X4 X4 Changes to Pulse Width and Dac Min

Discussion in 'X2 & X4 SoC' started by Jameson, Jul 19, 2018.

  1. Jameson

    Jameson New Member

    Hello,

    I am trying to improve my spatial resolution using the X4M300 by shortening the pulse width. It's not quite clear to me how to directly change the pulse width from the documentation provided. I know there is a finite number of phase bins, however, I seem to get a wide baseband signal when a person is present. Some other types of objects can appear to have a higher amplitude, but a more narrow baseband signal. I can increase the Dac_min value and I see what appears to be a big change for the better in target separation and less noise. I can also raise dac min to almost the dac max value and almost completely squash out the signal. This leaves me with a couple questions.

    Which X4 register can I use to change the pulse width or duty cycle of the pulsed radar?

    What are the pros and cons of adjusting dac_min?

    Best Regards,
    Jameson

    PS. I posted this in the X4300 Forums, but I realize that It actually applies more to the X4 SoC.
     
  2. NSANGOU CHOUAIBOU

    NSANGOU CHOUAIBOU Active Member

    Hi Jameson,
    First of all, X4 SoC is ETSI / FCC (European Telecommunications Standards Institute / Federal Communications Commission) and KCC/FCC (Korean Communication Commission / Federal Communications Commission) compliant at respectively 7.29 GHz and 8.848 GHz center frequency of the transmitted pulse. These are locked values.
    You can't change the pulse length (tau ~= 0.67e-9 s) which is based on the 1.5 GHz TX bandwidth evaluated @ -10 dB) !
    For Dac_min or other radar parameters settings, you need a thorough understanding of swept threshold sampling to perform the correct settings according to your application.
    P.S : X4M300 is not a development, it is a turn key radar module with Novelda algorithms already implemented in it.

    B.R
     
    Aksel Johnsby likes this.
  3. Georg Michel

    Georg Michel New Member

    Hi Nsangou,

    can you explain why the recommended setting of dac_min is so close to dac_max? This is a ratio of less than 1.2. What would be the meaning of setting dac_min close to zero?

    Cheers
    Georg
     
  4. NSANGOU CHOUAIBOU

    NSANGOU CHOUAIBOU Active Member

    Hi Georg,

    1 - DAC_min and DAC_max set the boundaries of the incomming radar signal. Sweeping the DAC is however a time consuming task as we also need a big process gain to have a good SNR. That is why adjusting down the extent of the DAC sweep which limits in consequence the dynamic input range is a compromise in many applications cases, just because of the low signal levels that is normally seen at the input of the receiver.

    2 - Setting DAC_min close to zero will only result in extending DAC swept and thus time consuming during frame building. At the same time, UWB radar is based on emetting ultra short pulses in time domaine, and in consequence a very wide band in frequency domain. The result is the spreading of the energy and your returned pulses are pratically at the noise level. That is why NOVELDA radar technology takes advantage of a technique called Suprathreshold Stochastic Resonance
    At the same time, having DAC_min sets above zero can help clamping interfering incoming signals with higher levels than the useful scattered pulses.
     
    Last edited: Jul 26, 2018
  5. Georg Michel

    Georg Michel New Member

    Hi Nsangou,

    thank you for your answer. I have the impression that the zero level is at DAC value 1024. This is because the recommended values as well as the values in some of the examples are symmetric around 1024. This would mean that values below 1024 are negative levels and values above 1024 are positive levels.

    In other words, the corresponding voltage is proportional to an eleven-bit signed integer, whose most significant bit must be inverted to obtain the respective DAC value. Do you think this statement is correct?

    Cheers
    Georg
     
  6. NSANGOU CHOUAIBOU

    NSANGOU CHOUAIBOU Active Member

    Last edited: Jul 26, 2018
  7. Georg Michel

    Georg Michel New Member

    Hi Nsangou,
    I have no sources. This is just my personal suspicion. The recommended settings for DAC_min/max are (949,1100), see e.g. here. And the settings in some example code are (900,1150). Both value pairs are symmetric around 1024.

    The DAC values are always positive but the received signal is bipolar. So there must be some zero level in the DAC range. I suspect that this zero level ist at 1024 and I am looking for confirmation or falsification of this suspicion.The datasheet says nothing about this.

    Cheers
    Georg
     
  8. NSANGOU CHOUAIBOU

    NSANGOU CHOUAIBOU Active Member

    Hi Georg,
    Yes DAC values are always positive from DAC_min to DAC_max. The mean value of the full DAC range is 1024, which is equivalent to 0 level.

    B.R