Other Parts Discussed in Thread: AWR2243
Tool/software:
Hi Team,
i think i found a Bug in the Configuration of the IWR6243. I am using a custom Sensorboard and configure the Chip with a custom software using the mmWave Link Library. I am currently using the Test Source Mode to validate my csi receiver for the raw data.
I configured the device with the following parameters:
- Centerfreq = 57 GHz
- Slope = 300 * (2.7e9 * 900 / 2^26) = rougly 10 MHz / us | (the value 300 is what the mmWave Link API takes, the rest is the conversion factor for one LSB defined in mmWave Radar Interface Control Document)
- idleTime = 50us
- adcStartTime = 10us
- rampEndTime = 50us
- txStartTime = 2us
- Samplerate = 2 MHz
- Complex Samples per Chirp = 64
I configured the device, captured one Frame of Raw CSI Data, extracted the first receiver channel of the first chirp and calculated the range fft on it. This FFT shows a big single peak at around 500kHz.
According to the formular:
range = c * f_if / 2 * slope
with slope = 300 * (2.7e9 * 900 / 2^26) * 1e6 (the 1e6 is the scaling for hz / us --> hz / s),
an IF Freq of 500 kHz corresponds to roughly 7m in target distance. This does not fit to the configured 5m target distance. When i do the calculation with a slope of
slope = 300 * (3.6e9 * 900 / 2^26) * 1e6
the calculation results in the correct value of 5m target distance. I also tried this for other Test Source Object distances like 2 or 3m and this error occurred with every configuration.
3.6e9 * 900 / 2^26 is the Conversion Factor for Configuring the Slope of the AWR2243. This seams like that somewhere in the configuration chain the old conversion factor is used.
The following screenshot shows an extract of the mmWave Radar Interface Control Document, where the Conversion Factors for the Slope is given.
Is this a known Issue? Do you know if this problem is caused by the Test Source Mode and is not present in normal Operation?
Best regards,
Tim
