AWR1243BOOST: 1243 cascading angle estimation documentation

Tiancong Gu

Part Number: AWR1243BOOST

Hi,

I am cascading two 1243 devices as a 4*8 MIMO radar and collecting data using DCA 1000 EVM. I followed the cascaded documentation building the hardware. Now we can get TDM data and do the range and Doppler fft. But when I did angle estimation, I found that there is a random phase difference between data of each receiver received from tx1 and tx2 which is same for tx1 and tx3, tx1 and tx4. Is this randomness avoidable? I know there is a calibration algorithm for 1642, but that is calibration for receivers of one device, does TI have any documentation for the phase calibration between different tx antennas?

over 7 years ago

0 Vivek Dham over 7 years ago

TI__Guru 58035 points

Hello,
There would be a phase mismatch between channels due to board routing and antenna mismatches also. These cannot be avoided and would exist in every solution. Hence these mismatches are measured once on a board and compensated for in the algorithm.

Regards,
Vivek

0 Tiancong Gu over 7 years ago in reply to Vivek Dham

Expert 1070 points

Yeah I understood this part, and I think compensating this constant mismatches is what TI did for one 1642 deivce.

But as for 1243 device(s), could you confirm that there is a variable time difference between each ADC sampling which may cause a variable phase difference between the data received from tx1 and tx2?

0 Tiancong Gu over 7 years ago in reply to Vivek Dham

Expert 1070 points

I just read this thread who seems like share the same issue: e2e.ti.com/.../719635
So it could be the issue of using MIMO?

0 Vivek Dham over 7 years ago in reply to Tiancong Gu

TI__Guru 58035 points

Hello,
The phase variability/mismatch between TX1 and TX2 inside the chip is typically < 3deg. The variation is dominated by PCB and routing.

Regards,
Vivek

0 Tiancong Gu over 7 years ago in reply to Vivek Dham

Expert 1070 points

I did experiment on 1243Boost as well. There is always a 6-degree difference between tx1-rx4 and tx1,2-rx8 which is because of the phase discontinuity.

This is what exactly the other thread's issue as well I think:

Could you please confirm that 1243 does have this issue and how to do the phase calibration?

0 Vivek Dham over 7 years ago in reply to Tiancong Gu

TI__Guru 58035 points

Hello ,
As mentioned above there would be a static phase difference between chips and between antennas. Typically this is just compensated for in the angle estimation algorithm. You measure the phase difference , as you have done above, and add these offsets in your computation.

Regards,
Vivek

Sensors

Sensors forum

AWR1243BOOST: 1243 cascading angle estimation documentation