Connect with us

How to amplify a square wave cheaply?

Discussion in 'General Electronics Discussion' started by plukens, May 23, 2012.

  1. plukens

    plukens

    8
    0
    May 23, 2012
    I need to amplify a square wave that is driving a piezo disk at 100 Hz (or less). I'm limited by the output of my function generator to 10 V. I need 50 or even 100 V. What is a cheap way of amplifying the output of my signal generator that will preserve the square wave at 100 Hz? Will an audio amplifier work? They are intended to drive low impedence loads (for instance 8 ohm). How will that work with a high impedence load (my piezo disk)?

    My budget is $50 or less. OK, $100 or less max.

    Thanks for your help.

    Peter
     
  2. Harald Kapp

    Harald Kapp Moderator Moderator

    9,136
    1,844
    Nov 17, 2011
    You don't need a fancy circuit. As simple one can do the job:
    [​IMG]
    For the transistor use any suitable NPN, suitable meaning: collector-emitter breakdown voltage less than battery voltage, collector current at least as high a the current through the piezo. For a current < 60 mA use for example this one http://www.onsemi.com/PowerSolutions/product.do?id=MMBT5551M3T5G or any similar type.
    The base resistor is calculated from
    RB= (10 V -0.7 V)/Icollector * Gain
    where:
    - 10 V = square wave generator output voltage (0 ... 10 V=
    - 0.7 V = base-emitter voltage of the transistor
    - Gain is the gain of the transistor at the specified collector current (for the transistor mentioned above it is Gain=30 at 50 mA).

    If the signal generator cannot source enough current, you need to select a transistor with higher gain or use a Darlington pair (http://en.wikipedia.org/wiki/Darlington_transistor).

    Cost should be less than 1$.

    Harald
     

    Attached Files:

  3. Wabajig

    Wabajig

    75
    0
    Apr 14, 2012
    Not a safe idea, you don't want a 100V where you are hooking up speakers. I'm not sure how to input schematics on this forum, but I'll try to explain verbally. In the above schematic where the piezo is use a high gain PNP with the emmitter tied to the + supply, and where the NPN transistor is you put your piezo to ground. Now use a digital transistor for input into base and tie the collector to the base of the high gain output and emmitter to ground. Experiment with that, it would be safer.
     
  4. TedA

    TedA

    156
    16
    Sep 26, 2011
    plukens,

    Can you give us some more information?

    What is the impedance of your piezo disk?

    ( Often these look mostly like a capacitance.)

    Is your 50 or 100 volt figure average, or peak to peak?

    How fast a risetime do you require?

    Will the piezo disc tolerate a DC component on the driving signal?

    Do you have a variable DC source for power?

    ( Such that the output can just swing between rails.)

    Or, does the amplifier have to define the squarewave output voltage?

    Do you really mean 100Hz? We often think of piezo transducers as being for higher frequencies.

    I should note that the circuit shown by Harald might be good for only one click, if the piezo device has high DC impedance.

    Ted
     
  5. plukens

    plukens

    8
    0
    May 23, 2012
    DC resistance is Mega ohms, it's just a ceramic disk bonded to a brass disk (which acts like a spring when it vibrates).

    The impedence at 100Hz (my target freq. to drive the disk) would be 1/wC = 57 KOhm.

    My signal generator produces a square wave at 10V p-p and I wish to amplify that up to a max of 50 V p-p. Or maybe as high as 100 V p-p to experiment with the disk behavior.

    Yes, just 100 Hz. I'm driving a liquid droplet generator and want a stream of drops at about 100/sec or 100 Hz.

    I'm sure the piezo can withstand a DC component, it's just a ceramic disk.

    Ideally,I'd like an off the shelf solution that I can drive with my signal generator, rather than build something.

    Not sure of the required rise time... The sharper the edge, the better the droplet formation Since I wish to drive it at 100 Hz or less, I probably need an amplifier bandwidth of 3x that or better yet 10x that? (10 KHz)

    Thanks for helping!
     
    Last edited: May 23, 2012
  6. TedA

    TedA

    156
    16
    Sep 26, 2011
    plukens,

    Thanks for answering my questions.

    Sounds as if you really want a box that plugs in the wall for power, and has connectors for in and out cables, rather than a construction project.

    Nothing comes to mind immediately, in the way of inexpensive off the shelf hardware. I'm sure someone makes just the thing, but not for cheap.

    It would not be a very complex construction project, using inexpensive parts, to make such a device from scratch. But this would require some electronics and construction skills, as well as time.

    If you don't need a really good squarewave, you might try an audio amplifier driving a step-up transformer. A stereo or PA amplifier good for a few watts, plus an audio transformer or low voltage power transformer turned around to step up the voltage. Perhaps a damping resistor or network.

    Another question: Does the disc remain displaced for 5ms, after each edge of the squarewave, or does it relax toward the unenergized position? I'm wondering if you could get away with more of a narrow pulse, every 10ms.

    Do you know any folks named Chambers in San Markos? Bill was an Uncle, and Helen is still there, I think. Cousins nearby, too.

    Talk to you later.

    Ted
     
  7. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    Harald, are you sure about this? I've tried it as you've shown but it produced very little drive to the Piezo. You're not providing a DC path for any Collector current. These are the two options I use.

    Chris
     

    Attached Files:

  8. Wabajig

    Wabajig

    75
    0
    Apr 14, 2012
    It just might not be responsive at that frequency. Do you have the specs on your piezo?
     
  9. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    I just re-read this. I could be wrong, though I doubt it. Common Piezos (CYA) aren't meant to operate at low frequencies like your driving it with. They have a resonant frequency, where their output will be maximum; see my previous post. If you think the output is low now, try switching your generator to Sine wave. It should produce just about no audio output at all. The reason you hear anything now is because the Piezo is trying to resonate to its natural frequency on each rising edge of your 100Hz square wave.

    If you're going to drive a Piezo at something other than its resonant frequency it's best to drive them from a pulse source rather than a square wave. Assuming that your function gen has a 50 Ohm output impedance I'd try it with a switching diode in series with the Piezo. The diode should prevent the low generator impedance from loading the Piezo after the pulse has goosed it into resonance and is long gone till the next one comes along.
     
  10. plukens

    plukens

    8
    0
    May 23, 2012
    Thanks for the feedback... (bad pun). I'm using the disk as a part of a droplet generator tha produces droplets of liquid. I want to produce single drops (1 pulse) or trains of droplets at 100 Hz (100 Hz square wave). The literature that I'm following finds success with a square wave. I believe the reason is that the rapid acceleration forward then backward due to the sharp rise and sharp fall causes the liquid to pinch off at the nozzle tip and eject a droplet. I am not trying to drive this disk in resonance. Or maximize the power I couple into the media (like you would with an ultrasonic cleaner). I am using it as a diaphram pump.

    The amplitude of the displacement of the disk is a function of the amplitude of the voltage applied. The literature gets theirs to work at 30 V p-p. Thus, I need to amplify the 10 V p-p from my signal generator.

    I will play around with the duty cycle to determine the best pulse width to produce uniform droplets. Right now, I can see the liquid miniscus vibrating at the nozzle tip, but droplets are not being ejected.

    Thanks for the help.
     
  11. plukens

    plukens

    8
    0
    May 23, 2012
    Someone suggested using a low voltage square wave from my signal generator to turn a MOSFET switch on and off, allowing a higher voltage supply to drive my disk. Since I'm only seeking 100 Hz or less, perhaps this is a good way to go? Will this unit do the trick?

    http://www.ebay.com/itm/130611734387?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

    So then I need a 50V DC supply to be switched?
    Any suggestions? Ideally, I'd like a variable Voltage supply. Do inexpensive power supplies go as high as 50 (or even 100) volts?
     
  12. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    Last edited: May 24, 2012
  13. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    Oops! If your gen is outputting +- 10V you'll want a 1N914 on the base of the driver transistor. Connect the Cathode to the Base and Anode to GND. Most BJTs have a maximum Vbe (reverse bias) of 5V. The diode with protect it from exceeding -.7V when the gen goes negative. ;)
     
  14. Ian

    Ian Administrator

    1,476
    427
    Aug 23, 2006
    Let me see what I can do about the permalinks :)

    edit: permalinks should now be visible to members (in the top-right corner).
     
    Last edited: May 24, 2012
  15. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    Google PA240U OpAmp data sheet. It's a HV/HP driver. Even includes a sample Piezo driver circuit.
     
  16. CDRIVE

    CDRIVE Hauling 10' pipe on a Trek Shift3

    4,960
    648
    May 8, 2012
    Wow, that was quick!! Can we have a Scientific Notation Editor too, or am I pushing my luck? Sure would be nice to have an Ohms, Deg, Delta, etc characters. :D
     
  17. BobK

    BobK

    7,599
    1,641
    Jan 5, 2010
    And while you are at it, schematic drawing.

    Bob
     
  18. Harald Kapp

    Harald Kapp Moderator Moderator

    9,136
    1,844
    Nov 17, 2011
    Been away for a while...

    O.K., my simple circuit misses a path to discharge the piezo once the transistor has opened the first time. A parallel resistor to the piezo should do the trick. Or one of the circuits as suggested by Cdrive.

    The 100 Hz issue is something different. You probably don't need a single 100 Hz signal to generate droplets at 100 droplets per second. As others have mentioned this is not the best operating frequency for the piezo. This is, as far as I know, also not the operating mode of piezo vaporizers. I think you should operate the piezo at least at 20 kHz or even more. This will generate the vaporizer effect without being audible. This 20 kHz frequency should be switched on and off at a rate of 100 Hz. So you have 100 droplet generating pulse pakets per second.

    Harald
     
  19. plukens

    plukens

    8
    0
    May 23, 2012
    Thanks for the circuit help.
    This is not a vaporizer creating a mist. It uses the piezo disk, bonded to a brass disk, as a diaphram which deforms with voltage across it. This pumps liquid out of a small orifice. When the voltage is switched off, the diaphram returns to its equilibrium position (flat). There is an ideal frequency (based on orifice size and surface tension of the liquid) for droplet formation, but as the literature shows, you can produce single droplets. No resonance is required. Just simple mechanical displacement of the liquid in one direction (out the nozzle) and rapid acceleration in the opposite direction (suck back) which breaks the column of liquid and pinches off a drop which then falls under gravity. Droplet size is determined by orifice size. Droplet diameter is roughly twice orifice diameter for water. For reference, here is the device I'm building. Figure 2 shows the setup:

    http://www.nist.gov/el/fire_research/upload/Yang-A-Simple-Piezoelectric-Droplet-Generator.pdf
     
    Last edited: May 25, 2012
  20. plukens

    plukens

    8
    0
    May 23, 2012
    Thanks for pointing me in this direction.
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-