Ultrasonic bath. Part 1

Hamsters welcome you friends.

Today's post will be devoted to the creation of an ultrasonic cleaning bath, which is based on a 60 W Langevin piezoceramic emitter. In the process, we will look at what the device consists of, how to set it up so that nothing burns out and at the end we see purification abilities that surpass Mr. Proper and all his friends in their action. An ultrasonic bath has many fields of application and it is almost impossible to list everything, since most of them will depend only on your imagination.



Before starting to dissolve your fingers in an ultrasonic bath, let's look at how mechanical vibrations arise on simpler systems. One example of such oscillatory mechanisms are magnetostrictors, which can be compressed or stretched under the influence of a magnetic field. These parameters are possessed by ordinary ferrite from the old grandfather's receiver, which for sure everyone is lying somewhere in the garage.



To start the experiment, we need: a signal generator, a pulse density modulator for adjusting power, a half-bridge, an adjustable power supply and an oscilloscope for visual evaluation of the signal. Further on a small mandrel we wind a coil of thick copper, in my case about 50 turns of 2 mm wire came out. Ferrite will be inserted right in the middle of this gun of the house. We set the power at 100 percent on a pulse modulator. Turning the knob on the generator, we find the resonance of the system, which in a particular case will look like two mountains whose tops need to be aligned.



The frequency of a particular rod is 8.5 kHz.Approaching mechanical resonance, it is seen how a drop at the top of a ferrite rod begins to vibrate, while changing its original shape. At some point, the amplitude of the vibration reaches such a value that it breaks the water into thousands of small particles and visually it seems that the liquid turns into fog in a split second. The size of each such drop depends on the mechanical system; the higher the frequency, the smaller the drop.

Such a magnetostrictive system is bad in that, at a certain power threshold, brittle ferrite is torn to pieces, as has happened now.15 watts were not allowed. In the middle of the rod, maximum mechanical stress occurs, and it breaks it. If after that you try to glue two halves of the rod, then there will not be such active work as was originally, since each individual piece will have its own mechanical resonance. During the shooting, I had three such rods torn apart.



As an experiment, we connect the most ordinary piezoceramic emitter to the generator. Rotating the handle of the generator we find the moment when the water begins to actively perturb. As can be seen, the droplets that formed have a slightly larger size than in the previously presented version, since the resonant frequency here is 2 times lower and corresponds to 3.6 kHz.

For reference.In ultrasonic evaporators and humidifiers, the same principle is used, only the frequency here lies already in the megahertz range. The size of a drop of water can reach several tens of microns.



Now we turn exclusively to the Langevin radiator, named after the French physicist who was engaged in magnetism. The electromechanical frequency of this piece of iron is 40 kHz, and the evaporation of water on it is more like the eruption of a volcano. At such idle speed, the emitter is very hot, so I do not recommend doing this.



In the next experiment, we will try to obtain ultrasonic levitation.At a resonance in Langevin, a standing ultrasonic wave with an antinode at the end of the radiating lining is formed. This is the main longitudinal mode. In this case, the particles of matter at the end of the patch oscillate in the vertical direction with an amplitude of tens of microns. These vibrations are easily transmitted to the air.

If you install a reflecting surface at a certain distance from the emitter, then the radiated and reflected waves will add up, forming standing sound waves in the air that have nodes - areas of minimum pressure, and antinodes - areas of maximum pressure. In order for the ball with polystyrene foam to levitate, it must be placed exactly in the sound pressure unit. If you turn off the system, the entire house of cards will immediately collapse.



With the principle of work, Langevin figured out.Now you can take a closer look at the emitter. On the front side you can see the sandblasted matte surface, which provides better adhesion to the glue, which will fasten the emitter to the gastronomic capacity.



The volume of such a trough is one and a half liters. Vessel size 1/6, depth 100 mm, stainless steel material. We center the emitter at the bottom of the vessel and mark the place where it will be located. In fact, this is necessary so that traces of sandpaper do not get out of the border and do not spoil the appearance. Ideally, this place is best treated with sandblasting, but I don’t have one on the farm. When the surfaces are prepared, degrease them with acetone and dilute the epoxy glue.



We apply it with a thin layer on the trough itself and carry out the same procedure with the emitter. There should not be any gaps, since we need to ensure good acoustic contact of the entire radiating surface. When docking the shuttle, Langevin tries to crawl somewhere. So that he does not run far, he needs to be rubbed a little, and then pressed down so that all the excess glue crawls out.



After polymerization, the epoxide will acquire the so-called metallic hardness. For fans, this option to start working with powerful ultrasound can be quite uplifting.



Now is the time to make the case. We mark pre-measured dimensions on 10 mm chipboard and begin work with a jigsaw. It is advisable to do such an operation at night, when all the neighbors are sleeping)

In the end result, 5 even pieces will come out, all that is needed is to secure the plywood walls more reliably so that nothing falls apart. We try on the bath by inserting one into the other. Ideally, the box should come out slightly smaller than the dimensions of the gastronome.



We pass to the electronic part. To control the operating time of the bath you need a timer. A suitable circuit was found on the Internet, but the circuit board itself had to be bred as it simply was not in the description. The result is a small scarf with a fairly modest size. Exactly what is needed.

We serve food and see how something lit up.A short press on the encoder button turns the timer on and off. Turning the knob allows you to select the time in minutes from 1 to 99. After the specified interval has elapsed, music plays, and then a siren sounds, which can be turned off once by pressing the encoder. Work is nowhere easier. If someone is annoyed by sound signals, a jumper is provided on the board to disconnect the speaker.



Now it's up to the generator, which will pump the speaker system. I bred the board exclusively for the dimensions of the parts that I dig in the pantry. I tried to place the elements as tightly as possible so that there were no high-frequency pickups. Although the option assembled from shit and sticks on the knee also did not work badly, it is not worth doing this.



The generator is called a push pool.In the beginning, it had transistors IRFZ46, then 2SK1276, then IRFP460 they all seemed to work in a kind of dull. The IRFZ44 transistors worked best of all, and stopped on them. Management comes from the driver chip IR2153.

Since the frequency control will be manual in some modes, the transistors will be very hot. Therefore, it is necessary to provide for a good heat dissipation. It is advisable to use a radiator with a thick base, since its heat dissipation will be much more efficient than that of a piece of aluminum located on the left, which overheats as a first grader on a first date. In any situation, it is necessary to provide good heat dissipation and air cooling. The temperature will be displayed on a Chinese thermometer with an LCD screen. It costs about 2 bucks.

All the energy in the bath will be pumped by a pulse transformer from a computer power supply. From practice, the size of the transformer does not matter, everything worked equally on both small and large such fucking. 60 watts for them like two fingers. We will evaluate the consumption of the entire circuit according to the readings of the ammeter of a powerful shunt connected in parallel. The power supply for our task needs not weak. This board is picked up from charging from some laptop. If you believe the characteristics, then it produces 65 watts at a voltage of 20 volts. Dividing the first by the second we get a current of three and a quarter amperes, which is very pleasing.



Now this bunch of parts needs to be staggered.To do this, on wooden boards we include all of our artist skills and mark the previously planned places where the controls will be inserted. The clean work has been completed, it is time to turn the carpet over with sawdust from chipboard, which pours like snow during drilling holes. The rough marks from the drill are removed with a drill. Since the nozzle is round, it remains to align the corners and here the file comes into play. But you need to work with it carefully, since on the decorative coating chips are obtained. After the dust has settled throughout the hut, decorative woodworking can be considered complete.



We place all the electronics.Good tone when all the details come in tightly. We place the timer board on the reverse side, and on the front, a Chinese thermometer that shows the temperature in tenths of a degree, we also set the remaining switches and switches. The result will be something like this.



Inside we place the power supply, as it can be seen near the blow hole for better cooling. We place the generator board opposite the fan and place the last element - the inductor.



How does this whole pile of iron work ?!Now let's figure it out. To start tuning, we set a voltage of about 14 volts on an adjustable power supply. We check the stabilized voltage to power the driver chip, it should be 12 volts. Using the oscilloscope probe, we cling to the transistor's gate and check if a signal is present in the form of a meander. If everything is in place, we change the frequency with a variable resistor and make sure that the signal does not twitch and is even in the entire adjustment limit. In this case, the upper limit is about 80 kHz, and the lower one is in the region of 34 kHz. The stock is large enough and the pocket as they say does not press.



We turn on the divider by 10 and connect to the middle leg of the field pole - this is the drain.At idle, you can see how at the moment the transistor is turned on, a high-voltage surge occurs, followed by a free damped oscillation compared to a shock on the water. At the moment the key is turned off, we see another peak. Ideally, this place should have a clean meander. But it seems he was swelling. Let's try to connect the load in the form of a Illich lamp. We see how the attenuation disappeared, the front edge of the meander is in the blockage, and inductive emissions reach about 700 volts. Such a picture is no good.

Part of this horror occurs even in the board, even a finger affects it. The same signal will be repeated at the transformer output. One can see how between the inclusions of each shoulder a deadweight of 1.2 milliseconds is formed. Exactly, in addition to the waveform, the work goes in the right direction.

High-frequency ringing can be crushed by the snubber. This is the name of the chain of resistor and capacitor. In this case, the resistor should be powerful, about 5 watts, since it is very hot. We place them in the zone of cooling the radiator. By connecting the RC chain to one of the shoulders of the push pool, one can see how the waves are quenched, though with a slight disturbance at the moment of switching on. This is the best thing that could be achieved experimentally by selecting the capacity and resistance of the snubber for this circuit. In any case, even under load, the signal at the output of the high-voltage part of the transformer tends to be similar to a square wave. With this sorted out, we move on.



Since the emitter is a capacitive load, it is necessary to calculate a resonant inductor, which will increase the efficiency of work.We measure the capacitance and get about 5 nF. The frequency of this Langevin is 40 kHz. We go into the "Electrodroid" program and enter these parameters there. An ingenious program for doubles, you do not need to count anything, just enter numbers, the program will do everything for you. According to the calculation results, the inductance reached 3.2 mH. We will wind the transformer with a double wire to reduce the total resistance. Less resistance, less losses that will be dissipated as heat.

The first version of the inductor was wound on the core of an unassembled transformer. It took about 4 hours, since it was difficult to lay copper coil to coil. The final inductance with all efforts exceeded 0.6 mH. I was upset. You can wrap a sample in one wire on a regular piece of ferrite, there will be many losses, but this option will work for tuning.



And so what are we seeing here ?! At one end of the emitter sits a current transformer, in the future it will be of little use. On the hot end of the inductor we hook up a neon bulb for a visual assessment of the voltage. Pour some water into the gastronome containers, about 1/3. Connect the oscilloscope probe to the high voltage output of the transformer.

We raise the voltage and see ...Yes, hell, understand that! At resonance at maximum consumption, the meander sits down at the very most I want, forming two peaks as in the movie The Lord of the Rings. I suspect this is how the inductor affects the low-voltage part. The magnitude of the voltage seems to be considerable, so I do not recommend doing it any further. We connect the probe with the divider to the hot end, adjust the frequency and see how the voltage amplitude waves out of the scope of the oscilloscope. A span of about 1000 volts. The second end of the neon lamp is plucked when touched.

Let's see what's on the current transformer. The picture is jumping due to poor synchronization of the oscilloscope. Anu synchronized old junk. Do not take me out! The current at resonance is growing, which is as it should be. If the water in the bath hangs, then the system becomes unstable.

An interesting effect discovered during the experiments. If one end of Langevin is not connected to a common circuit wire, then the entire voltage potential in kilovolts appears on the bath body, this is clearly visible on the neon bulb. Even small sparks slip when touching a piece of iron. On the board, a grounding jumper is provided in advance.



The circuit of the electronic part. I tried to indicate everything in it, even the transistor pinout. On the throttle of the resonant part is a contactor. I noticed that sometimes a bath works better without it than with it, and sometimes vice versa.



For clarity, below are shown two pictures with signals. The first work with capacitive load, and the second with resonance. Archive with all the necessary materials for assembling the bath.



We figured out this part, like nothing burned out, we move on. We connect all the connectors with power, control, variable resistors, Keller, etc. Since the temperature sensor of the thermometer has a very convenient form for mounting, I did not come up with anything else to attach it to a piece of foil adhesive tape, although it would be more correct to drill a hole in the radiator and put it there together with thermal grease for better thermal contact.

The body of the bath is made of chipboard, and as you know, he is afraid of water, or rather, its unprotected sidewalls. Waterproof silicone copes with such tasks. We separate a piece of this muck and rub it into the ends of the trees. Here it is important not to rush anywhere for ourselves. Also, a damping tape will be better supported on silicone, which will isolate the body of the gastronomic container from the device body so that useful vibrations are not damped.

For fastening Langevin to a stainless trough instead of epoxy, you can use Poxipol type cold welding. It seems like manufacturers of bathtubs use it. Let them use it, an ordinary epoxy is several times cheaper.

For reference.Do not leave things unattended, otherwise hamsters will run in and gnaw all the wires. But do not be afraid if you can always fight back with a soldering iron) To say that the bath turned out to be compact is to say nothing compared to the Chinese, but how much power is there ...



Part two

---

Utility archive
Full project video on YouTube
Our Instagram