• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 30 subscribers
  • Views 538 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

AWRL6432: The relationship between DC leak level and noise increase

TAK
Prodigy 150 points
Part Number: AWRL6432
Other Parts Discussed in Thread: AWR6843

When we evaluate our AWR6432 module, sometime the noise level is drastically rising up. And it depands on the module design.
Seeing range profile of each antenna, the module waht the DC leak level is higher would be that the noise level is also higher.
The DC leak level of TI EVM is lower and the there is also no noise increasing.
We also tried to control the Rx gain, HPF norner and Tx Backoff. If we change the each value to suppress the DC leak level (that means Rx Gain lower, HPF corner is higher, Tx Back-off is larger), the noise increase is disappered.

And this phenomenon is not observed about AWR6843 module evaluation.

So then, should we supress the DC leak enough when we use AWR6432 module?
And if you know would you tell something about the relationship between DC leak and noise increase?

This is the range profile result of our module and TI EVM. (FW is our original based on Motion and Presence from SDK5.3)
This evaluation was conducted in chamber box. We are thinking that the peak around 50cm would be the reflection from inner wall of chamber.

・TI EVM


・Our module

Thank you,
Best regards

Tak

  • 0
    TI__Mastermind 40495 points

    HI, Tak:

    I would recommend you check whether there is ADC clipping with your module.  The noise level could be due to the clipping.  

    The DC signal is what we call TX coupling.  When the DC signal is much higher, the phase noise can increase too.  So you can potentially see the noise increase.  To actually observe the noise level, you should plot the range-Doppler FFT output for your static test.  The zero Doppler represent the signal, and non-zero Doppler usually represent noise floor.   The noise you observe could be just the FFT spreading effect or ADC clipping, they are not really noise floor we usually talking about. 

    You can refer to my measurement below with TI EVM.   After range-Doppler FFT, you can see the zero-doppler has show high peaks in close-DC (represent the TX coupling) and around range bin 60 (represent the corner reflector).  And the other non-zero Doppler bin represent the noise level.   When you evaluate your results, you need to first make sure there is no ADC clipping (check the ADC data), and second, apply hanning window to avoid strong range FFT windowing effect. 

    Best,

    Zigang

  • 0 in reply to zigang Yang
    Prodigy 150 points

    Thank you for your reply, I understood well.
    We would try to get and check the ADC data.
    Does AWR6432 tend to get the phenomenon more easily than AWR6843?

  • 0 in reply to TAK
    TI__Mastermind 40495 points

    HI, TAK:

    The default of AWR6432 signal chain uses BPM which has 6dB higher maximum power.  But AWR6843 usually by default uses TDM mode.   So it is possible that BPM mode can have a higher chance to see ADC clipping.

    Best,

    Zigang

  • 0 in reply to zigang Yang
    Prodigy 150 points

    This phenomenon is also observed using TDM mode of AWR6432 module.
    AWR6843 output is higher of 1dB than AWR6432. So I am thinking that normally AWR6843 is easier to have ADC clipping.

    Does clipping of both IC occur at the same level?

  • 0 in reply to TAK
    TI__Mastermind 40495 points

    HI, there:

    We have not noticed any ADC clipping on TI EVM for normal configurations on both devices EVMs.  

    Can you send me your configuration file to review? 

    Best,

    Zigang

  • 0 in reply to zigang Yang
    Prodigy 150 points

    This is the .cfg file used in our evaluation.

    Ghost_Evaluation_using_TDM.cfg

    BR,
    Tak

  • +1 in reply to TAK
    TI__Mastermind 40495 points

    HI, Tak:

    You are just using one TX, and your RX gain setting is 30dB.  I will not expecting any ADC clipping with these settings. 

    The only thing is that your HPF setting is really low.  In xwrlx432, we only have one stage of HPF, and in xwr6843, we have two stages of HPF.  So, I would recommend to set HPF to be at least 1. 

    chirpComnCfg 18 0 0 128 1 30 0

    Best,

    Zigang

  • 0 in reply to zigang Yang
    Prodigy 150 points

    Thank you very much for your answer and I'm sorry for confuse.
    We are changing the Rx Gain, Tx Back-off, Filter cut-off in the evaluation.
    I understood that it is difference of number of filter. We try to adjust the setting of HPF.

    Your explanation really helped, I appriciate.

    Best regards,
    Tak