Driving a piezo with a step-up transformer

[CPR]

I'm being offered everything but using a step up transformer. Is the consensus that it will not work? If so then I'm going to concentrate on just using the 12v I have for now.

Why don't you just try the step up transformer for yourself and then let us all know if it works? That's what I'd do.

[TimB]

As trastikata said

"It is unlikely you can find small transformer with the correct winding ratio for less than 2-3 USD. But then you have to consider the switching frequency of this transformer - the datasheet says the resonant frequency for the piezoelement is 26KHz. At 26KHz things get more complicated with the transformer and noise can appear."

[JonW]

Be careful with boost converters and low current drain applications as they may run discontinuous in light load conditions and suffer from very poor efficiencies if they are not designed specifically for it and may cause problems  If you are building one off products or small quantities and are not familiar with designing Boost converters then opt for a module that can handle light loads or load the output, alternatively drive your circuitry from a separate linear reg running from the boosted voltage to place the boost under the required load for optimal operation.  Also consider looking at the boost converter in the PICKIT and use the  PID routine Les wrote, load it to get good efficiency to produce your own boost, you literally only need an inductor, diode, BJT and a few low ESR ceramic capacitors.  Then drive the transducer from a single BJT via PWM.  There are also switched capacitor boost regs that are used for this purpose here.

Have you asked Murata about the drive circuit they use as the information only states what they drive their circuit from and not what's applied to the transducer itself

[rick.curl]

Hi Tim-
Take a look at a PAM8904E, made be Diodes, inc. It is a piezo driver that contains a charge pump.It can drive a piezo with 27Vp-p from a 4.5 volt source. Look here: https://www.diodes.com/assets/Datasheets/PAM8904E.pdf

-Rick

[TimB]

JONW

Yes Murata has a circuit, very expensive parts. But remember what I have been saying all the way through. I will only be driving the device for a few ms. Not mins not seconds but around 1-2ms.

If you use a charge pump you are going to have it running all the time. Then once every few seconds it is used for a couple of ms. If you turn the charge pump on when you need it, it takes time to build up the charge.

This is why I was looking at the pulse transformer. Ideally a piezo should be driven by AC (so it was pointed out to me). Or the disk is only flexing one way.

I'm going to try just direct driving it with 12v using a hbridge motor drive chip as I have one on a board I make so will save me time messing with development boards

[rick.curl]

Yes Murata has a circuit, very expensive parts. But remember what I have been saying all the way through. I will only be driving the device for a few ms. Not mins not seconds but around 1-2ms.

The PAM9804 is only 74 cents each, quantity 10, and it gets down to 30 cents each for a full reel.

If you use a charge pump you are going to have it running all the time. Then once every few seconds it is used for a couple of ms. If you turn the charge pump on when you need it, it takes time to build up the charge.

In the PAM9804 the charge pump will run all the time but it only needs a few microamps. As for the H-bridge you mentioned, remember that the piezo device is purely capacitive, so it does not hurt to have DC voltage on it when it is not running.  Look at the block diagram on the PAM9804 spec sheet. 

-Rick

[shantanu@india]

How do electronic blood pressure monitors generate pneumatic pressure? My Omron machine generates a decent amount of pressure to stop the blood circulation in just 10 secs!
If it is peizo then it needs only 6V DC(4X1.5)

[ken_k]

Hi I missed all this post, I'm certain you have resolved your problem. Here is my offering. Stack a voltage doubler on top of your 12V rail, a gate drive IC running on 12VDC will provide 6VAC to drive it, if the circuit is driven from a PIC charge up the 16V cap only when needed.
I cannot see how to upload a schematic so here is a spice file to the schematic this may not work cut and paste into note pad, change the .txt to .asc and it might!! run in LTspice.
Ken K

Version 4
SHEET 1 880 680
WIRE 384 48 288 48
WIRE 624 48 384 48
WIRE 688 48 624 48
WIRE 288 64 288 48
WIRE 384 112 384 48
WIRE 288 144 288 128
WIRE 288 144 128 144
WIRE 288 160 288 144
WIRE 624 192 624 48
WIRE 624 192 528 192
WIRE 288 240 288 224
WIRE 288 240 -256 240
WIRE 384 240 384 176
WIRE 384 240 288 240
WIRE 528 240 528 192
WIRE 624 240 624 192
WIRE 128 272 128 224
WIRE -256 320 -256 240
WIRE 128 336 96 336
WIRE 32 352 -112 352
WIRE 528 384 528 304
WIRE 624 384 624 320
WIRE 624 384 528 384
WIRE -112 400 -112 352
WIRE -256 512 -256 400
WIRE -112 512 -112 480
WIRE -112 512 -256 512
WIRE 32 512 32 384
WIRE 32 512 -112 512
WIRE 624 512 624 384
WIRE 624 512 32 512
WIRE -256 560 -256 512
FLAG -256 560 0
FLAG 688 48 Vout_16VDC
IOPIN 688 48 Out
SYMBOL schottky 304 128 R180
WINDOW 0 24 64 Left 2
WINDOW 3 24 0 Left 2
SYMATTR InstName D2
SYMATTR Value 1N5817
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL schottky 304 224 R180
WINDOW 0 24 64 Left 2
WINDOW 3 24 0 Left 2
SYMATTR InstName D1
SYMATTR Value 1N5817
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL cap 368 112 R0
SYMATTR InstName C1
SYMATTR Value 220n
SYMBOL voltage -256 304 R0
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR InstName V1
SYMATTR Value 12
SYMBOL voltage -112 384 R0
WINDOW 3 -103 206 Left 2
WINDOW 123 0 0 Left 0
WINDOW 39 0 0 Left 0
SYMATTR Value PULSE(0 12 1n 10n 10n 50u 100u)
SYMATTR InstName V2
SYMBOL Digital\\buf 32 288 R0
WINDOW 3 0 0 Invisible 0
SYMATTR Value Vhigh=12
SYMATTR InstName A1
SYMBOL cap 112 272 R0
SYMATTR InstName C2
SYMATTR Value 220n
SYMBOL res 608 224 R0
SYMATTR InstName R1
SYMATTR Value 1k
SYMBOL res 112 128 R0
SYMATTR InstName R2
SYMATTR Value 10
SYMBOL cap 512 240 R0
SYMATTR InstName C3
SYMATTR Value 10µ
TEXT 216 32 Left 2 ;schottky diode
TEXT -168 536 Left 2 ;maybe a PIC
TEXT -344 552 Left 2 !.tran 1
TEXT -152 224 Left 2 ;12V rail
TEXT -96 344 Left 2 ;10kHz
TEXT 48 400 Left 2 ;FET GateDrive IC on\na 12V rail
TEXT 656 344 Left 2 ;test load
TEXT -72 112 Left 2 ;R2 is a current limit\nfor the gate drive IC
RECTANGLE Normal -32 576 -192 368 2

[ken_k]

I forgot you will have to name text text file as something. Maybe 16VDC.txt and then change it to 16VDC.asc

[shantanu@india]

Hi Ken,
Seeing you first time in this new forum.
Hope all is well in the Kangaroo Continent!!

[ken_k]

Hi Ken,
Seeing you first time in this new forum.
Hope all is well in the Kangaroo Continent!!

H Shantanu
All is well in Australia, thank you for asking. Did the file open OK in LTspice?

[shantanu@india]

I'm not a regular user of PS spice or LT so I cannot comment.
Do you need license for it?

[shantanu@india]

Downloaded LYspice XVII v16 but the 'Run' failed with the message'multiple instances of flag'

[ken_k]

Thanks for giving that a try. LTspice is very popular and worth learning, it is one of the best free programs out there.

I tried saving the text from the forum back to my machine and it ran ok.
I don't want to waste your time.
Here is another attempt. Copy from forum open notepad and paste text save as power.txt the after saving rename power.asc and see if it opens in LTsice. If it fails to work I will try downloading it on my wife's laptop and see if it works. BTW I simplified the schematic.
Regards
Ken K

This failed I will try again.

[shantanu@india]

Thanks Ken...it worked...screenshot attached.
Learnt something new from you today.
Shall try to start from the basics

[ken_k]

Excellent!

Basic LTspice Instructions:

open file

right mouse click and press run.

mouse over any node and left click to have a look.

The "scope" view screen can be magnified by dragging a rectangle out with a left click.



At the tabs at the top of the screen click "simulate"
click edit simulation command to change things around.

Regards
Ken K

[See_Mos]

Tim, you are correct about driving the piezo with AC, consider it as a capacitor that has to be charged then discharged.

[JonW]

You should be able to drive it from a pulse transformer with centre tap to VCC and drive the primary from a H bridge and secondary with the required turns ratio, most can do tens if not hundreds of kHz and you will discharge ok with a push pull or H bridge driver.  Many years ago we used to use this method for research in piezo uses in ion thrusters and it worked pretty well.  As mentioned the switched cap method and a single BJT with a resistor in parallel with the transducer should also work, it will be low cost and be efficient.

Should be easy to test the transformer method on breadboard quickly.  There are many 3:1 or more pulse transformers available from the disti's and you can construct a model of the Piezo (if you don't have one) knowing the capacitance, resonance and resistance at resonance.

[See_Mos]

Some years ago I was tasked with designing a unit to test sonar transponders at specific frequencies in the 25KHz to 35KHz range.  This required a high power transducer and I did some research and came across this web page:-

http://www.midnightscience.com/ultra-articles.html

From the main page click on 'Ultrasonic Articles and download' then look for '40KHz Piezo Ceramic Transducers'

I decided that optimizing the drive was too much work so I used the brute force method shown here. It works as a full bridge driver thus increasing the effective drive to the transducer. Microchip AN799 has some useful information on the TC4428.

The unit worked on a nine volt battery and I looked at using the MC34063 to boost the supply voltage but in the end this was not needed.

SCR pulse transformers work surprisingly well at higher frequencies so your idea of using one might work.  You could try single FET flyback topology or connect the primary in place of the piezo in the circuit below.  As Jon said you could use two to one plus one with the secondaries wired in series to double the voltage or a three to one wired in reverse.
 

[ken_k]

Years ago I designed and sold some therapeutic ultrasounds, all were required to deliver 15 watts acoustic into water loads at 1.1 MHz.
The topologies used are as shown. The first circuit allows for a good sine wave at over 100 volts.
The second simple circuit does the job and can also provide a high output voltage, the waveform is a little ugly, it may be the best option for a mini air pump.
This is a good sight for design calculations.
https://home.sandiego.edu/~ekim/e194rfs01/jwmatcher/matcher2.html

If you go down this path it pays to capacitor couple the FET gate drive so if the PIC stops with drive to the FET no smoke is released.