hi all...
I need to drive a special piezo air pump https://www.murata.com/en-global/products/mechatronics/fluid/overview/lineup/microblower_mzb1001t02
I will have 12v and 5v on my board I have free P1 and P2 on the pic to generate a half bridge PWM. As the device will run from 8-18v I was going to use a couple of transistors PNP/NPN to drive it.
Then I though what about a pulse transformer! A 0.333 wind direct to the pic pins may be enough to get 15v or there abouts.
Does that sound feasible? If so any pointers to what type would do? BTW it will only be driven for < 1ms and a duty of about every 3-5 seconds
Thanks Tim
Interesting device Tim!
We are regularly driving peizo buzzers directly from 5V/12V without isolation which is structurally similar to an air blower. So isolation through a pulse transformer/ optocoupler is definitely not mandatory. On the other hand there remains a possibility of impedance mismatch when using a pulse transformer which might result in incomplete energy transfer.
Direct drive is better. Ensure that no nasty spikes are generated by the peizo(freewheeling/P6KE??)
Hi shantanu
My aim is not to isolate using the transformer, it was to step up the voltage I need ~18v ideally. I have 12v and 5V if I can step up the 5v to 18v direct from the pic. It would be ideal.
Tim
Quote from: TimB on Sep 11, 2022, 06:26 PMMy aim is not to isolate using the transformer, it was to step up the voltage I need ~18v ideally. I have 12v and 5V if I can step up the 5v to 18v direct from the pic. It would be ideal.
Hi TimB, there are many step-up converters, requiring very few external passive components - is it not feasible in the design to use one such IC?
Hi
My thoughts are to use a step I need
1 A step up chip
2 Inductors etc for the step up
3 Switching circuit of now step upped voltage to drive the piezo
If you use a step up transformer its just one component if it is of cause possible
I have seen a video of it using a step up
I think I will investigate that
Tim
Quote from: TimB on Sep 11, 2022, 06:43 PMHi
My thoughts are to use a step I need
1 A step up chip
2 Inductors etc for the step up
3 Switching circuit of now step upped voltage to drive the piezo
If you use a step up transformer its just one component if it is of cause possible
Using a transformer you still will need the bridge driver to convert the DC to AC voltage.
In first case you have DC-DC step-up IC followed by an DC-AC bridge converter.
In the second case you have the DC-AC bridge converter followed by the step-up transformer.
Remember that I have a half hbridge output from the pic so while its not true AC it will swing around 2.5v
Very interesting device, I wonder if it can be used for CPAP devices.
Quote from: Parmin on Sep 11, 2022, 11:06 PMVery interesting device, I wonder if it can be used for CPAP devices.
Normal breath about 500ml. For say 12 breaths per minute = 6L/min.
This device 1.2L/min.
John
Quote from: John Drew on Sep 12, 2022, 12:54 AMNormal breath about 500ml. For say 12 breaths per minute = 6L/min.
This device 1.2L/min.
John
I agree if the device is to be used to SUPPLY breathing air then the flow is not enough.
On the other hand, CPAP is to provide pressure rather than supply air.
If somehow the pressure can be maintained then the system should work, I think.
There is also the possibility to chain say 5 devices to provide a higher flow rate.
Nevertheless, this is just a thought to explore posibilities ;)
It seems that the CPAP possibilities for this device has been explored by this company https://www.qtflow.co/
I believe a few others are doing the same as well.
Quote from: TimB on Sep 11, 2022, 06:26 PMHi shantanu
My aim is not to isolate using the transformer, it was to step up the voltage I need ~18v ideally. I have 12v and 5V if I can step up the 5v to 18v direct from the pic. It would be ideal.
Tim
Then better use a standard boost converter. I used to design flyback power supplies myself using TOPSWITCH and buck converters using Viper 22A myself a few years back. Now modules like these are so so cheap...
https://www.electronicscomp.com/mt3608-2a-max-dc-dc-step-up-power-module-booster-power-module?gclid=Cj0KCQjwjvaYBhDlARIsAO8PkE1xsTc1h-kqU9n1Smr0ZvyAueD7W1hEPGNMzPu50RFiqLCucp19bY4aAt8EEALw_wcB
My requirement with this device is to generate micro bubbles one at a time. So a step up chip with all the components and cost is IMHO a waste
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.
Tim,
I suggest you try out your pulse transformer idea with your pump and see if it works. But instinctively I feel that there would be a significant loss since pulse transformers are basically used for triggering (thyristor gating etc.)which traditionally consumes very little energy.
BTW.. these pumps can replace the electromechanical monsters that I use in my fish aquarium.
https://www.aliexpress.com/item/1005002234592016.html?spm=a2g0o.detail.1000060.3.5dab6896LaaBAo&gps-id=pcDetailBottomMoreThisSeller&scm=1007.13339.300834.0&scm_id=1007.13339.300834.0&scm-url=1007.13339.300834.0&pvid=f3d56f6d-ebd0-4c7e-87b0-8f50770551a3&_t=gps-id%3ApcDetailBottomMoreThisSeller%2Cscm-url%3A1007.13339.300834.0%2Cpvid%3Af3d56f6d-ebd0-4c7e-87b0-8f50770551a3%2Ctpp_buckets%3A668%232846%238116%232002&pdp_ext_f=%7B%22sku_id%22%3A%2212000019467558116%22%2C%22sceneId%22%3A%223339%22%7D&pdp_npi=2%40dis%21BRL%212.22%211.95%21%21%21%21%21%40210312eb16629890150357100e6aef%2112000019467558116%21rec&gatewayAdapt=glo2bra
0.4 USD :)
That price is insane. Multiturn preset pots cost me more than that.
John
Quote from: TimB on Sep 12, 2022, 07:20 AMSo a step up chip with all the components and cost is IMHO a waste
I'm being offered everything but using a step up transformer.
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.
Usually a simple higly integrated DC-DC step-up IC costs about .60$ -.70$, then you need a small inductor, a Schottky diode, two resistors and one or two small caps. BOM comes to 1USD ;) The most switching ICs work around the 1 MHz switching frequency, usually the noise is negligible.
trastikata
Thanks, I appreciate that info.
I'm going to try first switching with 12v as I need to drive it at a low level. I have 12v that the whole system runs off.
Tim
Quote from: TimB on Sep 12, 2022, 07:20 AMI'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.
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."
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 (https://e2e.ti.com/blogs_/b/powerhouse/posts/drive-a-piezoelectric-buzzer-with-a-simple-boost-converter).
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
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 (https://www.diodes.com/assets/Datasheets/PAM8904E.pdf)
-Rick
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
Quote from: TimB on Sep 13, 2022, 05:52 PMYes 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.
QuoteIf 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
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)
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
I forgot you will have to name text text file as something. Maybe 16VDC.txt and then change it to 16VDC.asc
Hi Ken,
Seeing you first time in this new forum.
Hope all is well in the Kangaroo Continent!!
Quote from: shantanu@india on Sep 14, 2022, 07:12 AMHi 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?
I'm not a regular user of PS spice or LT so I cannot comment.
Do you need license for it?
Downloaded LYspice XVII v16 but the 'Run' failed with the message'multiple instances of flag'
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.
Thanks Ken...it worked...screenshot attached.
Learnt something new from you today.
Shall try to start from the basics
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
Tim, you are correct about driving the piezo with AC, consider it as a capacitor that has to be charged then discharged.
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.
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.
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.
ultrasonic drive.jpg
I expect there are many in the forum that a much better at RF design than myself. 26kHz is quite a reasonable frequency to work with. To make calculations easier and more predictable it might pay to add a load resistor across the transducer, power can be controlled by PWM, Q and moving off frequency to reduce power. Some sort of gate drive I.C. will be required at 26kHz, I used a very simple NPN PNP emitter follower buffer at 1.1 MHz (cost way less than a chip).
BTW the 1.1MHz transducers were PZT ceramic transducers they had both a series and parallel resonant frequencies, both worked fine for driving.
Ken
It just dawned on me that you don't need a transformer, charge pump, or H-bridge. Use a CMOS HEX inverter like a CD4049. Parallel three gates and have them drive another set of three paralleled gates. Tie the Piezo across the second set of gates. If you supply it with 12 volts this will give 24 volts P-P across the piezo.
(https://i.postimg.cc/NFsMSdf1/image-2022-09-14-203634030.png)
-Rick
Rick you are correct See_Mos also posted a similar higher current drive in post number 38.
My submission is very much an overkill, I see the maximum drive for the device is 30Vpp.
BTW this is what the waveform looks like from a small single FET circuit. If the load resistor is changed to 100 ohms the output will still be over 28Vpp across the load.
26kHz a.jpg
26kHz b.jpg
I would be very interested to see a spice model of the air pump, what is it's actual impedance?
Not much information is given.
Hi All thanks so much for the feed back
I had an old board that had a Brushed motor controller. So I wrote code to reverse the direction at approx. 22khz ( I cahned it later to 26)
What I noticed was that I was generating 25v PP AC of my 12v input. Well that is what my scope said. This was using a mains powerbrick
The test on the piezo pump was disappointing. A little buzzing but no air. So I changed to a desktop dev power supply and reduced the voltage to 8v and my PP dropped to 7v????
After upping the the voltage to 18v the device came into life a little, but nothing like it should do.
I conclude the device is really sensitive to the drive circuit so If I continue with it I will use the Murata design. Its not as simple as I hoped.
What I have to state is what I need to achieve. I have a flow of water around 0.4ml a second in a 2mm id pipe. I need to inject a micro bubble around 2mm long in the flow. It needs to be a whole bubble. From experience bubbles in small tubes are a Fing pain. They contract in volume with pressure, they change the nose shape with velocity. I have managed to develop what I call a bubble accumulator that joins little bubbles and makes them big.
Below are some images I have of the scope readings and what I'm looking at next. Using a venturi to suck a bubble into the feed line and use just a small solenoid valve to let the air in. Ideally I would use a solenoid and piston to "punch" a set volume bubble into the stream. But the volumes I'm working with mean its will be very hard.
I need to design a system I can produce in the 100's and be VERY reliable as they may run 24/7
meassured wave form.png
Venturi.png
Ken's post #39 reminded me of the time I had to repair our Branson ultrasonic tank which used the same resonant drive.
Tim,
Another method of bubble formation is to lower the pressure below vapour pressure of water at that temperature ( vacuum pump?).
The dissolved air in the water shall form natural bubbles.
But very difficult to control.
Maybe injecting a miniscule amount of say CaCO3 to create a bubble of CO2?
Hi all
So as a final update, I came up with a system. It meets and exceeds my expectations first in simplicity, reliability, cost and a few others.
I now have an outlet pipe that is approx 200mm long and hangs so water when it exits it, is causing a small draw on the line. I have a tee in the line and that is connected to a small solenoid valve. All I need to do is open the valve for around 50ms and the draw on the line sucks in the bubble.
Thanks every one for their help.
Tim
Ah, the venturi effect.
Nice one Tim!