• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 10 subscribers
  • Views 233 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.

AFE8030EVM: observing phase compensation after DAC

Shiva Pinagani
Prodigy 60 points
Part Number: AFE8030EVM
Other Parts Discussed in Thread: AFE8030
We are observing some rotation on phase compensated de-mapped data (5G NR OFDM waveform) when processing the IQ dumps, collected at TI 4.4 GHz output, offline in MATLAB. Our setup is:

IQ generated in RTL -------- JESD interface --------> TI DAC (4.4 GHz) -> VSA (4.4 GHz)
We also passed the dump collected at the input of the JESD interface i.e., the RTL generated Tx output data. There we didn't observe any rotation. Please refer to the snippets attached below. Please help us resolve this issue.
  • 0
    TI__Guru 69584 points

    Hello Shiva,

    The TX portion of the AFE80xx will be stable at a fixed frequency as long as the Fref (reference frequency) and also the AFE80xx on-chip PLL is stable and remains locked. The rotation of the QPSK happens when the TX output of the AFE80xx is rotating with respect to the observation path to the Matlab script.

    The Matlab and the observation path to the TX output is the source for the frequency offset compensation in the QPSK demodulation. If the initial frequency offset estimation is different than the final offset estimation at the end of the 5G NR frame, then the phase rotation will happen.

    Please advise if the AFE8030 EVM is phase locked to the observation path/instrument through a common reference. This is one key point at the test setup side. 

    In terms of the AFE80xx itself, you will need to check whether the reference frequency and the on-chip PLL is locked and stable. The AFE80xx itself, if function properly at the clocking level, is at a lower level than the frequency offset compensation done at the 5G NR level.

  • 0 in reply to Kang Hsia
    Prodigy 60 points

    Hi Kang ,

    We Actually Synced TI with VSA(observation path) by Taking the output 10Mhz REF from VSA to TI.

    Regards,

    P Shiva

  • 0 in reply to Shiva Pinagani
    TI__Guru 69584 points

    Hi Shiva,

    Please check your EVM% demodulation process by cross checking the result with a known 5G NR demodulator like the Keysight VSA or the R&S FSW 5G personality. When we use the two tools highlighted above, we do not see phase rotation issue. This will help you check your Matlab script on frequency/phase offset compensation.

    -Kang