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

Best way to produce a 10-40kHz square wave

  1. Oct 28, 2012 #1
    Hi! I should probably start off by saying that I did attempt to search for this, but I wasn't able to find a thread that had a similar question.

    I'm driving an RLC network (in effect) with a square wave to saturate a ferrite core. The image is pretty close to what I'm trying to achieve. I need the circuit to be in the frequency range of 10-40kHz to resonate the secondary winding with the capacitor, but I also need to be able to output enough current to saturate the core (apprx. 200mA?).

    Right now I'm using an LM392 opamp as a astable multivibrator to generate a square wave, but I haven't been able to get it to go over 2kHz...and the chip is limited to about 20mA on the output. I looked into 555 H-bridge circuits, but I need the circuit to produce a ±V square wave to saturate the core in both directions, and it didn't seem like the 555's would handle negative voltage. I looked at using a mixer and crystal oscillator to boost the frequency, but that still doesn't solve my current issue.

    Does anyone know of a good way to achieve this? I just didn't know the words to search for in looking for other options, so I thought this would be a good place to ask for help.

    Thanks for your time!

    Attached Files:

  2. jcsd
  3. Oct 28, 2012 #2
    The two windings are at right angles to each other. So in an ideal world, there would never be any signal on the oscilloscope.
    Also I have some question how you believe this circuit should work.
  4. Oct 28, 2012 #3
    Well, the primary winding doesn't induce a voltage onto the secondary directly. What happens is if you line the primary winding up with the field lines of a magnetic field that you want to measure, and then drive the primary winding with a sinusoid that will saturate it. Because the winding on one side is in the direction of the field that you're trying to measure and the other is 180 out, one core will stay in saturation longer and the other will come out of saturation earlier...and it's this difference in the cores that will produce a voltage in the secondary winding (at the second harmonic of the primary frequency).

    You are correct though, in a perfect world with no external field there would be nothing produced on the secondary.

    I'm just having trouble finding a circuit that can source the required amount of current, as well as be at the resonant frequency of the coil/cap on the secondary.

    I might not have done a perfect job explaining how it works, but it's called a fluxgate magnetometer if you're curious about it.
  5. Oct 28, 2012 #4
  6. Oct 28, 2012 #5
    I'd looked at the toroidal form of the fluxgate, I was just hoping to get away from winding a toroid by hand (I have to do 3 of these for a 3 axis magnetometer). I also have the FGM-3 that the article referred to, which it's pretty fun to play with as a pre-made piece of equipment... I was just interested in making my own. Thanks for the article though, it was well written.

    The image below is the circuit as it is right now with a square wave driving the primary winding. Signal A is the voltage across the primary winding, and Signal B is the voltage across the secondary (which I had my leads reversed by accident, so it's upside down). When I plugged the primary winding in, the frequency jumped quite a bit...but the voltage as plummeted, so I'm trying to figure out a way to counteract that. I also discovered I need to do some filtering/demodulating...I wasn't expecting all the harmonics in the secondary.

    But anyways, thanks again for the reply...you wouldn't happen to have any ideas about my initial question about alternative circuits for producing a square wave would you?

    Attached Files:

  7. Oct 29, 2012 #6
    Some things that you can do to locate the information that you require are:
    Google operational amplifier astable multivibrator
    Go to Mouser, search for LF355, open data sheet, Figure 23 is an astable multivibrator.
    This is a low power operational amplifier and you may have to add a power amplifier after it.
    Go to Mouser, search for LM7171, open data sheet, first figure in the application notes is an astable multivibrator.
    This operational amplifier is high frequency and the circuit layout is critical.
    Both these circuits may have the time the output is positive greater than the time it is negative. Or it may be less time. If this is a problem, there are ways to make the positive and negative output time the same.

    If you are interested in multivibrators, the book IC Op-Amp Cookbook by Walter G. Jung has some useful information.

    Good Luck
  8. Oct 29, 2012 #7
    Thank you sir, I appreciate the help! I'll take a look at those chips when I get off work tonight.
  9. Nov 2, 2012 #8
    A Schmitt trigger makes a good oscillator, easy to tune through a single resistor.
    It will need some amplification to deliver 200mA, but you could use several logic buffers in parallel.

    I too have big doubts about the primary winding of your first picture. I expected it to make a field parallel to the sensed field in one core, antiparallel in the other, with the secondary summing both cores.

    If the sensed field is huge, you could live with square waves filtered by an LC... But if you sense for instance Earth's field, you will need much cleaner sine waves. This can be made by active filters (but then, you need a very linear amplifier to deliver the 200mA) or by several stages of LC filters.

    Depending on how clean the sine must be:
    - The filter coils must have no magnetic core (very probable)
    - The capacitors must be of polypropylene or type-I ceramic (sensitive fluxgate)
    - The resistors must be metallic (normally they're all presently)

    Waveforms are always dirty on fluxgates. They reflect the desaturation of a magnetic core, which produces pulses. This waveform contains the 2*F harmonic and many more; fluxgate circuits use this harmonic rather than the pulses because analog filters can be made with a huge selectivity, first to produce 1*F without 2*F, then to pick 2*F without 1*F.

    The effect of the sensed field is to shift the pulses in time, one forward and the other backward, and this produces even harmonics. If it weren't for sensitivity, we could measure the time position of these pulses to know the external field.

    Quantitative measurements add one DC field more that the circuit adjusts to zero the second harmonic; the current making this DC field is then the measure.
  10. Nov 2, 2012 #9


    User Avatar

    Staff: Mentor

    It seems you aren't needing a "negative voltage", you just need a square-wave whose current alternates in direction.

    If you take a standard 555 astable and feed its output to another 555 having pins 2 & 6 joined together, the second chip acts as an inverter producing a square-wave in antiphase. Connect your transformer winding between their two pin 6 outputs to get the desired push-pull drive. Both 555s operates off the same single-ended power supply.

    I think your application isn't critical in demanding that the drive level be exactly symmetrical, as here you are contending with variation between IC's for this. (I think asymmetry could be fixed by operating the inverting 555 from a separate power supply and tweaking its voltage very slightly, but probably totally unnecessary.)

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook