• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 335 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.

IWR1642: Is initial chirp random?

Marvn Arcangel
Genius 12415 points
Part Number: IWR1642

Hi Team,

Our customer has a question about the initial phase of each chirp of IWR1642 kindly see below:

"Since I know the transmitter signal is x(t) = exp(j*2pi*f_c*t)*exp(j*pi*S*t*t)*exp(j*phi), in my understanding, the phase phi is random. In the receiver, after a time delay of tau, the signal is x(t-tau) = exp(j*2pi*f_c*(t-tau))*exp(j*pi*S*(t-tau) *(t-tau) )*exp(j*phi). Therefore, after mixing the two chirp, which means we do the dechirp, r(t) = x(t-tau)*x(t)', so the initial phase phi is eliminated.

My question is: Is my understanding right? The inital phase phi is random for each chirp.

Here, I add more details of the issue I faced.

I use two IWR1642 to communicate. I use external trigger to get the two radars synchronized. Radar A start frequency is 77.0000 GHz, Radar B start frequency is 77.002 GHz, Sample rate is 10MHz and other parameters are the same. Radar A transmits 128 chirps, radar B receives these chirps and does the dechirp itself. Finally, we do the FFT on the IF data of radar B, we find the peaks of different chirps are 2MHz which is totally right for our experiment. However, the phase of the peaks in different chirps are not the same and the phase of peak for each chirp seems to be random. 

Therefore, I have the guess(The inital phase phi is random for each chirp ), which casues the r_B(t) = x_A(t-tau)*x_B(t)' = ()*exp(j*(phi_A-phi_B)), and the phase is random."

Can you help clarify?

Regards,

Marvin

  • 0
    TI__Mastermind 39895 points

    Hi, Marvin:

    The signal from two different device have independent LO.  Because of that, the phase will keep drifting.    Maybe you can try to apply the same external clock in order to avoid that.

    Best,

    Zigang

  • 0 in reply to zigang Yang
    TI__Genius 12415 points

    Hi Zigang,

    Thanks for the help. Here is the response from our customer:

    Since I know the TX signal is x(t) = exp{j*2π*f_c*t}*exp{j*π*S*t^2}*exp{j*φ}, in my opinion, the phase φ is random for different chirp(I am not sure about this). In the RX, after a time delay of τ, the signal is x(t-τ) = exp{j*2π*f_c*(t-τ)}*exp{j*π*S*(t-τ)^2}*exp{j*φ}. Therefore, after mixing the two signal, which means we do the dechirp, r(t) = x(t-τ)*x(t)'=()*exp{j*φ}*exp{-j*φ}, so the initial phase φ is eliminated and all the chirps are coherent.

    My questions are:
    1. Is my understanding right?
    2. If so, is the initial phase φ random for each chirp? And Rx uses the mixer with the Tx signal to realize chirps coherency in a frame.
    3. Or the initial phase φ is fixed and known for each chirp in a frame, so all the chirps are coherent.

    Thanks!

    Regards, 

    Marvin

  • 0 in reply to Marvn Arcangel
    TI__Mastermind 39895 points

    HI, Marvin:

    1) The delay part is correct.  But the received signal should apply a gain/phase compare to the transmit signal to reflect the target reflection. But as long as the target is static, you can assume the received signal will have constant gain/phase change compare to the transmit signal. 

    2-3) My understanding is that initial phase is constant for each chirp.  But I am going to double check and get back to you.

    Best,

    Zigang

  • 0 in reply to zigang Yang
    TI__Mastermind 39895 points

    HI, Marvin:

    Here is the answer we aligned internally about initial phase of each chirp.

    The initial phase is not controlled, but it is not purely random as well.   We have seen that phase change from chirp to chirp to remain roughly the same. But it is not predictable and can vary over time.     

    In addition, the reason the customer is seeing the phase difference is because the two devices are running on independent LO clocks , so there would be a phase drift between the two. The only way to sync them is to use common LO, like in the cascade system.

    Best,

    Zigang