# Ultrasonic transducer driver

#### marestudio2004

##### Newbie level 4
Hello folks! I am new in this area in analog domain...

I have piezo ceramic ultrasonic transducer for humidifier...

Specs:
Voltage: 5-12V
Res. Freq: 1,7 MHz
Res. Impedance: <2""> I dont now what is this...
Coupling factor> 52%
Static capacitance: 1800pF

I want to make driver for this transducer as simple as posible. Do you have any idea?

##### Super Moderator
Staff member
A full H-bridge is the convenient method to create AC from a single DC supply. It's a variation of the astable multivibrator.

You can adjust values so large supply current goes through a load.

Your specs mention a resonant frequency. To clarify how oscillations are created through a load, my simulation has an inductor & capacitor load.

The op amp automatically detects the LC resonant frequency. It does so by sensing zero crossings in the low-ohm resistor.

marestudio2004

### marestudio2004

points: 2

#### marestudio2004

##### Newbie level 4
Thanks! This is fantastic!:razz:

Can this example generate 1,7MHz?

##### Super Moderator
Staff member
Can this example generate 1,7MHz?
My schematic illustrates a concept of operation, an automatic method to detect resonant frequency. Where I have a series LC, there needs to be the correct model for your transducer.

Or instead you may decide it's easier to construct a 1.7 MHz oscillator, to drive a transistor (or mosfet), which sends pulses of current to your transducer. There are types of the 555 timer IC which can create up to 2 MHz.

It's a question what is the proper waveform to apply, whether true AC, or whether pulsed DC.

You also need to consider what components can operate at the speed you require. Not every op amp can attain 1.7 MHz.

#### marestudio2004

##### Newbie level 4
Specs of driver:

Input: 9-12v DC
Output: Sinewave (AC), 9-12v, Freq of signal 1.7MHz

##### Super Moderator
Staff member
This driver starts with a square-wave oscillator. Frequency is adjustable. (It does not auto-detect resonant frequency.) You must find a frequency which gets best response from the transducer.

A capacitor filter turns it into a quasi-sine. A series capacitor blocks DC so that AC bias is applied to the half-bridge.

If you're willing to make a +10V -10V bipolar supply, this method is simpler than driving a full H-bridge.

#### KlausST

##### Super Moderator
Staff member
Hi,

There are a couple of mistakes.

Klaus

#### marestudio2004

##### Newbie level 4
Hi,

There are a couple of mistakes.

Klaus

#### KlausST

##### Super Moderator
Staff member
Hi,

What´s the idea of V1?
It seems the schematic is made with a simulation tool. Why not simply start the simulation to see if/how it works?

Mind: You post a schematic (named: "This is nice solution") in a forum. Many people see it and try to build it.
Please feel responsible for the informations you post. Thus my kind recommendation to verify it before you post it.

Klaus

#### marestudio2004

##### Newbie level 4
Okay, sorry for the writing method ...

The main idea is described in my first post.

So a driver for a ceramic piezo transducer with the given characteristics.

I have mentioned several times before that I am not an expert in this field.

My posts are mostly about finding a solution (i.e. help ), from people who are experts in this field.

Given scheme is an example of the Cholpitts OpAmp oscillator, simulated in Proteus. The simulation does not work properly.

V01 is just internal name of this example...

#### KlausST

##### Super Moderator
Staff member
Hi,

V01 is just internal name of this example...
No, "V1" is the name of a voltage source connected to the output of the amplifier.

***
I repect that you not that experienced, and that you are looking for a solution: And this is all right the way it is - nobody is perfect.
But on the other hand you write "This is nice solution" (let´s me think that you tested this and it´s working well for you) and then "The simulation does not work properly".... somehow contradictory in my eyes.

Klaus

#### marestudio2004

##### Newbie level 4
Misunderstanding. The name of the PDF document eventually contains 01, I meant version 01 .... Never mind.

The forum should be a place that can help people. Unfortunately, I'm going to have to spend a lot of time reading professional literature to come up with a solution.

All the best.

#### KlausST

##### Super Moderator
Staff member
Hi,

I miss a "power" or "current" specification.

In the headline you ask for a "driver". But you post a schematic with an OPAMP used as socillator.
Is the OPAMP suitable as driver? Or are you still looking for a driver?

***

The forum should be a place that can help people.
I agree. (I´d like to underline the word "help".)
And a forum is meant that not only the OP gets help, but additionally all other visitors with similar problems by reading through your thread.

Unfortunately, I'm going to have to spend a lot of time reading professional literature to come up with a solution.
Yes - that´s the usual way. Call it unfortunaltely or sadly...

A forum can´t replace school and can't replace reading books or working through tutorials. And a forum is not meant that others do your job. It is meant to help.

****

I compare it with climbing on a mountain:
Let´s say you want to climb a mountain. You will find others also climbing the mountain.
And maybe there are some difficulties - then everybody will lend you a hand (for free) to help you to get over the problem.
But no one will carry you all the way up - unless you pay for it.

And maybe there is a difficult place on the way up the mountain and you find it useful (for you and for others) to fix a rope around a rock. (ie posting a schematic).
It should be tested and solid .. to give others a benefit. Otherwise others may get hurt more than without the rope.

Klaus

##### Super Moderator
Staff member
A Colpitts (or Clapp, its cousin) oscillator requires many adjustments of values, to get 1.7 MHz. Then to get oscillations started and sustained also requires a lot of adjusting. These produce a sinewave, however the waveform does not necessarily have the correct voltage range, or sufficient power, to send a few Watts at 10V through your transducer. (Internet articles suggest 2-4W of power for a humidifier transducer.)

A Colpitts is able to generate a bias waveform sufficient to drive a half-bridge or H-bridge. By doing more adjustments you might succeed at amplifying the sinewave sufficiently without distortion or clipping, as it drives your transducer.

In contrast, notice how easy it is to arrange two invert-gates (post #6). Make the resistor a potentiometer adjust.

My schematic shows one capacitor as a one stage low-pass filter, however if you want a better semblance of a sinewave, then add one or more similar stages.