Data structure

Discussion in 'Chips+Salsa' started by Hanna Hebestreit, Apr 5, 2016.

  1. Hanna Hebestreit

    Hanna Hebestreit New Member

    Hello everybody,
    How is the output data of the development kit? A matrix with 256 or 512 samplers to get information over the distance and then the "time" samplings? so 512 x time sampling?
    What kind of information do we get? Amplitude of phase information? in Voltage? DAC counts? etc.
    Last edited: Apr 5, 2016
  2. Michael Lehmann

    Michael Lehmann Moderator

    Hey Hanna,

    Sorry for the late response.

    The Cayenne has an X1 radar chip which has a 512 sampler which is a default distance of 2 meters. This is also with the default setting of the Range Accuracy of 4mm. I would recommend looking at the Cayenne Datasheet for more information on that.

    The Ancho has an X2 radar chip which has a 256 sampler which has a default distance of 1 meter.

    Both systems can be frame stitched to get longer distances.

    Finally, the information you get are DAC counts which are related to the voltage. You can also use our library functions to get more information out such as the amplitude of the phase.


    Mike Lehmann
    Olav Liseth likes this.
  3. Hanna Hebestreit

    Hanna Hebestreit New Member

    Thanks a lot!
  4. Hanna Hebestreit

    Hanna Hebestreit New Member

    Sorry to bother you again with some stupid questions...
    Why do we have oscillations in the raw radar data? Do the oscillations occur due to the gaussian pulse? So the interference of the reflected pulses from varios objects?
    Last edited: Apr 12, 2016
  5. Will

    Will Moderator XeThru Partner


    I assume you are referring to the oscillations in the pulse itself? The Cayenne (X1) is a 1st order Gaussian pulse approximation, while the Ancho (X2) is an 11th order Gaussian pulse approximation. Please refer to the respective radar chip data sheets for more information on this. Depending on the environment, i.e. multi-path, and the number of targets in the antenna field of view; yes, there could be numerous pulses converging in similar spots that would cause the oscillations to merge and appear lengthened, or cancel each other.

  6. Hanna Hebestreit

    Hanna Hebestreit New Member

    Regarding the pulse shape I have a question. You try to approximate a Gaussian modulated sinusoidal pulse or a pure Gaussian pulse? I uploaded a graph with both of them. Because if I look at the time domain of your pulse (Ancho X2) it looks more like the Gaussian modulated sinusoidal pulse. I think sometimes its also called Gaussian pulse, but I have few experience on this topic, so I want to be sure which shape you try to approximate (red or blue).

    Attached Files:

  7. Stig Støa

    Stig Støa New Member

    Hi Hanna.

    I think the short answer you are looking for is "blue"

    If you do for example a down conversion to baseband the answer is "red".

    Either way, pulse absolute bandwidth and envolope will be the same.

    See also the X2 datasheet for information about the different pulse generator settings, PGSELECT in particular, and it's characteristics.

    Best regards