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

Impedance matching

  1. Apr 21, 2008 #1
    A friend of mine is having problems with a measurement signal he is getting from a Rogowski coil. When he changes the coaxial cable length the amplitude of the signal changes. It’s been a while since I’ve studied this type of thing but to me it sounded like an impedance mismatch issue and thought that I would suggest using a balun. Does this sound like a sensible solution?

    The Rogowski is a clever design but quite simply a single loop of sold aluminium with a very low inductance, around 1 x 10^-10H. The signal is going straight in to an integrator circuit, and from there in to a scope.

    I’d really appreciate any advice on the subject so I can help my friend out.

  2. jcsd
  3. Apr 21, 2008 #2


    User Avatar

    Staff: Mentor

    Sounds like the first preamp needs to be nearer to the coil. The output Z of the coil is nowhere near the Zo of coax. What frequencies is the coil being used for, and how long is the coax?
  4. Apr 21, 2008 #3
    It's to measure a current pulse which is around 70ns, so GHz. I assume this will mean the integrator will need to be effectively directly coupled to the coil?!

    I'm sure the coax at the momnent will be in the region of a few meters in length.

    Since I've had a quick look in to baluns and I can see that wouldn't be suitable due to the impedance of the Rogowski (next to nothing).

    Can the length of the coax effectively be tuned?

    Thanks for your quick response BTW :smile:
  5. Apr 21, 2008 #4


    User Avatar

    Staff: Mentor

    At those frequencies, yes, transmission line considerations and antenna considerations definitely come into play. And the inductance and size of the coil could become an issue.

    The first improvement would be to put a preamp right at the coil output, and then drive any coax that is needed to get to the main integrator and other electronics. The preamp would need to be able to drive the 50 Ohm coax (or 100 Ohm twisted pair, or whatever) directly, and that transmission line would be terminated in its characteristic impedance at the integrator end of the line.

    Since you are sensing a pulse, you cannot tune the length of the coax. The pulse has too broad a spectra for that.

    You need to look at the size of the coil and its inductance and parasitic capacitance, to see if any of that is going to be an issue for the GHz frequency range you want to operate in. Also, your 70ns pulse does not necessarily have frequency components up into the GHz range, depending on the characteristics of the pulse. In general, you will want to be sure that the size of your sensing coil is much less than a wavelength at the highest frequency component, and be sure that the coil inductance and parasitic capacitance (including the input capacitance of the preamp) do not resonate until well above the frequencies that you are concerned with.
  6. Apr 21, 2008 #5
    Thanks for both of your replies, I'll look in to your suggestions.

    It's funny how something that someone mentions in passing can consume your thoughts and distract you from what you're meant to be doing!
  7. Apr 21, 2008 #6
    I'm sure there was another reply suggesting some transformer that I should look in to but it seems to have gone!
  8. Apr 21, 2008 #7


    User Avatar

    Staff: Mentor

    No worries. There was another reply, but the poster realized that his suggestion was for narrow-band impedance matching to the coax, and that your question was more broadband, so his trick would not work, so he deleted his post.

    That happens to me sometimes as well, where I realize a bit later that what I've posted has a limitation. I usually go back and add an "EDIT -- ....." line to the end of my post, so that both the suggestion and the associated limitation are still posted. But deleting the post is an option too, at least for 30 minutes or whatever the current PF timeout is.
  9. Apr 21, 2008 #8
    I like berkemans answer.....

    Would it also do well considering that the integrator is somewhat acting like a charge amplifier for him to stick that stage at the front end also and then transmit the signal as a voltage signal using a fet follower?
    I find that charge transmission is unduly worried even by cable movements.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Impedance matching
  1. Impedance matching! (Replies: 3)