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

Gain Spikes

  1. Jun 18, 2010 #1
    Gentlemen and ladies,

    I have an RF device and an in-line amplifier (specifically an AMMC-6220 - http://www.datasheetcatalog.com/datasheets_pdf/A/M/M/C/AMMC-6220.shtml). Certain units develop gain spikes at the high-freq band. They are in very specific frequency locations, and vary in amplitude from unit to unit.

    Lowering the voltage to the amp eradicates the spikes. Raising it has the opposite effect. The bias voltage is well within spec.

    Another thing: playing with the input match to the amplifier has an effect on the spikes. However, the AMMC-6220 is (supposed to be) an unconditionally stable amplifier. This shouldn’t happen, assuming that what I’m seeing is oscillation.

    I’m stumped, and can’t do much except keep replacing this relatively expensive amplifier. The manufacturer doesn’t know what’s going on, nor do engineers a lot smarter than I. I can’t even seem to isolate the problem to a bad lot of amps. Any advice, theories, ideas, words of emotional support, would be appreciated.

    Attached Files:

    Last edited: Jun 18, 2010
  2. jcsd
  3. Jun 18, 2010 #2


    User Avatar

    Staff: Mentor

    Could you post a link to that AMMC-62220 amplifier? I'm having trouble finding it with Google. What is the RF device that feeds the amplifier? Where does the output of the amp go? If you go straight into the input of the spectrum analyzer with coax and no other feeds to other devices, do you still get the same results?
  4. Jun 18, 2010 #3
    Thanks for the reply and I apologize for the typo. It's an AMMC-6220 and the datasheet can be found here http://www.datasheetcatalog.com/datasheets_pdf/A/M/M/C/AMMC-6220.shtml

    The problem is the amp is part of a series of cascading amps, so I would have to tap directly onto the trace after this amp. There are a few attenuators, an equalizer, some diodes, some other amps and various filters along the way. I will have to give more details on Monday.
  5. Jun 20, 2010 #4
    Any way you can get an S-parameter sweep?
  6. Jun 20, 2010 #5
    attached to original post
  7. Jun 20, 2010 #6
    The amp may be unconditionally stable but your total system may not be.

    I can't see the frequency range on your plot, but I'd hook up a DC to 30ghz spectrum analyzer and look for an oscillation that depends on the bias voltage.

    The second spike on your chart looks to me like what happens when the network analyzer sweeps past a resonance.

    You need to set up your analyzer to do a gain-phase plot and start looking for ways that the output is leaking back to the input or bias supply.
  8. Jun 20, 2010 #7
    My bad, didn't see that. Is this a sweep of the rf device and the amp? If that's the case, is your rf device stable? Also, how are the other S-parameters impacted?

    I'm going to agree with Antiphon, and I definitely would check on the isolation of your power supply/amps/inputs.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook