TRF372017: LO Isolation & Phase Noise

Part Number: TRF372017

Hello,

I am using the TRF372017 in fractional mode with a carrier around 2GHz. I have the loop filter set near the 30.72MHz fPFD example within table 30 in the datasheet. With only a carrier on the RF output the phase noise on the LO port is measuring ~1degree in my band of interest. When I input pseudo random data onto the BB input the RF spectrum appears as expected on the RF output. The modulated output also appears on the LO output, looking like an RF output with horrible carrier suppression. How much isolation should I expect between the LO output and the RF output? Are there possibly settings within the device that I could have not tuned properly that would cause this?

Thanks,

-Brandon

  • Phase noise is not measured in degrees. Any positive number would be very, very bad. 8^) What are you actually trying to measure?
  • In reply to Keith Barkley:

    Phase noise when integrated across a specific band can be measured in degrees rms. This is commonly how it is specified to us. The issue that I am seeing is that when data is being input into the baseband inputs, modulation shows up on the LO output and the RF output. This modulation on the LO output is causing my rms phase noise measurements to be fictitiously high. When 0s are being input to the BB input, the LO output and RF output are simply the carrier as expected. how many dB of isolation should there be between the LO output and the RF output. I expect that the LO output would not be affected by the baseband inputs.

    Thanks,

    -Brandon
  • Hi Brandon,

    I will try to reproduce your issue in the lab and will get back to you as soon as I have results.

    Regards,

    Abdallah

  • Hi Brandon,

    It appears that there exists coupling between the baseband input (that gets modulated up) and the LO output port. There isn't anything that you could do via the registers to correct for this, as far as I am aware. I measured the isolation on the LO port over various frequency offsets and have attached them to this post. If this is an issue for your application, I believe that the TRF3722 may be a better fit. Its overall performance is improved over the TRF372017, and I believe that it will have better isolation as well.

    Regards,

    Abdallah

  • In reply to Abdallah Obidat:

    Hey Abdallah,

    Thanks for the info. I don't think this is the issue that I am seeing though. It is the modulated output that is showing up on the LO. What is the isolation between the RF output and the LO output? Could a bad output match and high VSWR somehow make it's way back onto the LO output? This seems to be the case as we are able to tune the output of the device and see the power level of the modulated output on the LO output change. Which in turn increases or decreases the phase noise depending on how good the tune is.

    Thanks again for the support.

    -Brandon
  • In reply to Brandon Sorn:

    Brandon,

    The data above was measured on the LO output port and compared the power of the LO against the power of the modulated RF output. This issue of coupling between the baseband input and LO, manifests itself as a diminished RF output on the LO port. Tuning the output impedance will probably affect the isolation to a degree. Also, the isolation does improve when the output is loaded with a 50 ohm termination. I do not believe, however, that there is a way to completely eliminate the RF output from the LO port without attenuation. You could try attenuating your input signal and you’ll notice an improvement in isolation.

    Regards,

    Abdallah