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.

IWR1843: The impact of runtime calibration function closure

Part Number: IWR1843

Hi,

Problem phenomenon: Long-time running, the detection rate is found to decrease.

It is found to have a certain relationship with temperature. In the Application Report Self-Calibration in Ti's mmWave Radar Devices document Figure1-2, the runtime calibration function is described. The chip will self-adjust the TX\RX Gain value as the temperature changes. If we put this The gain is fixed to compensate a relatively high value. What impact will this have on subsequent mass production?

  • Hello Nancy,

    I'm afraid I was not able to understand your concern much. Can you maybe rephrase it?

    Are you questioning on the need for calibrations in the sensor? 

    Regards,

    Ishita

  • Hi,

    Customer don't want to enable calibration and he wants to fix the gain to a relatively high value, will it have an impact?

  • Hello Nancy, 

    Just to give a brief background on this, the Rx gain or TX power (or any other RF/Analog aspect) may vary from device to device due to manufacturing process variations and also across temperature. Hence, the purpose of calibration is to ensure the RX gain and output power are maintained as configured by the user despite variations in process and temperature. To achieve this, the internal processor adjusts the mmWave circuit configurations at initialization (to mitigate effects of process variation) and periodically at runtime (to mitigate effects of temperature drifts).

    Now coming to your question, Yes, you can disable the periodic calibrations if not desired, and only enable one-time calibrations instead (fixing the gain to a particular value rather than periodically letting the system update) 

    You can refer to the details in the following API: "AWR_RUN_TIME_CALIBRATION_CONF_AND_ TRIGGER_SB". This API is also used to trigger one time calibrations instantaneously for various RF/analog aspects.

    With this, the user can accordingly decide themselves on the need for calibrations in the device. 

    Hope this helps and gets you the required information.

    Regards,

    Ishita

  •  Hello Ishita,

    Are periodic calibrations and one-time calibrations two different things?How can we close the periodic calibrations?Which API we can called?AWR_RUN_TIME_CALIBRATION_CONF_AND_ TRIGGER_SB can disable the periodic calibrations function?

    BR,

    Bryant

  • Hi,

    Thanks for your reply. And i have a few more questions.

    1. If we only enable one-time calibrations, will the performance of the same device start at different temperatures also be different?

    2.Currently, I have read the rlTxGainTempLutGet/rlRxGainTempLutGet LUT table through the interface, and found that the value we read each time it is started may be different. Is the difference between each startup caused by a one-time calibration? -What are the influencing factors of time calibration, and which factors are related to the fixed scene?

    3. The purpose of periodic calibrations is whether there are other considerations besides maintaining the overall performance. If it is only to maintain the performance, then we set a fixed value in the LUT, and the performance of the device should be more stable and predictable. So can you explain the disadvantages of setting a fixed value, so that we can make our changes more confident

  • Hello Nancy, 

    Sorry for the delay. Let me get back to you on this tomorrow. Hope that's okay with you. 

    Regards,

    Ishita

  • Hello Bryant, 

    Yes, the API mentioned by you can disable the periodic calibrations. Please refer to the mmWave ICD document for details.

    Regards,

    Ishita

  • Are there any updates to the three questions Nancy raised on July 12? I'm also paying attention to them

  • Hello Users,

    Apologies for the delay on this, had to get some information for this from the internal team. Please find my response below : 

    1. Yes, that is correct. As you can see from the graph illustrated below, the performance varies at different temperatures and degrades at high temperatures. 

    3. As I mentioned in my previous response, the RF parameters of the device may vary from device to device due to manufacturing process variations and also across temperature. Calibrations help to ensure those values remain AS PROGARMMED BY THE USER across all temperatures and process variations. If calibrations are not enabled, RF parameter say TX Power will degrade a lot at higher temperatures.

    Now coming to the types of calibrations, we have periodic calibrations enabled within the device, so it basically runs the calibrations routines at a periodic rate (as entered by the user) and based on the temperature of the device at that point in time, calibrates the value. There is an LUT for this purpose and it has multiple values. When you set a particular gain setting in the profile config, the firmware computes the Rx gain LUT values (there are 18 value for -40C to 140C) temperature range (and similarly for TX). 

    This is the disadvantage of keeping a fixed value throughout. If the device works across temperature, the calibration value/LUT value will also vary across temperature, and hence the user/host need to take that into consideration while programming a fixed value. You can run the one-time calibrations instantly any number of times by issuing AWR_RUN_TIME_CALIBRATION_CONF_AND_ TRIGGER_SB API. For example, you can keep a temperature range as a deciding factor and if the device at any point in time changes the range, you can issue the run time API with one time calibrations enabled to account for that temperature change. 

    I skipped Q2 because I felt it was more or less covered in the above comments. Let me know if you need any clarification on that.

    Hope this gets your started. 

    Regards,

    Ishita