SPRACV2 November   2020 AWR1843 , AWR2243

 

  1.   Trademarks
  2. 1Introduction
    1. 1.1 Background – Simple Single-Chip Applications
  3. 2Cascade Incoherence Sources and Mitigation Strategies
    1. 2.1 PCB Routing Imbalances and Device Processes
    2. 2.2 Temperature Drifts
    3. 2.3 Scheduling of Run Time Calibrations
  4. 3Enabling Cascade Coherence and Improved Phase Performance
    1. 3.1 High-Level Summary
      1. 3.1.1 Sequence of Proposed Steps and Introductory Flow Diagrams
    2. 3.2 Saving RF INIT Calibration Results at Customer Factory
      1. 3.2.1 Note on LODIST Calibration
      2. 3.2.2 TX Phase Shifter Calibration and Saving Results at Customer Factory
    3. 3.3 Corner Reflector-Based Offsets Measurement at Customer Factory
      1. 3.3.1 Corner Reflector-Based Inter-Channel Imbalances
      2. 3.3.2 Corner Reflector-Based TX Phase Shifter Errors
    4. 3.4 Restoring Customer Calibration Results In-Field
      1. 3.4.1 Restore RF INIT Calibrations Results In-Field
      2. 3.4.2 Restore TX Phase Shift Calibration Results In-Field
    5. 3.5 Host-Based Temperature Calibrations In-Field
      1. 3.5.1 Disabling AWR Devices’ Autonomous Run Time Calibrations
      2. 3.5.2 Enabling Host-Based Temperature Calibrations of Inter-Channel Imbalances
      3. 3.5.3 Switching of DSP Imbalance Data
      4. 3.5.4 Enabling TX Phase Shifter’s Host-Based Temperature Calibrations
        1. 3.5.4.1 Estimating TX Phase Shift Values at Any Temperature
        2. 3.5.4.2 Temperature Correction LUTs for AWR1843TX Phase Shifter
        3. 3.5.4.3 Temperature Correction LUTs for AWR2243 TX Phase Shifter
        4. 3.5.4.4 Restoring TX Phase Shift Values – Format Conversion
        5. 3.5.4.5 Restoring TX Phase Shift Values – Transition Timing and Constraints
        6. 3.5.4.6 Typical Post-Calibration TX Phase Shifter Accuracies
        7. 3.5.4.7 Correcting for Temperature Drift While Sweeping Across Phase Settings
        8. 3.5.4.8 Amplitude Stability Across Phase Shifter Settings
        9. 3.5.4.9 Impact of Customer PCB’s 20-GHz Sync Path Attenuation on TX Phase Shifters
      5. 3.5.5 Ambient and Device Temperatures
  5. 4Concept Illustrations
  6. 5Miscellaneous (Interference, Gain Variation, Sampling Jitter)
    1. 5.1 Handling Interference In-Field
    2. 5.2 Information on TX Power and RX Gain Drift with Temperature
    3. 5.3 Jitter Between Chirp Start and ADC Sampling Start
  7. 6Conclusion
  8.   A Appendix
    1.     A.1 Terminology
    2.     A.2 References
    3.     A.3 Flow Diagrams for Proposed Cascade Coherence Scheme
    4.     A.4 LUTs for TX Phase Shifter Temperature Drift Mitigation
    5.     A.5 Circular Shift of TX Phase Shifter Calibration Data Save and Restore APIs

3.3.2 Corner Reflector-Based TX Phase Shifter Errors

Customers may choose to calibrate the TX phase shifters (e.g. for better accuracy) using corner reflectors in the factory. This procedure is an alternative to using the devices’ self-calibration of TX phase shift using RF INIT. An example measurement procedure is as follows.

  1. RF INIT should have been performed before this step with TX phase shift calibration disabled. This is so to measure the raw analog’s nonlinearity, which is what’s compensated in-field.
  2. The devices should be configured to use per chirp phase shift mode.
  3. Use profile config API to set the RF frequency at the sensor’s in-field operating range, and configure the TX/RX for good SNR.
  4. Configure multiple chirps with different phase shift index/indices. E.g. 0, 1, 2, 3, 4, … 63, 0 (effects of temperature drifts during measurements, if any, should be mitigated). Configure multiple TXs to transmit these chirps in sequence, one at a time, within a frame, to avoid significant temperature drift during successive phase measurements. An example chirp configuration for such a calibration frame is shown in Table 3-3.
  5. Collect RX ADC data, and process it to find the corner reflector tone’s phase for each phase shift index for each TX. This data must be stored in the sensor’s non-volatile memory for each TX in the cascade.

If deriving the phase shifter calibration values from corner reflector measurements in customer factory:

Equation 4. Factory Measured Phase Shift ArrayCornerReflector,TXm (0 to 63) = Measured Corner Reflector tone phase in RX ADC output for phase shifter settings 0 to 63

The arrays Factory Measured Phase Shift ArrayCornerReflector,TXm (0 to 63) collected at factory temperature are necessary for restoring back to the device (explained later in this document). Example values for the Measured Phase Shift Arrays of Equation 2: [0, 5, 11, … 356] degrees, corresponding to [0, 5.625, 11.25, … 354.375] degree phase shifter settings. These correspond to INL error values of [0, 0.625, 0.25, …, –1.625] degree, i.e. the deviation from ideal expectations. Here, INL error refers to Integrated Non Linearity error.

Here are some APIs relevant for achieving this:

  1. AWR RF INIT CALIBRATION CONF SB (field: “Enable TX Phase calibration” = 0 to disable the device’s self-calibration of TX phase shift errors at RF INIT).
  2. AWR RF RADAR MISC CTL SB (field: PERCHIRP PHASESHIFTER EN) and AWR PERCHIRPPHASESHIFT CONF SB for controlling the phase shift values.
Table 3-3 Example Chirp Configuration for Factory Calibration of TX Phase Shifter Errors
Device 1 Device 2
Chirp Index Enable TX1 Enable TX2 Enable TX3 Enable TX1 Enable TX2 Enable TX3 Phase Shift Index (common to all TXs)
0 1 0 0 0 0 0 0 (0o)
1 1 0 0 0 0 0 1 (5.625o)
2 1 0 0 0 0 0 2 (11.25o)
: 1 0 0 0 0 0 :
63 1 0 0 0 0 0 63 (354.375o)
64+0 0 1 0 0 0 0 0 (0o)
64+1 0 1 0 0 0 0 1 (5.625o)
64+2 0 1 0 0 0 0 2 (11.25o)
64+: 0 1 0 0 0 0 :
64+63 0 1 0 0 0 0 63 (354.375o)
2*64+0 0 0 1 0 0 0 0 (0o)
2*64+1 0 0 1 0 0 0 1 (5.625o)
2*64+2 0 0 1 0 0 0 2 (11.25o)
2*64+: 0 0 1 0 0 0 :
2*64+63 0 0 1 0 0 0 63 (354.375o)
3*64+0 0 0 0 1 0 0 0 (0o)
3*64+1 0 0 0 1 0 0 1 (5.625o)
3*64+2 0 0 0 1 0 0 2 (11.25o)
3*64+: 0 0 0 1 0 0 :
3*64+63 0 0 0 1 0 0 63 (354.375o)
4*64+0 0 0 0 0 1 0 0 (0o)
4*64+1 0 0 0 0 1 0 1 (5.625o)
4*64+2 0 0 0 0 1 0 2 (11.25o)
4*64+: 0 0 0 0 1 0 :
4*64+63 0 0 0 0 1 0 63 (354.375o)
5*64+0 0 0 0 0 0 1 0 (0o)
5*64+1 0 0 0 0 0 1 1 (5.625o)
5*64+2 0 0 0 0 0 1 2 (11.25o)
5*64+: 0 0 0 0 0 1 :
5*64+63 0 0 0 0 0 1 63 (354.375o)

At 25C, the measured phase shift array from corner reflector-based experiments from TX1 of typical AWR2243 device (nominal process) is tabulated in Table 8 in the Appendix section. The same in the form of phase shifter INL plot at 25C is represented in Figure 3-1(b). These have been obtained in TI lab evaluation of few nominal process based devices. They provide representative information and can be used for example in initial sensor development. There can be some amount of process variation in the INL but the trend is expected to be roughly similar (based on evaluation of DoE devices in TI lab).

For post-calibration phase shifter accuracy with 25C factory calibration using corner reflector, refer to Figure 5(b).