Ancho Kit (X2) Problem about max distance setting and signal strength raw data

Discussion in 'Chips+Salsa' started by ZhangJunpeng, Nov 10, 2016.

  1. ZhangJunpeng

    ZhangJunpeng New Member

    Now I have the Ancho Kit, and I'm studying the demos pre-installed in the BBB board(with the path ../Demos/Ancho). There are two questions confused me at present.
    1.The default max distance that the Kit can measure is just around 1 meter, how can I change it and make the distance larger than 1 meter?
    2.A "signal strength" question. From RangingDemo, ZoneSense demos we can easily get the raw distance data using "printf" function in C language. But can we get the "signal strength" data from any demo as well? From the, there is a SalsaScope B-Scan figure demonstracts signal strength in the Starting SalsaWeb. Can we get the real signal strength data just like the way we get the distance data?
  2. Will

    Will Moderator XeThru Partner

    1. To change the default distance, just increase the register setting "FrameStitch" to more than 1. Every time you increase this variable by 1 you add approximately a meter of data (depending on your sampling rate), which increases your data size by 256. However, increasing "FrameStitch" will decrease the frame rate because the radar sweep time has been increased.
    2. Absolutely, the signal strength data is essentially the raw data, as this is a DAC counter value (normalized to voltage) vs. time (normalized to distance). Every time there is an update, a new radar frame has been acquired. Just grab the data in the loop right at this point. For instance, in the Ranging Demo the raw radar data is the return of the function call radarHelper_getFrameRaw (line 386).
  3. ZhangJunpeng

    ZhangJunpeng New Member

    Thanks, Will. Your reply really helped me continue to study the demos. But:
    1. It seems that there is a problem happened to my kit that the ranging accuracy is no longer precise. A week ago when I first got the kit, the RangingDemo worked good and the distance it measured seemed to be the real distance * 2. But yesterday when I had a try, the real distance(manually measured with rular) vs the distance the kit measured became strange. Here are a few data:
    manually measure the kit's raw data
    0m 0.2957m
    0.1m 0.3498m
    0.2m 0.4747m
    0.3m 0.5913m
    0.4m 0.7037m
    0.5m 0.7787m
    I have changed the "FrameStitch" from 1 to 3 but finally back to 1, but it still produced such data above. How can the problem probably come from? Should I do some other change or settings?

    2.I found the radarHelper_getFrameRaw function, but I think the function normalizeVolts(line 432 in rangingDemo) directly provide the raw data including "singal strength" information. And in the line 453,the parameter max in the function radarDSP_dp_Max is just the max envelop of the measured signal, and is that the "signal strength" you mean? Or is there a way to compute signal strength from the max envelop(the "max" parameter provide)?
    Thank you. Looking forward to your reply.
    Zhang Junpeng
  4. Will

    Will Moderator XeThru Partner


    First, please familiarize yourself with the radar settings and radar data format. See the PDF Salsa Radar Primer supplied in the Salsa documentation package (I have also attached it for your convenience).

    Next, whenever the FrameStitch setting is changed (or any timing-based register) you should complete a timing calibration, i.e. ExecuteAction("MeasureAll"). For example, set all your registers first, then run a timing calibration. For accurate ranging results you need to have an accurate zero reference, i.e. setting the SampleDelayToReference properly, and completing a timing calibration for accurate timing results over temperature drift.

    Whether you are getting the DAC counter values, normalized voltage, envelope, etc. they are all related to the signal strength and will be relative values. So, maybe I am misunderstanding what you mean by "signal strength", please explain in more detail.


    Attached Files:

    Last edited: Nov 14, 2016