AWR6443: Noise peaks near zero range bin observed in one of Tx antenna but not in other

V Prathith Vasudev

Part Number: AWR6443

Hi Team,

I and my team have been experimenting with changing profile configuration in the demo for xwr64xx in mmwave sdk version 3.5.0.4. I am looking at 1DFFT plot
and 2D FFT plots after capturing the radar cube data. I am running the following profile configuration:
tx start time: -8 us
number of adc samples: 50
freq. slope: 256 MHz/us
number of transmitters used: 2
number of receivers used: 4

ramp end time: 15.6 us

sampling freq.: 6000 ksps
other parameters suitably assumed.

We are observing noise peaks near zero range bin in the 1DFFT plot of Tx1. However this is not there in Tx2's plot. The plot is of same Rx for both the Txs.

We are observing asymmetry in the plots of Tx1 and Tx2 for same Rx.
These noise peaks are changing w.r.t to time, i.e. the noise peaks are changing through successive frames even though my test object kept is at same place.

PFA plot of 1DFFT for Tx1 as well as Tx2.

Please let me know any reason for this strange behavior and how to solve this problem

Thanks & Regards,
Prathith

over 4 years ago

0 Ishita Kochar over 4 years ago

TI__Mastermind 31680 points

Hello Prathith,

Sorry for the delay. I have reached out to our SDK team for this. Please expect a response by the end of this week or early next week.

Thankyou for your patience.

Regards,

Ishita

0 zigang Yang over 4 years ago

TI__Mastermind 40085 points

Hi, Prathith:

The chirp you designed have some problem. Can you use "rampTimingCalculator" tab in the radar studio to make sure it is OK? The following figure shows that you can only consider a ADC start time around 5.52us ~ 6us. And you need to adjust the ramp time probably.

Best,

Zigang

0 V Prathith Vasudev over 4 years ago in reply to zigang Yang

Prodigy 70 points

Hi Zigang,

I need more clarification on this.

I have some more data that I want you to look in.

I am using the following profile config:

		Profile Identifier	Strt Freq	Idle Time	ADC Strt Time	Ramp End Time			Slope	Tx strt time	ADC samples	Fs
*profileCfg*		0	60	15.65	7	15	0	0	256	1	40	6000	0	0	48

I have changed the ADC start time here, but I am seeing the same behavior in the plots

I kindly request you to please look into this and give more clarity on the same.

Thanks & Regards,

Prathith

0 zigang Yang over 4 years ago in reply to V Prathith Vasudev

TI__Mastermind 40085 points

HI, Prathith:

You can adjust the last column (highlight in red box) in "rampTimingCalculator" to make sure the ADC start time is bigger than the column of 99% settling (highlight in green box). There is no red alert coming out of it. Bigger idle time is not a problem. Therefore,,this chirp configuration looks good to me.

Best,

Zigang

0 V Prathith Vasudev over 4 years ago in reply to zigang Yang

Prodigy 70 points

Hi Zigang,

Thanks for confirming the correctness of the config. But we still see the noise peaks with this configuration.

And I think this issue is not related to ADC settling time.

Please let me know what do you think of this issue as early as possible.

Regards,

Prathith

0 zigang Yang over 4 years ago in reply to V Prathith Vasudev

TI__Mastermind 40085 points

Hi, Prathith:

I was trying to do a data capture with your configuration in radar studio and check on my side. However, I found that the number of ADC samples can not programmed to be less than 64 samples in radar studio. I will consult with some design expert and get back to you.

Best,

Zigang

+1 zigang Yang over 4 years ago in reply to zigang Yang

TI__Mastermind 40085 points

HI, there:

I have confirmed that ADC samples needs to be 64 and above.

Best,

Zigang

0 V Prathith Vasudev over 4 years ago in reply to zigang Yang

Prodigy 70 points

Hi Zigang,

Thanks for your support

Regards,

Prathith

Sensors

Sensors forum

AWR6443: Noise peaks near zero range bin observed in one of Tx antenna but not in other