• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 88 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.

AWR1843BOOST: mmWave post processing discrepancy

Mark Passia
Prodigy 40 points
Part Number: AWR1843BOOST
Other Parts Discussed in Thread: DCA1000EVM

Tool/software:

Recently I have been recreating the plots from mmWave's post-processing in Python, during which I noticed that the range within the MATLAB (,i.e. post processing) plots seems to be off. It turns out that this discrepancy is due to the post-processing somehow using a different bandwidth than what is shown in mmWave Studio. Remarkably, though, the frequency slope is the same in both cases.

Please take a look at these pictures. In mmWave Studio, the frequency slope was set to 50 MHz/us, which yields a bandwidth of 3001 MHz. The post-processing reports the same frequency slope but a bandwidth of only 1280 MHz. Interestingly, though, the post-processor reports the same characteristic chirp times that are seen in mmWave Studio, in this case, T_idle = 100 us and T_ramp = 60 us. Using T_ramp and f_slope, we get B = 3000 MHz. So where does the post-processing get this value of 1280 MHz?

Please note that the bandwidth value within the post-processing is used for the calculation of the range. For example, in the range-Doppler plot for N_samples = 256:

  • At 3000 MHz, the range would be:
    256 * 3e8 / (2 * 3e9) = 12.8 m
  • At 1280 MHz, the range would be:
    256 * 3e8 / (2 * 1.28e9) = 30 m

The latter one is what can be seen on the range-Doppler plot from the post-processing. 

Does anyone know why this mismatch occurs? Any insights would be greatly appreciated!

(The data were gathered with an AWR1843BOOST in conjunction with the DCA1000EVM)

  • 0
    TI__Genius 14780 points

    Hi Mark,

    Check the following computation for bandwidth calculation -

    samplingTime_sec = numAdcSamples/samplingRate_Hz;
    rfBandwidth_Hz = freqSlope_Hz_per_sec*samplingTime_sec;

    Plug in your inputs to this formula and you can understand why the bandwidth is equal to 1.28GHz. 

    For more details refer to,

    Regards,

    Samhitha

  • 0 in reply to Samhitha Bobba
    Prodigy 40 points

    Hello Samhitha,

    Thank you for your reply! This, of course, explains the 1280 MHz bandwidth seen in the post-processing. However, my question is: what does the "Bandwidth (MHz) 3001.08" value shown in mmWave Studio actually refer to?

    My question is more about the concept rather than the calculation. I understand that after discretization, we effectively use only 1280 MHz of the bandwidth. But since we ramp for T_ramp = 60 µs, does this mean that the total bandwidth we sweep through is 3000 MHz, which is what is shown in mmWave Studio? And that we effectively discretize/use only 1280 MHz due to the ADC settings?

    Thank you for your help.

    Mark

  • 0 in reply to Mark Passia
    TI__Genius 14780 points

    Hi Mark,

    Mark Passia said:
    But since we ramp for T_ramp = 60 µs, does this mean that the total bandwidth we sweep through is 3000 MHz, which is what is shown in mmWave Studio? And that we effectively discretize/use only 1280 MHz due to the ADC settings?

    Yes, your understanding is correct. Since the sampling time is only 25.6us, effective bandwidth is 1.28GHz.

    Regards,

    Samhitha