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

Inverting a 130 Vrms, 40-45 kHz Signal?

  1. Jan 24, 2008 #1
    Inverting a 130 Vrms, 40-45 kHz Signal??

    Hello everyone, I was wondering if anyone might be willing to help me out on a little problem I am having:

    I have an amplifier output of 130*sin(f*t), f = [40 - 45 kHz], and I need to run this into two parallel lines such that initially, line 1 carries 130*sin(f*t) and line 2 carries 130*sin(f*t). Next, I need to flip the sign of line 2 such that the signal becomes -130*sin(f*t).

    So my question: Can anyone describe a circuit that can take the high voltage and frequency of the amp. output and invert the waveform?

    I was looking into using a high voltage op-amp from MSkennedy.com, but it seems they have op-amps that can take enormous supply voltages, but whos input voltages are limited to ~25 V. I am a complete moron when it comes to electronics, so I'm lost here.

    Another option would be changing the phase of signal, but I would have to do so in a fashion that is frequency independent, as part of my testing involves altering the frequency of the signal to observe changes in my actuator performance.

    Thank you for taking the time to read this, and thank you for any comments you might have in advance.

    Best,

    Devon
     
  2. jcsd
  3. Jan 24, 2008 #2

    berkeman

    User Avatar

    Staff: Mentor

    Just use a transformer. You will need one that runs at that frequency, and has adequate safety precautions.
     
  4. Jan 24, 2008 #3
    I took a look at inverting transformers, but I am not really seeing sine--> (-sin) inverters. I see that I can invert DC signals to AC signals, but that's about it.

    If you have a moment, would you mide ellaborating a little bit about how I might use the transformer to alter the wave sign?

    Thank you,

    Devon
     
  5. Jan 24, 2008 #4

    Integral

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    There is a big difference between a transformer and a wall wort. If you do not know the difference I would question the wisdom of attempting this project.

    If you want us to help, you need to explain in detail what it is you need to accomplish. It may well be that what you need to do is much different from what you are trying to do.
     
  6. Jan 24, 2008 #5
    Alright, here is the setup:

    First off, the reason for the request: I have a Traveling Wave Ultrasonic Motor that I am trying to run tests on. The motor is actuated by 4 sets of piezoelectrics attached to two separate stators within the motor package. The first set of PZTs on stator 1 requires two signals, a sine and a cosine wave, 130sin(~40kHz*t) and 130cos(~40kHz*t). The second stator has the same size PZTs, but they require two different signals, -130sin(~40kHz*t) and 130cos(~40kHz).

    In the lab, I have two high power amplifiers that are capable of outputting up to 400W WAY more than I need. But, they serve the purpose. The first, (a.), recieves a 1V, ~40kHz cosine wave from dSpace and amplifies it to around 130V. The second, (b.), recieves a 1V, ~40kHz sine wave from dSpace and amplifies it to around 130V.

    Now the problem:

    The motor requires four signals.

    (1.) cos
    (2.) sin
    (3.) cos
    (4.) -sin

    I can use the amplifier (a.) to supply signals 1 and three with no problem.

    I cannot for the life of me figure out how to use amplifier (b.) to provide signals (2.) and (4.), as I do not know how to change the polarity of the signal.

    I think my first post was unclear... I must have given the impression that I wanted to create a high frequency AC signal from a DC source, "invert" the signal. I'm sorry, I used the wrong terminology. What I would like to do is change the SIGN of the sin() wave produced by amplifier (b.) from POSITIVE to NEGATIVE. The only other stipulation is that whatever device is used, it cannot alternate the frequency or amplitude of the signals, as all four signals being passed to the motor must be of the same amplitude and frequency.

    If that isn't enough info, I'll be more than happy to include more. I'm sorry about the initial confusion.

    Best,

    Devon
     
  7. Jan 24, 2008 #6

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    If amplifier (b) accepts a differential input (i.e. floating ground), all you'd have to do is reverse the input cables. If your signals are single-ended, you could use any run of the mill op-amp to do the "inversion" on the input to amplifier (b).

    - Warren
     
  8. Jan 24, 2008 #7

    berkeman

    User Avatar

    Staff: Mentor

    Couple things. I was referring to a signal transformer, 1:1 in turns ratio. You put your signal in one side, and then choose which polarity you want to use from the output (the output is just two wires, so use them one way to get an output in phase with the input, or the other way to get an output 180 degrees out of phase). You can also wind the transformer with two output windings, so that you can use one output in-phase with the input, and the other output out of phase with the input. This is a cleaner solution than using the input and the output of the transformer as your two different signals (depending on what the load impedance does).

    Second, you are mentioning that you really need all 4 signals, and the cos signals are in quadrature with the sin signals (90 degrees out of phase). That would seem to be your harder problem, not just inverting the signals (180 degrees is easy with the transformer as I already described).

    For the setup that you describe, I would recommend generating the sin and cos signals with signal generators that provide a phase-lock capability. That is pretty common on laboratory-grade signal generators. Then generate the opposite polarity signals with small signal transformers that you wind yourself (or maybe can find at Digikey or somewhere -- be sure that they accommodate your frequency range), and then finally, use 4 power amplifiers to make the high-voltage signals.
     
  9. Jan 24, 2008 #8
    Right, but if I understand you correctly, I would be inverting the sine wave PRIOR to amplification, I would then wind up with this:

    (1.) cos
    (2.) -sin

    out of the two amplifiers.

    In which case, I am left with the same problem as before, except that I would be needing to change ONE of those sine outputs into a positive (+) sine wave while leaving the second negative (-).

    I need to split the cosine signal into two and the sine signal into two, and THEN I need to change the SIGN on the sine signal.

    So it looks like this:

    (1.) cos: splits into two separate branches, a. and b.
    a. cos
    b. cos

    (2.) sin: splits into two separate branches c. and d.
    c. sin
    d. sin --> NOW, once in a separate branch, I want to manipulate this signal such that the SIGN changes from a positive (+) sine wave to a negative (-) sine wave.

    Thus, after the processing I should have four signals.

    (+) cos
    (+) sin
    (+) cos
    (-) sin


    Thank you for your suggestion, and any further advice would be great.

    Best,

    Devon
     
  10. Jan 24, 2008 #9

    berkeman

    User Avatar

    Staff: Mentor

    What I said was use two signal generators, locked at 90 degrees of phase shift. That will give you your small signal sin and cos waveforms. Then put those each through 2 separate transformers. Each transformer should be wound with three windings, with a 1:1:1 turns ratio. The transformer needs to be a good transformer (low loss) at the signal frequencies your are mentioning. The 4 outputs of the 2 transformers give you the sin, -sin, cos and -cos small signal waveforms. Amplify those 4 signals to get your power signals.
     
  11. Jan 24, 2008 #10
    berkeman,

    I have no problem generating the pre-amplified signals, as they are being sent via dSpace. The problem is that I only have 2 amplifiers capable of producing a 130V, 40-45kHz signal. The amps are ridiculously expensive, and we can't afford 2 more.

    Can a transformer handle these high frequencies? If so, could I possible get one such that I can take one line off the sine output of one amplifier and run it to the motor, and then I can take another line and run it to the transformer with a 1:1 wrapping, and then switch the wiring on the transformer output to flip the sign of the output into the motor?

    Am I correct in assuming that I can branch the cosine signal into two wires and input them into different places on the motor as well as branching the sine output from the amplifier, modifying one of those branches signals, and then plugging each of those into the motor?

    Thank you for your time and patience.

    Devon
     
  12. Jan 24, 2008 #11

    berkeman

    User Avatar

    Staff: Mentor

    You can put the transformers on the 2 power amp outputs -- that just means that the transformers have to be bigger, and you need to think a bit about safety when dealing with those high voltages.

    So use the two sig gens to make the small-signal sin and cos signals, go through the power amps to make the high power versions, and then go through the two transformers that I described in my last post to make the high power sin, -sin, cos, and -cos signals. Take all 4 power outputs from the transformers.

    Yes, you can wind up transformers to handle the power and frequency requirements. It's done all the time. It would be best for you to find some technician or engineer there where you are doing this -- find somebody who has wound up audio-type or data-signal transformers in the past, and ask them to help you figure out how to do it. Is this work at a university or something?
     
  13. Jan 24, 2008 #12

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Ah, I see, you need to flip the sign AFTER amplification.

    Well, are the inputs to the motor differential or single-ended? If they're differential, can you just reverse the wires going to one of the sin inputs?

    - Warren
     
  14. Jan 24, 2008 #13
    -Ok, Sounds good.

    --Alright, here. I only need to flip the sign on the sine, not the cosine. So it is alright if I just use two parallel lines for the cosine run directly off the amplifier? Next, if I only want to change the sign on SINE, couldnt I run the output from the power amp into two parallel branches and use the only one transformer (of high quality to retain the signal quality), and flip the outputs of the transformer to achieve a negative (-) sine wave?

    With your description above, it sounded to me like I would have needed four transformers instead of 2, one for each of the two branches coming off amp 1 to create cos and -cos, and two more transformers for the two branches coming off amp 2 in order to create sin and -sin signals. Is this correct?

    --You got me. Yeah, I am working on my Masters at Virginia Tech, as a dynamics/smart-materials guy (I am a DOLT when it comes to electricity...) The work involves the Traveling Wave Ultrasonic Motors, and running them without their company provided drivers is a complete pain.

    I'll certainly not try to do the winding myself. Thank you for everything; especially your patience.

    Devon
     
    Last edited by a moderator: Jan 24, 2008
  15. Jan 24, 2008 #14

    berkeman

    User Avatar

    Staff: Mentor

    You could use separate 1:1 transformers, or you could use 1:1:1 3-winding transformers. It just depends on whether you want to buy or wind them, and what kind of performance you need.

    I suggested using two 1:1:1 transformers and using the outputs (rather than two 1:1 transformers where you used both the input and inverted output as your two power drive signals), because you will get a little more consistent performance when both loaded signals are from the outputs of the transformer, instead of one from the input and one from the output. The loss through the transformer and the leakage inductance of the transformer are the main things that I had in mind. But for a simple lab experiment, that may not matter.
     
  16. Jan 25, 2008 #15

    f95toli

    User Avatar
    Science Advisor
    Gold Member

    As Chroot has already pointed out is that you should be able to "invert" one sine by just swapping the two wires at the motor.
    You would need two pairs of wires going from each amplifier, by swapping the cables "+" and "-" in one of the pairs coming from the "sine" motor you get -sin.
    This would only work if input is differential (no ground wire) but unless there is somehing I am missing it should be.
     
  17. Jan 26, 2008 #16
    F95toli and Chroot:

    We actually tried this initially, but it didn't work, as the SIN and COS signals share a ground inside the motor. Huge bummer... There is no way to add an additional ground without altering the motor and killing out warrantee.

    Thank you for the suggestions though.
     
  18. Jan 26, 2008 #17
    I'll be taking a look into the transformers. I think it would actually be pretty cool to wrap one up and see it work, whether to our desired efficiency or not. We'll see. I'll post again once we get this thing going.

    I appreciate everyones help, sincerely.

    Devon
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Inverting a 130 Vrms, 40-45 kHz Signal?
  1. Vdc and vrms (Replies: 3)

Loading...