• State TI Thinks Resolved
  • +1 person also asked this people also asked this
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 53 subscribers
  • Views 598 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.

LAUNCHXL-CC26X2R1: Sampling times for SDK 5.20 AoA examples

Stephen Endicott
Prodigy 20 points
Part Number: LAUNCHXL-CC26X2R1


Hi, 

I have been having some issues trying to get my IQ samples to match what I expect in terms of antenna numbering and timing and want to confirm that I have the correct understanding.

Below are the details of one trial I ran where I only initially looked at the I data to try and make sense of the data being collected:

The below parameters were calculated using a Sine fit function using the 8 samples from the reference period and look pretty close to what I was expecting.
i amplitude: -65.900
i frequency: 251.589kHz
i Phase: -0.187

These are the parameters that I used. I'm using antenna array 2 on the AoA booster pack.
angle_of_arrival_params={
    #enable filter means only samples during the reference period and sample slots are recorded. Samples from the guard and switching slots are discarded.
  'enable_filter': True,
  'slot_duration': 2, #2uS slot duration
  'sample_rate': 1, #1MHz sampling rate
  'sample_size': 1,
   'sampling_control' : 0x20, #Default filtering using antenna array 2
  'cte_length': 20,
  'aoa_pattern_len': 4,
  #Antennas 1,2,4 map to 2.1, 2.2 & 2.3 respectivly
  #Antenna 2 (middle antenna) is the reference antenna
  'aoa_ant_pattern': [2,1,2,4] 
}

Here is the (xlxs) results of the 1st packet with my calculated times and antenna numbers.


Below is a plot of the 'I' data for each antenna, along with a sine wave generated from the 8 reference samples.

Note 1: I am starting at time = 0. I.e. Sample idx=0 at time 4uS is the first sample of the reference period and sample idx=8 at time 15uS is the first sample in the second half of the first 2uS sampling window.
Note 2: I am using the 5.20 SDK.
Note 3: The data was taken with a clear line of sight, however, it was not as controlled as an Anechoic chamber.
Note 4: I have already read this post on sampling, but I don' think it applies as I am getting filtered results and don't need to discard the guard period and switching period samples myself.
Note 5: Based on this post I am only using a single antenna array. I used array 2 is more easily mapped to the array in iCallBLE/ble_user_config.c

Questions

1. For 1 MHz sampling, where within the reference period and sample slots is the sample taken? Are they sampled at the same point for both reference and sample slot samples? I.e. are they taken in the middle of the uS (first reference sample would be 4.5uS and proper sample at 15.5uS) or at some other point within the slot.

2. For 2, 3 & 4MHz sampling, how are they spread out within each 1 uS reference slot and each .75uS sampling slot?

3. Does the antenna switch during the first testing slot as per the specs (e.g. in my case antenna 1 is sampled in the first switching slot)? Looking at the data in the plot, it appears as though the reference antenna (antenna 2) and the antennas on either side (1&4) might have been swapped. Or is there anything else that might be causing this?

4. Based on the plot and the xlxs results is there anything else that I need to fix us to get more accurate results?

Cheers,
Stephen

  • 0
    TI__Mastermind 37805 points

    Hi Stephen,

    Stephen Endicott said:

    1. For 1 MHz sampling, where within the reference period and sample slots is the sample taken? Are they sampled at the same point for both reference and sample slot samples? I.e. are they taken in the middle of the uS (first reference sample would be 4.5uS and proper sample at 15.5uS) or at some other point within the slot.

    2. For 2, 3 & 4MHz sampling, how are they spread out within each 1 uS reference slot and each .75uS sampling slot?

    While I cannot tell you the exact time within each slot the sample takes place, I can point you to this section of our user's guide, which references the BLUETOOTH CORE SPECIFICATION Version 5.1 | Vol 6, Part B | 2.5.4 which (in turn) states

    "IQ samples should be taken at the same point within each IQ Sampling Window; this starts 0.125 µs after the beginning and ends 0.125 µs before the end of each microsecond period"

    If you're interested in diving even deeper, you could potentially set a GPIO to toggle every time RF activity is detected, capture some samples and review.

    Stephen Endicott said:
    3. Does the antenna switch during the first testing slot as per the specs (e.g. in my case antenna 1 is sampled in the first switching slot)? Looking at the data in the plot, it appears as though the reference antenna (antenna 2) and the antennas on either side (1&4) might have been swapped. Or is there anything else that might be causing this?

    What specs are you referring to? We recommend against sampling during the reference period in the user guide linked above. That section also has important info on what samples to keep and which to discard.

    Stephen Endicott said:
    4. Based on the plot and the xlxs results is there anything else that I need to fix us to get more accurate results?

    Unfortunately I cannot review your XLS because we don't allow google drive access. I think your estimation is pretty close given the constraints of the AoA system, but I will ask my colleagues if they see anything too.

    Best,

    Nate