Driving ultrasonic transducers

[Los Frijoles]

I understand this sort of question is asked often, but most of them have at least started something and are mainly looking for ways to fix it. I am on a limited budget and want to make sure I get this right before I go blowing things up.

I have recently come into a couple 40Khz unlabled ultrasonic transducers. The place I got them gives capacitance and such along with an optimal voltage of 20V. They are similar to the ones seen on the back of cars. So far, I understand ultrasonic transducers to behave like a series LRC circuit with a resonant frequency of whatever they are rated for (40Khz in my case). These seem to be low-power transducers.

Here are my questions:

* I have a step up thing I found on dealextreme to get me the appropriate driving voltage. Some places I have looked say that it needs to be driven by a sine wave. Others seem to show that inputting a square wave is just fine. If I use a square wave (lets say off of a microcontroller pin) as the frequency source (40Khz of course), will that work fine for driving this sort of thing after being amplified to the proper voltage? Or do I need to make a sine wave generator at that frequency?
* I understand that these things apparently create quite a bit of back-emf spike. Is there a more effective way than a 100 ohm resistor and a zener to take care of this? Its been a while since I have considered this kind of thing. Most of what I have designed for myself have been digital circuits that didn't need to take care of this a lot and I don't have a lot of money to be burning out parts because of back-emf spikes.
* The pins seem to imply that this thing is polarized in some way (one is longer than the other). Is this thing polarized? I was under the impression that piezo-based things weren't exactly polarized (I probably am quite incorrect).

There may be more as things come up, but a lot of my questions can be answered by hooking this thing up to an oscilloscope and seeing what happens.

 

[pakitos]

a)I have worked with several ultrasonic circuits. All work fine with square wave inputs.
b)When supplying from a microcontroller a proper npn pnp amplifier stage is sufficient to drive the ux tx.
c) check on the oscilloscope the level of back emp to solve that problem.
d)yes they are polarized usually the lower side is connected to the body of the sensor.

If you need more info do not hesitate and ask
regards

 

[jiripolivka]

I understand this sort of question is asked often, but most of them have at least started something and are mainly looking for ways to fix it. I am on a limited budget and want to make sure I get this right before I go blowing things up.

I have recently come into a couple 40Khz unlabled ultrasonic transducers. The place I got them gives capacitance and such along with an optimal voltage of 20V. They are similar to the ones seen on the back of cars. So far, I understand ultrasonic transducers to behave like a series LRC circuit with a resonant frequency of whatever they are rated for (40Khz in my case). These seem to be low-power transducers.

Here are my questions:

* I have a step up thing I found on dealextreme to get me the appropriate driving voltage. Some places I have looked say that it needs to be driven by a sine wave. Others seem to show that inputting a square wave is just fine. If I use a square wave (lets say off of a microcontroller pin) as the frequency source (40Khz of course), will that work fine for driving this sort of thing after being amplified to the proper voltage? Or do I need to make a sine wave generator at that frequency?
* I understand that these things apparently create quite a bit of back-emf spike. Is there a more effective way than a 100 ohm resistor and a zener to take care of this? Its been a while since I have considered this kind of thing. Most of what I have designed for myself have been digital circuits that didn't need to take care of this a lot and I don't have a lot of money to be burning out parts because of back-emf spikes.
* The pins seem to imply that this thing is polarized in some way (one is longer than the other). Is this thing polarized? I was under the impression that piezo-based things weren't exactly polarized (I probably am quite incorrect).

There may be more as things come up, but a lot of my questions can be answered by hooking this thing up to an oscilloscope and seeing what happens.

Driving piezo ultrasonic transducers can be easy or difficult, depending upon what acoustic power you need to generate.
The simplest way to drive a 40 kHz transducer is to connect it to the output (pin3) of a 555 connected as multivibrator at 40 kHz. To see the best function,use another 40 kHz transducer connected to oscilloscope input (1 Ohm, 50 pf). You will see that the resonance at 40 kHz is quite narrow, so a fine generator tuning is needed.
To increase the output power ~4 times, connect a 7404 or a CMOS multiple gate after your 555, and connect the transducer between the input and output of the last gate.

For a higher output power, check transducer rating for a maximum voltage. Piezo transducers are capacitive, so you can build a transistor power amplifier with an output transformer like 500 to 5000 Ohms, and with added capacitors tune the output for resonance at 40 kHz, with transducer connected.

High-power transducers allow the peak-to-peak voltage over 100 V, but they can be damaged by overvoltage as well as by mechanical overload.

 

[honeybadger]

I understand this sort of question is asked often, but most of them have at least started something and are mainly looking for ways to fix it. I am on a limited budget and want to make sure I get this right before I go blowing things up.

I have recently come into a couple 40Khz unlabled ultrasonic transducers. The place I got them gives capacitance and such along with an optimal voltage of 20V. They are similar to the ones seen on the back of cars. So far, I understand ultrasonic transducers to behave like a series LRC circuit with a resonant frequency of whatever they are rated for (40Khz in my case). These seem to be low-power transducers.

Here are my questions:

* I have a step up thing I found on dealextreme to get me the appropriate driving voltage. Some places I have looked say that it needs to be driven by a sine wave. Others seem to show that inputting a square wave is just fine. If I use a square wave (lets say off of a microcontroller pin) as the frequency source (40Khz of course), will that work fine for driving this sort of thing after being amplified to the proper voltage? Or do I need to make a sine wave generator at that frequency?
* I understand that these things apparently create quite a bit of back-emf spike. Is there a more effective way than a 100 ohm resistor and a zener to take care of this? Its been a while since I have considered this kind of thing. Most of what I have designed for myself have been digital circuits that didn't need to take care of this a lot and I don't have a lot of money to be burning out parts because of back-emf spikes.
* The pins seem to imply that this thing is polarized in some way (one is longer than the other). Is this thing polarized? I was under the impression that piezo-based things weren't exactly polarized (I probably am quite incorrect).

There may be more as things come up, but a lot of my questions can be answered by hooking this thing up to an oscilloscope and seeing what happens.

Hey,

In your case if you want to generate a 20V p-p to drive the transducer, its easiest to go with the 555 method as jiripolivka suggested. However, it may be a bit harder with a microcontroller alone. However, with or without a controller, you can go with a similar method to the one stated in this TI application note. I've found that it does the job with my application requirements.
https://www.ti.com/mcu/docs/litabsmultiplefilelist.tsp?sectionId=96&tabId=1502&literatureNumber=slaa136a&docCategoryId=1&familyId=914

Cheers

 

[Los Frijoles]

Wow, that is an excellent application note and is very helpful. I may try both the 555 and TI driving ideas since they are both very similar and require minimal parts. My transducers are small, so I imagine it won't take an incredible amount to drive them.

 

[jiripolivka]

Wow, that is an excellent application note and is very helpful. I may try both the 555 and TI driving ideas since they are both very similar and require minimal parts. My transducers are small, so I imagine it won't take an incredible amount to drive them.

The piezo ultrasonic transducers are behaving like a ceramic capacitor. At resonance, they become resistive in their equivalent circuit. As their load is only air, they transmit some power by ultrasound with not much of efficiency. I used the 24 and40 kHz devices to transmit simple commands over a couple of meters, with a 555 as a generator, and some IRF power FET as the power amplifier. The FET was powered by 12 VDC through a transformer on a ferrite core, and at resonance the voltage peaked over 100 V across the transducer.
As I wrote, to check the function of my transmitter, I used another transducer connected to an oscilloscope. At resonance (and quite sharp, like < 1%) the system response was clear and the waveform on the scope was sinusoidal. On the transmitting transducer, the waveform was rather pulse-like. I think it does not matter if one drives the transducer by a rectangular or sinusoidal voltage.

I did not make a high-power "cleaning" type devices but I repaired one. The power driver is typically a switching-type power amplifier, with a transformer. The water-loaded transducers do not need an exact frequency ; the resonance is wider ~5%, and efficiency, too. A typical electrical drive power is 50-100 W and liquid cooling removes the heat. It is not recommended to run them dry.

When power FETs are used with not exactly resistive loads, they need the overvoltage protection. Some devices have built-in diodes. To be sure, connect one more diode to swallow the reactive overvoltage, and select rather high-voltage device to prevent burnout. Also, be sure you do not touch the live line at operation, 100 V peak and more hurts quite well. Also, do not expose your ears to the high-power ultrasound. You cannot heAr it but it can hurt your inner ear.

 

[phongphanp]

Where can I get psychology paper for safe usage ? I had heard about calculation the maximum power output in DB rating. Many application just as in datasheet is interesting , if we can predict source of error, we found the way to cope with them is good thing to adjust your circuit for all reason: specific work: music consert in engineer work , safe for human perception in medical work .

I had found new chip paper , you may make the look up ,again .

In some Physics book had omit this sound content, because this subject has property like electromagnetic theory. That for Ideal physician , I think . Nowadays , the ultrasound scan had role on medical in the second-third world country than the industrial country . :grin: