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.

AWR1642: Calculating approximate receive power from PeakVal

Part Number: AWR1642
Other Parts Discussed in Thread: DCA1000EVM,

Hello,

For my research, I need an approximate receive power for objects detected to allow for a backscatter coefficient calculation. I found a forum post here that describes decoupling the gain values from the received peakVal to get an approx. value. I would like some more information regarding this.

I am able to extract out the rxGain value from the profile_cfg. However, I'm having more trouble with the other:

1. fft-processing gains (range, dopper, angle) -> Do I just divide by the (number of samples, number of chirps, number of antennas)?

2. DFE gain -> Can't find a specific value for this, or how to get the specific value used during processing.

The sentence was also appended with "etc", so I assume there are more that need to be de-coupled to extract out the correct value (LNA)?

Overall, I would like help trying to convert PeakVal to approximate receive power.

Thanks,

Austin

  • Hi,

    Please see some older threads on computing power

    Thank you

    Cesar

  • Hi Cesar,

    Thanks for the links. If I understand them correctly, I can get the RX power value from the ADC signal in and remove the RX Gain (which I have). However, is this data only available when the DCA1000EVM is used? Or can it be collected by analyzing the Heat Azimuth data returned to the radar unit when running the demo software? That data seems to be 2D ADC complex data organized by virtual antenna.

    I have already created software interfacing with the demo application, so I would much prefer to use my current software than write new software using the DCA1000.

    Thanks,

    Austin

  • Austin,

    The raw ADC values are only available when using the DCA1000EVM. By the time you arrive at the range-azimuth heatmap, the ADC values have been removed by several FFT calculations, depending on the specific processing chain used.

    The range azimuth heatmap is a 2D FFT in the range direction, not 2D ADC complex data. ADC values are typically not stored in mmWave reference examples as they are immediately fed into a range FFT after being sampled. Please refer to the Doxygen on the mmWave demos for more detailed information.

    Regards,

    Kyle

  • Hi Kyle,

    Thanks for the clarification. However, back to my original question, you mentioned in the post I linked that it is possible to get a estimated return power using the peakVal. Currently in my code, I’ve divided peakVal by 256 as my guess for removing the FFT gain, then converted peakVal to dbm, then subtracted the rxGain, then converted back to mW.

    estimated_rxPower = 10*np.log(peakVal / 256) - rxGain 
    
    return 10**(estimated_rxPower / 10)

    Above is my current code. Does this seem correct? The values I’m getting seem to be in the general right range. Using the demo software, with aluminum foil 0.5m away from the unit, with the unit pointed directly towards It, I’m getting an estimated receive power back of around ~11.5mW. I’m assuming the TX output from the Radar is about ~15mW With 0db back off.

    My next step was to do a power calibration calc to confirm.

    Thanks,

    Austin

  • Austin,

    TI only provides information in our datasheet that specifies that TX output power at the ball. For the AWR1642, the output power is listed as 12 dBm. While there are calculators online that will convert dBm to mW, you should not assume that since the output power at the ball matches the output power at the antenna. There are losses between the ball and the antenna and can be affected by the frequency of your chirp. The type of reflector used does have an impact as well, although aluminum foil would be a strong reflector.

    It appears that your calculation is arriving at a reasonable answer based only on what you have provided in this forum.

    Regards,
    Kyle