• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 12 subscribers
  • Views 238 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.

AFE7950: Intrinsic Noise

Tom Hastings
Prodigy 10 points
Part Number: AFE7950
Other Parts Discussed in Thread: LMX2820

Tool/software:

Is the phase noise plot on page 87 of the data sheet "AFE7950_datasheet_revC_12jun2023.pdf" truly the phase noise due to the part intrinsic noise in external clock mode with the external clock phase noise backed out?  Basically with a perfect clock, -100dBc/Hz at 1KHz is the best I can do?

  • 0
    TI__Expert 4190 points

    Hi Tom,

    The Figure 7-280 in the datasheet does have the input clock phase noise removed. Is there a specific requirement/level you need for the phase noise?

    Best,

    Camilo

  • 0 in reply to Camilo Garcia Trujillo
    Prodigy 10 points

    Thanks for the quick response Camilo.

    This is for a pulse-doppler system.  Attached is a first cut graph of spec vs. calculation with no contributions considered for vibration, temp, or other circuitry. 

    There is also an assumption of coherance between tx and rx that I'm not completely comfortable with having just started turning on the evaluation module.  You can see that the math shows we are just meeting the spec and any margin will likely be lost under environmental conditions and complete system stackup.

    I wanted to make sure that my assumtions on coherance are correct and that there are no added phase noise contributions elevating the phase noise indicated on the plot.

    I can of course choose a lower phase noise reference oscillator and improve the performance but we're rapidly approaching a cost vs architecture threshold.

    Regards,

    Tom

  • 0 in reply to Tom Hastings
    TI__Expert 4190 points

    Hi Tom, 

    Sorry for the delay on this. I will check with our systems team to see if they have any extra recommendations.

    But like you said, to improve the performance you could move to a different clock source.

    Best,

    Camilo

  • 0 in reply to Camilo Garcia Trujillo
    TI__Expert 4190 points

    Hi Tom,

    We are planning to take some data to check the correlation between TX and RX. I will keep you updated on the results.

    Best,

    Camilo

  • 0 in reply to Camilo Garcia Trujillo
    TI__Expert 4190 points

     Hi Tom,

    We took a measurement in the lab to verify, and as you can see in the plot below we are able to see significant cancellation of the input clock phase noise when feeding the TX output to the RX input.

    We used the AFE7950 with 12GSPS DAC, 3GSPS ADC, and an external 12GHz clock from an LMX2820 .

    The traces in the plot below are the following:

    Yellow: Normalized single sideband phase noise of the clock, in this case we used an LMX2820

    Red: Single sideband phase noise when giving the output from an SGMA SGS100A to the RX input. As you can see this follows the clock’s phase noise closely.

    Blue: Single sideband phase noise when giving the output from a Tx channel to one of the RX inputs. As you can see there is >20dB of cancellation with no sign of being limited by the cancellation.

    Hopefully this can clear some of your doubts.

    Best,

    Camilo