based on your latitude as the Earth turns. But maybe this is how they keep Foucault pendulums going without affecting their plane rotation.
See the GIF in my initial post to see how it swings.