Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Wave Generator with Crystal Quartz Oscillator?

  1. Feb 11, 2015 #1
    Hi All!
    A little background info. I'm working with cardiac pacemaking devices with the overall scope of implementing piezoelectric energy harvesters into the devices. I'm in the initial stages of research and development.

    Okay, so this is my conundrum: I have an output pulse circuit as shown on page 13 of the attached journal (Fig 11.7). Essentially, the circuit acts as a voltage doubler by supplying controlled pulses of -2Vdd to the output (heart). The circuit requires a "pulse of sufficient magnitude to place the transistor [Q1] in the conductive state" thus turning the gears of the circuit's functionality (from steady state to conductive).

    Here in lies my problem. In my prototype I have used a 13.0 MHz 20 pF crystal (digikey# 887-2036-ND) as the microprocessor/pulse source and, upon initial testing, received garbage results. I was informed that I would probably need to connect a wave (square) generator to the crystal for proper functionality. I have tried testing the circuit with just the crystal, just the wave generator, and with the generator supplying waves to the crystal. Any thoughts/advice, I'm pretty stumped.

    Also, I'm certain the circuit is put together correctly as I sought help from an EE PhD student and similarly found the correct voltage drops across each capacitor during the steady state. Thanks!

    https://galileo.seas.harvard.edu/images/material/3794/1321/CardiacPacemakerdesign.pdf

    QL
     
    Last edited: Feb 11, 2015
  2. jcsd
  3. Feb 11, 2015 #2

    Svein

    User Avatar
    Science Advisor

    I have looked at your circuitry and it does not make sense.
    upload_2015-2-11_22-9-41.png
    Here is a simulation of your circuit (I have guessed at some parameters, and repeated the "on" pulse). If you want a voltage doubler, try this circuit:
    upload_2015-2-11_22-23-43.png
    Feed it with a square wave of somewhere between 1 and 100 kHz. A simulation of the circuit is shown here:
    upload_2015-2-11_22-30-8.png
     
  4. Feb 12, 2015 #3

    Baluncore

    User Avatar
    Science Advisor

    The circuit does work. It needs time to recover between triggers. Attached is pacemaker_1.txt rename to .asc to run with LTspice.
    Pace 1.png
     

    Attached Files:

  5. Feb 12, 2015 #4

    Svein

    User Avatar
    Science Advisor

    OK. To me it looks like an AC-coupled cascode circuit, which is basically a high frequency amplifier.
    I may have gotten some of the values wrong in my simulation (it wouldn't have been the first time).
     
  6. Feb 12, 2015 #5

    Baluncore

    User Avatar
    Science Advisor

    It is not really cascode because the base of Q2 is not clamped to a fixed DC potential.
    I would call it a Marx Impulse Generator because it switches parallel capacitors to series connection.
    http://en.wikipedia.org/wiki/Impulse_generator#Marx_generator

    It needs a low impedance path through the body to charge and discharge the system. That is my R6 = 1k.
    C1 and C2 are gradually charged to Vdd.
    When Q1 turns on, C1 generates –Vdd on the emitter of Q2.
    Q2 is then turned on and bootstraps C2 down to –2*Vdd.
    So it is a pulse voltage doubler that operates by charging two capacitors in parallel.
    When triggered it connects them in series. Hence the Marx topology.

    Note: my value of R5 should be 15k, the values of R2 and R4 should have been 33k, but it makes no real difference to the simulation or function.

    The problem with the OPs prototype was probably that a short positive trigger pulse was not applied to the base of Q1 through a limiting resistor each second.
     
  7. Feb 15, 2015 #6
    Thank you all for the help and taking the time to test the system. You are correct in that I was pulsing at too high a frequency and have not included a limiting resistor. I will make adjustments and test it myself on Monday!

    QL
     
  8. Feb 15, 2015 #7

    Baluncore

    User Avatar
    Science Advisor

    The trigger pulse should only remain high for a short time. The long period of low between pulses is needed to recharge the capacitors for the next pulse. This is a self limiting feature of the circuit. Be aware that without a base current limiting resistor your input transistor may have died earlier.

    You can download LTspice for free directly from the semiconductor manufacturers website; http://www.linear.com/designtools/software/

    Let us know how you get on with testing your circuit. I expect you to ask more questions here if you have problems.
     
  9. Feb 25, 2015 #8
    Sorry for the extended absence, busy with other courses and had to order a few components. Anyway, the circuit tested fine so I appreciate the help and input from everyone. The new issue is implementing into practical use. I have a few power management circuits that I'm currently looking at which, under sufficient harvesting, will provide Vdd to the circuit above which in turn will give the desired pulse to the heart. Since, in the overall scope, I'm working with very limited space I need to think conservatively and be space conscious. All in all its an autonomous system provided the mechanical vibrations to the piezoelectric harvester.

    The issue at hand is to make this fully inclusive; meaning, I need to find an autonomous wave/oscillation supply, i.e. a crystal quartz properly integrated into the circuit in place of a function generator (hence the original post title). Any sort of micro-controller is out of the question as these require power and consume far to much space. I've looked into MEMS but I'm cautious and hesitant due to their lack of robustness and wide spread use. Thus, the only logical/practical/relevant choice is the crystal. The problem is I don't know how to implement this in place of the trigger while keeping the specifications I desire from a trigger. I'm looking into the problem presently and would appreciate any feedback/suggestions, thanks!

    QL
     
  10. Feb 25, 2015 #9

    Baluncore

    User Avatar
    Science Advisor

    A digital watch contains a ceramic or quartz crystal oscillator and a divider chain that produces one second pulses.
    It uses very little power because it runs on very low voltage at a low frequency.
    The crystal frequencies used are usually typically 32,768. Hz (=2^15) or 32,000.Hz (=2^5 * 10^3).
     
  11. Feb 26, 2015 #10

    Svein

    User Avatar
    Science Advisor

    electronique_base_temps_001.gif
    This is an example of a low-frequency, low-power crystal oscillator of the type Baluncore mentioned.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Wave Generator with Crystal Quartz Oscillator?
Loading...