• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 1757 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.

Original question:

IWR6843ISK-ODS: Understanding of theoretical "chirp time" vs timing parameters within CFG file

IWR6843ISK-ODS: Understanding of idle time and its effects

Victoria Rudakova
Intellectual 770 points
Part Number: IWR6843ISK-ODS

Hello,

I am trying to tune up my chirp config file for an application of human presence detection for short range operation (e.g. 5-10 meters). I was able to establish the values of the main components within my CFG file such as timing parameters (ADC sampling time, etc), slope, ADC sampling frequency and number of loops. My objective was to create a config that is sensitive enough for smaller movements. That is, I am trying to optimize for velocity resolution parameter. However, it is still not clear to me how (or if) the idle time affects the performance. 

I learned that idle time may have a certain minimum value - Table 3 of https://www.ti.com/lit/an/swra553a/swra553a.pdf . However, I am trying to understand what happens if we have a much larger value for the idle time.

So, for my CFG file, I tested different values of idle time while keeping all the other parameters the same: from 10us to 500us. When looking at the data, I noticed that larger values of idle time help to distinguish smaller movements better. I'm trying to figure out why do I observe such phenomenon.  

My questions would be:

1. What would be a general advice on choosing the idle time given application of presence detection. Is it better to go as large as possible or as small as possible, or somewhere on the middle? My only interests would be ability to distinguish smaller movements while having as high SNR as possible. 

2. Is my understanding correct that the increased idle time helps to improve sensitivity to smaller movements, or I misunderstand something? How does this play out on theoretical level? From what I know, the idle time defines a period of time between two consecutive chirps, so I just cannot wrap my head around the fact that we would see data differently when the idle time is increased between the chirps while the ADC sampling time and number of loops remains the same. If my assumption is correct, what is the explanation of it?

3. Are there any cases when the idle time is too short? For example, I learned that it gives some rest time to DSP processes to be done - are there more to be aware of?

  • 0
    TI__Mastermind 23725 points

    Hello Victoria,

    Thank you for your interest in TI's mmWave device.  While we will have detailed answers for you soon to address each of the questions, I will try and set the direction of answer.

    Time  between two chirps (Tc = Total Chirp time)   has inverse relationship with Max velocity and Velocity resolution.  Tc includes Idle time.

    So you are on the right lines to think that increasing idle time gives more sensitivity to movement ( lower the velocity resolution higher the motion sensitivity) .

    Hope you have already seen the Max Velocity and Velocity resolution sections in the document you referenced.

    Device level constraints for programming chirp engine for various parameters are incorporated in the Ti's mmWave estimator tool.

    If would further help to See the Interface control document which calls out valid values and constraints on parameters

    MMWAVE ICD

    Thank you,

    Vaibhav

  • 0 in reply to Vaibhav Mahimkar
    Intellectual 770 points

    Hi Vaibhav,

    Thank you for the reply. It would be great to get the clarifying answers. Meanwhile, I will be looking through the document you attached.

    Regards,

    Victoria

  • 0 in reply to Victoria Rudakova
    TI__Mastermind 40355 points

    Hi, Victoria:

    You can check the radar training to under more about the basics on range estimation, velocity estimation at:

    https://training.ti.com/intro-mmwave-sensing-fmcw-radars-module-3-velocity-estimation?context=1128486-1139153-1128544

    The maximum unambiguous velocity  for FMCW radar is given by:  = /c, where Tc is the chirp duration.  And the velocity resolution will be vmax/numChirp.  This means the longer the chirp (including idle time), and larger the number of chirps (numLoops) in one frame, you will have better velocity resolution, i.e., more sensitive to the small movement.

    In general, the larger the idle time and the larger the numLoops, the system will be more sensitive to the small movement.  However, it has to be compromised with other requirement too like frame rate and the on-chip memory limitation. 

    Best,

    Zigang

  • 0 in reply to zigang Yang
    Intellectual 770 points

    Hi Zagang,

    Thanks for the reply. I had already seen the training video and read through some of the other documents. I guess, my original question was not very clear. 

    I am aware that sensitivity (velocity resolution) depends on the chirp time, they are inversely proportional. What is not clear to me is why does idle time influence the velocity resolution as well - since there is nothing happening during the idle time, no ADC sampling or anything of the sort, the idle time does not make the signal emitting longer. So how come the idle time makes it part of the chirp time when it comes to velocity equations, shouldn't it just be the ADC sampling time only? At least I got this sense from the TI training video - it's only the time when the chirp is emitted that matters; the idle time was never covered there.

  • 0 in reply to Victoria Rudakova
    TI__Mastermind 40355 points

    Hi, 

    Yes, the range estimation will not be affected by changing idle time.  However, the velocity estimation is based on the phase change from one chirp to the next chirp.  Therefore, if you increase the idle time, the neighboring two chirps will separated longer in time domain.  The same movement in speed will cause bigger phase change from one chirp to the next chirp.  

    Best,
    Zigang

  • 0 in reply to zigang Yang
    Intellectual 770 points

    Thanks for explanations, now it makes sense!