• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 32 subscribers
  • Views 231 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.

IWRL1432: I got ADC data , then use HWA to do a FFT, but have 3 weird spikes

Allen He
Prodigy 120 points
Part Number: IWRL1432

Tool/software:

Dear experts,    I encountered such a problem.

I   got adc data ,like blow  (  total 1024 points , I export adc data , and draw it)



then do fft with log  use HWA   ,get resluts below

There are alaways thress weird spikes.   I am sure  they are not real  reflected signal.   

（up curve is the result of doing it once. below curve is superimposed multiple times and some denoising and smoothing are done.）

This program is modified from the following program.   use mmave sdk 05_03_00_02  

The main parameters of chirp can be roughly expressed as follows using cli.

channelCfg 1 1 0

chirpComnCfg 8 5 0 1024 1 110 0
chirpTimingCfg 6 25 0 26.04 77.42
frameCfg 1 0 120 1 250 1
factoryCalibCfg 0 0 34 0 0x1ff000

What is the problem? Can you help me figure out how to solve it?

If you need any more information, feel free to ask me

  • 0
    TI__Genius 14910 points

    Hello.

    What environment are you running the data collection in?  Do you only have the one object in front of the radar sensor or are you in an environment with many reflective objects?  If you find that these are regular occurrences in even a clean environment, I would also recommend switching to a production level device as I noticed you are using the SDK associated with the pre-production devices.  You can also try filtering out these points using the rangeSelCfg command if you see these small spikes past a certain distance and any further distance is not required for you.

    Sincerely,

    Santosh

  • 0 in reply to Santosh Krishnan
    Prodigy 120 points

    Hi Santosh.

    I am sure these are regular occurrences in even a clean environment.
    I test  lastest sdk MMWAVE_L_SDK_05_04_00_01,  get the same result.
    rangeSelCfg does not meet the requirements either, because I want to use the full range.


    Based on your experience, which part might be the problem?
    Configuration problem? Chip bug? AD? Mixer? Antenna?

    I have no idea where to start investigating this issue

  • 0 in reply to Allen He
    TI__Genius 14910 points

    Hello.

    The latest SDK for the IWRL1432 is SDK 5.5, which is for production-ready devices, while SDK 5.4 is not.  I would advise to ensure that there are no reflective objects in those ranges as those spikes would indicate some sort of detection at that range bin.  Could you also clarify what ranges you are seeing them in and how often you are seeing them at if they are occurring at regular intervals?

    Sincerely,

    Santosh

  • 0 in reply to Santosh Krishnan
    TI__Genius 14910 points

    Also, the level sensing is doing a zoom-FFT on the adc data as opposed to a regular FFT data.  Can you please do that as well in your analysis to see if this issue persists?

    Sincerely,

    Santosh

  • 0 in reply to Santosh Krishnan
    Prodigy 120 points

    Hi Santosh.

    I only have the ES1.0 device now, but I will get the ES2.0 device next week, so I will try it again and see the results.

    I think I am basically sure that there is no reflective object at that distance.

    Based on the parameters I set
    chirpComnCfg 8 5 0 1024 1 110 0
    chirpTimingCfg 6 25 0 26.04 77.42
    The full range is 36M, and there are three strange peaks, corresponding to distances of about 18m, 24m, and 30m

    They have always been there.

    ____________

    I observed a phenomenon.

    When I move the target, for example, move closer (the target with the red arrow).

    I found that the three peaks also seem to move and branch out.

    I don't know this phenomenon. Can you help me analyze it?



  • 0 in reply to Santosh Krishnan
    Prodigy 120 points

    zoom-fft is to zoom in around a certain point, which I think is not relevant to this question.

  • 0 in reply to Allen He
    Prodigy 120 points

    Could it be that the problem is with the IF mixing part of IWRL1432, such as a large reflected echo causing errors in the mixing circuit?

  • 0 in reply to Allen He
    TI__Genius 9940 points

    Hi

    I do not suppose that there is any problem IF mixing part. But it could be due to multi path from the reflected object if these peaks move as you move the target. It would be an expected behavior due to multipath if the peaks move as you move the target

    Regards