• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 50 subscribers
  • Views 494 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

High frequency difference/Instrumentation amplifier?

Javcel
Intellectual 470 points
Other Parts Discussed in Thread: VCA824, THS3217

Hi,

I have two sinusoidal inputs at frequencies from 15MHz to 50 MHz with an amplitude of approx 1.4Vpp. I need to accurately measure the difference in amplitude between them (approx 0.1% to 1%) but I am having issues to find a device which fulfill the requirements.

I guess the main issue is the CMRR at these frequencies, so an off-the-shelf INA or Difference amplifier would not be good enough for this particular application...

Any ideas on how to implement this?

Thanks a million,

Kind regards,

  • 0
    Guru 140200 points
    Hi Javcel,

    I would use a sharp passive LRC low pass and high pass filter to separate them. We use 5th order elliptic filter for a similar task.

    Kai
  • 0
    TI__Genius 17666 points
    Hi Javacel,

    Can you use a discrete INA? Two closed loop buffers and a differential configuration op amp?

    Best regards,

    Sean
  • 0 in reply to Sean Cashin
    Guru 48395 points
    Hi Javacel, this used to come up in the earlier Comlinear days a lot - so I wrote the attached - which should be what you want.

    www.tij.co.jp/.../snoa393b.pdf

    Use this with the most recent versions of this device, VCA824 etc - these can do very high CMRR to high Freq., and you don't need to use the gain adjust feature, just lock them in highest gain, or if you need to have a wide range on the difference in amplitudes, you can use the gain adjust to attenuate the output when the difference in amplitudes is large -
  • 0 in reply to Michael Steffes
    Guru 48395 points
    Hold on Javacel,

    While the VCA824 approach might be enough, I was ruminating this morning what you really want is a buffered single stage diff to single operation.

    Well, that is exactly what the front end of the THS3217 is - a recent part we did for DAC output interfaces. Figure 17 in the data sheet shows about 35dB CMRR at 50Mhz. that is probably small signal, so you would need to be careful about large signal mismatches and slew limiting out of that first D2S stage. If large signal effects limit you, might want to attenuate each input by some amount - matched of course.

    Also, whatever your sources are, watch mismatched phase shifts to the inputs - at 50Mhz, won't take much to generate false diff out.
  • 0 in reply to Michael Steffes
    Intellectual 470 points
    Hi Michael,

    Thanks very much for your reply. it is extremely useful information.

    I will most likely build a small prototype using both solutions (VC824 and THS3217) and compare performances.

    I will mark the thread as solved and will reopen it or create a new one if anything arises while testing the circuit.

    kind regards,
  • 0 in reply to Javcel
    Guru 48395 points

    You bet Javcel,

    This is actually a pretty difficult task here - I suspect you will have large signal related mismatches, and maybe you will need to experiment with lowering your amplitudes to get the difference, then gaining that back up - good luck.

    I was doing some sims with the VCA824 model, and small signal wise it seems to have some mismatch built in - the beginning of that work is attached.

    Using the VCA824 as a wideband differencing stage.docx

    Wideband differencing VCA824.TSC