In this article, we will consider the electronic filling of such devices, the principle of operation and the configuration method. Until now, I have seen descriptions of finished factory products, very beautiful, and not very cheap. In any case, with a cursory search, prices start at ten thousand rubles. I offer a description of a Chinese kit for self-assembly for 1.5 thousand.
First of all, it is necessary to clarify what exactly will be discussed. There is a great variety of magnetic levitators, and the variety of specific implementations is amazing. Such options, when permanent magnets, due to design features, are located with the same poles to each other, are now of no interest to anyone, but there are more tricky options. For example like this:
The principle of operation has been described repeatedly, to say briefly - there a permanent magnet hangs in the magnetic field of the solenoid, the strength of which depends on the signal of the hall sensor.
The opposite pole of the magnet does not turn over due to the fact that it is mounted in a model of a globe, which noticeably shifts the center of gravity down. The electronic circuit of the device is very simple, and almost does not need to be configured.
There are options for implementing such projects on arduino, but this is from the series βwhy is it simple, when it can be difficultβ.
This article is devoted to another option, where a stand is used instead of a suspension:
Instead of a globe, a flower is possible, or something else, as fantasy tells. Serial production of such toys has been established, but the prices do not please anyone. In the vastness of Ali Express, I came across such a set of parts,
which is the electronic filling of the stand. The issue price is 1,5 thousand rubles, if "Seller`s method" is selected.
As a result of communication with the seller, , and Chinese setup instructions. What especially touched me was that the seller provided a link to a video where the specialist also tells everything in detail in Chinese. Meanwhile, the assembled structure requires competent and painstaking adjustment, it is not realistic to launch it βon the moveβ. That's why I decided to enrich the RuNet with instructions in Russian.
So, in order. The printed circuit board was made in a very good place, as it turned out, it even has four layers, which is completely unnecessary. The workmanship is top notch and the silk-screen printing is well drawn and detailed. First of all, it is more convenient to solder the Hall sensors, and it is very important to position them correctly. A close-up photo is attached.
The sensitive surface of the sensors should be at half the height of the solenoids.
The third sensor, which is curved with the letter "G", can be raised a little higher. Its position, by the way, is not particularly critical - it serves to automatically turn on the power.
I would recommend mounting the solenoids so that the leads from the beginning of the winding are on top. So they stand up more evenly, and the risk of a short circuit is less. Four solenoids form a square, it is necessary to connect the diagonals in pairs. On my board, one diagonal was marked X1,Y1, and the other - X2,Y2.
Not the fact that you will come across the same. The principle is important: we take the diagonal, we connect the internal conclusions of the coils together, the external ones - into the circuit. The magnetic fields created by each of the pairs of coils must be opposite.
Four columns of permanent magnets must be fixed so that they all look in the same direction. It doesn't matter if it's the north or south pole, it's important not to be in discord.
After that, we calmly deal with the details and stick them according to the silkscreen. Tinning and plating are excellent, soldering such a board is a pleasure.
Now it's time to delve into the operation of the electronic circuit.
The node J3 - U5A - Q5 is located a little separately. Element J3 is the Hall sensor that is tallest and on bent legs. This is nothing more than an automatic power-on device. Sensor J3 determines the very fact of the presence of a float over the entire structure. We put the float - the power turned on. Removed - turned off. This is very logical, because without a float, the operation of the circuit loses its meaning.
If you do not apply power, the float tightly sticks to one of the magnetic columns. I draw your attention: this is correct, as it should be. The float must be turned to this side. It starts to repel only when it is strictly in the center of the structure. But while the electronics are not working, he inevitably falls on one of the tops of the square.
The regulator is designed as follows: two symmetrical halves, two differential amplifiers, each receives a signal from its own Hall sensor and controls the H - a bridge, the load of which is a pair of solenoids.
One of the LM324 amplifiers, for example, U1D, receives the signal from the J1 sensor, the other two, U1B and U1C, serve as drivers for the H-bridge formed by transistors Q1, Q2, Q3, Q4. As long as the float is in the center of the square, the U1D amplifier must be in balance and both arms of the H-bridge are closed. As soon as the float moves towards one of the solenoids, the signal from sensor J1 changes, some half of the H-bridge opens, and the solenoids induce opposite magnetic fields. The one closest to the float should repel it. and which is further - on the contrary, to attract. As a result, the float goes back to where it came from. If the float flies back too far, the other arm of the H-bridge will be opened, the polarity of the power supply of the pair of solenoids will change, and the float will again go to the center.
The second diagonal on transistors Q6, Q7, Q8, Q9 Works in the same way. Of course, if you mess up the phasing of the coils or the installation of the sensors, everything will be completely different, and the device will not work.
But who's stopping you from getting everything right?
Now that we have figured out the electronic circuit, the issue with the setting has cleared up.
It is necessary to fix the float in the center, and set the sliders of the potentiometers R10 and R22 so that both arms of both H-bridges are closed. Well, let's say, "fix" - I got excited, you can probably hold the float with your hands, more precisely, with one hand, and turn two multi-turn resistors in turn with the other hand. As it turned out, these resistors are multi-turn for a reason - literally half a turn on one of them, and the setting flies. Where my hands grow from is a secret, but by touch I could not catch the changes in the behavior of the float depending on the position of the potentiometer slider. I dare to suggest that the developer experienced the same difficulties, and therefore provided two such jumpers on the board.
See the two jumpers at the top left and right? They break the circuit between the pair of solenoids and the H-bridge. The benefits of them are twofold: by removing one of the jumpers, you can completely turn off one of the diagonals, and by turning on the ammeter instead of the other, you can see what state the H-bridge of the other diagonal is in.
As a lyrical digression, I note that if the H-bridges are completely open on both diagonals, the current consumed can reach three amperes. Under such conditions, it will be very difficult for the Q5 transistor to stay alive. Fortunately, it withstands such a load for a short time, but you need to twist two multi-turn resistors, and it is not known in advance where.
So for preliminary tuning, I strongly recommend fiddling with each diagonal separately: turn off the second one with a jumper so that Q5 does not smoke.
Since the current passing through the solenoids can change direction, the Chinese have such ammeters on the farm, in which the arrow is vertically in the middle of the scale. That is why they feel good and comfortable: they pull out the jumpers, stick the ammeters into the gaps, and calmly turn the resistors until the arrows go to zero.
I had to leave one jumper open, and include an old Soviet tester in ammeter mode with a measurement limit of 10 amperes into the other gap. If the current turned out to be reversed, the tester went off scale dully to the left, and I patiently turned the screw until the tester returned to zero. That's the only way to get it set up right. Then it was possible to turn on both diagonals, and adjust the adjustment, achieving maximum stability of the float. You can also control the total current consumed by the device: the less it is. the more precise the setting.
Out of habit, I printed the Levitron body on a 3D printer. It didn't turn out as beautifully as in the finished toy for ten thousand, but I was interested in the technical principle, not aesthetics.
Source: habr.com