AWR2243: I/Q balance

Hi,

Our hw team will review your request and will get back to you early next week

thank you

Cesar

Hello Zhaohong,

Are you using complex 2x mode? in that case we send out the image band information also you would expect to see it.

regards,

Vivek

Vivek,

Thanks for the feedback. Would you please provide following two configuration settings for us to compare with ours?

(1) A setting which does not generate the image.

(2) A setting which generates the image on purpose.

If the image is generated on purpose, would the image have the same amplitude as the true target? In the FFT I shared, the magnitude of the image is only half of the true target. I am also attaching the time domain ADC data for the chirp (256pts).

Thanks,

Zhaohong

Vivek,

"If you use complex 1x mode you will get 0 to +fs information and again you will not see the image band information."

Do you mean that no image  should be visible in Complex 1x mode?

I understand that if the I/Q channels are perfectly matched, there should be no visible image. However, since  the I/Q channels can not be perfectly matched in reality, is there a spec about image level compared to target?

Thanks,

Zhaohong

Hello Zhaohong,

If you have not disabled the IQ mismatch calibrations at boot time, the IMRR is expected to be better than 20dB.

Regards,

Vivek

Vivek,

Would you please point to a document about how to enable and disable IQ calibration at boottime?

Thanks,

Zhaohong

Vivek,

We are using a two chip cascade configuration in the system. Would it affect the RX-IQMM calibration? Is it possible to run RX-IQMM calibration one more time after boot?

Thanks,

Zhaohong

Vivek,

Is there an updated version of this document? It mentions the RX imbalance once but does not show it in the calibration list.

Thanks,

Zhaohong

Hello Zhaohong,

The Interface Control Document (https://software-dl.ti.com/ra-processors/esd/MMWAVE-DFP-2G/latest/exports/mmWave-Radar-Interface-Control.pdf) has the details of all the APIs. You can refer to the "AWR_RF_INIT_CALIBRATION_CONF_SB" API to verify the calibrations that are enabled or disabled at bootup/RfInit.

Coming back to your data, what is the mode you are using in the FFT plot you have shared above? is it real mode/complex 1x/complex 2x? How are you relating the spikes in FFT to be image band related?

Regards,
Vivek

Vivek,

We are checking our configuration software. It seems that the complex 2x mode is used.

adcFmt=1;\n\

We are making the correction to complex 1x and rerun the test. I will u[date you with the results. By the way, what is the purpose for the complex 2x mode and how it works?

Thanks,

Zhaohong

Hello Zhaohong,

You can refer to the following white paper to better understand the difference between the different mode: https://www.ti.com/lit/pdf/spyy007

Regards,

Vivek