• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 3 answers
  • Subscribers 31 subscribers
  • Views 1718 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.

IWR1443BOOST: Strange interferences in noise floor profile

Patrik Reiske
Prodigy 40 points
Part Number: IWR1443BOOST

Hello everyone,

I coded sort of a Python client API for the mmWave demo program (version 1.0.0.5) running on my IWR1443BOOST which takes the incoming data -- object list, range and noise floor profile -- and processes it as in the DemoVisualizer's source code (mmWave.js). Also it's capable to display the data (see below).

Running my software, I notices that there are strange interferences in the noise floor profile (see below; red line in bottom axes). Those occur in irregular periods, seem to be kind of random (haven't figured out how to trigger them, yet) and are not produced by a bug in my software I think, since the very same code is used for the range profile (see below; blue line the the bottom axes), which does not show this behavior.

Any thoughts on why this happens and what's the problem?

  • 0
    Prodigy 40 points
    I forgot to mention this in my initial post: I'm measuring indoors with the antennas pointing against a wall in about 2 meters distance.
  • 0
    TI__Guru 50240 points

    Hello Patrik,

    It is difficult to interpret what is going on in your range profile. Do you have any other pictures of it? Are there any other objects around the wall that could be in the Sensor's field of view?

    One way to improve noise floor is through the chirp to decrease the ADC sampling rate or increase the number of ADC samples taken. You can see more about altering chirp parameters here: www.ti.com/lit/an/swra553/swra553.pdf


    Cheers,
    Akash

  • 0 in reply to Akash Gondalia
    Prodigy 40 points
    Hello Akash,

    It seems like those interferences occur more often when a fast moving (no matter what direction) is in the sensor's field of view. Also, frequency of occurrence and density of the interferences seem so correlate with the range of the signal source -- less often/dense near the sensor.

    I probably won't find time to do some tests on altering the ADC sample settings this week, but I'll reply here as soon as I did.

    Thanks.
  • 0 in reply to Patrik Reiske
    TI__Guru 50240 points
    Patrik,

    Any update here?

    Moving objects should almost always be easier to detect than stationary ones. This is because moving objects will be found when performing Doppler FFT while stationary objects will not. This results in a higher degree of precision, so it doesn't make much sense that moving objects are adding a component of interference to your scene.

    Since your GUI is custom, it is difficult to understand why it is so different from the standard mmWave Visualizer. Can you provide some information on changes you have made? Also a picture of your scene would be helpful.

    Have you tried decreasing sampling rate or increasing number of samples? Let us know what questions you have at this time.


    Cheers,
    Akash
  • 0 in reply to Akash Gondalia
    Prodigy 40 points

    Hi Akash,

    please excuse my late response. Sadly, I did not have time to do some tests on the ADC sample rate, yet.

    I agree with what you wrote about moving objects. But since the interferences do also occur while there's no moving object in the sensor's field of view and as I said the source code for processing range and noise profile data is the very same, I still wonder how/why those are caused! Any calculations in my source code are -- apart from that it is Python instead of JavaScript, entailing different syntax and called functions -- exactly as in the DemoVisualizer's mmWave.js file.

    Not sure if this assumption is correct, but isn't the occurrence of the interferences even without moving objects indicating, that it's not an ADC sample rate related problem?

    See a picture and a drawing with distances of the scene below.

    Thanks!

  • 0 in reply to Patrik Reiske
    TI__Guru 50240 points
    Patrik,

    I don't have enough information yet to rule out an ADC problem. I have several questions for you which would probably be more appropriate if taken offline. I've sent you a connection request, please accept it and we will proceed from there.


    Cheers,
    Akash