AWR2944: AWR 2944 Question about strong detection will increase the doppler bin's noise.

?? ?

Part Number: AWR2944

In AWR2944 ES2.0, we found a strange phenomenon: A far strong detection will increase all doppler bin's noise.We think this phenomenon is not corrent. Would you please help to analyze this phenomenon?

Some config about RF: the RX gain=24; the HPF config: 175KHz&350KHz

Test : We use a target simulator to simulate a far range detection in chamber with different power, and after 1D FFT and 2D FFT with windows, we plot the RD map, and calculated the noise of each rangebin, and result are as follows:

A high power detection: we can see the detection is located at rangebin=136, dopplerbin =376, the power is 86.73, we can see the noise of all dopplerbin in rangebin=136 has increase a lot.

the green line is the noise of each rangebin,(noise is calculated by OS-CFAR)， it's also easy to find the noise around rangebin=136 increased.

A lower power detection: we can see the detection is located at rangebin=136, dopplerbin =378, the power is 65.73, we can see the noise is not effected by detection

the green line is the noise of each rangebin,(noise is calculated by OS-CFAR),the noise is not effected by detection..

over 2 years ago

0 Pradipta Bandyopadhyay over 2 years ago

TI__Mastermind 32555 points

Hi,

We will need to check on this with the team internally. Kindly allow us some time to check on this and get back to you.

Thanks,

Pradipta.

0 Wang Qiyu over 2 years ago in reply to Pradipta Bandyopadhyay

Intellectual 260 points

May I know when you could get the feedback?

Thanks.

0 Kaushik Gowda over 2 years ago in reply to Pradipta Bandyopadhyay

TI__Mastermind 20605 points

Hi,

With a stronger signal you can see a small increase in surrounding noise as well. A stronger signal would have more signal to noise ratio as opposed to just increased signal strength or decreased noise level.

I have some additional questions:

Have you performed windowing prior to applying the FFT in both range and doppler dimension?
Can you give a description of the simulated target from the target generator? Would it be possible to look at the spectrum of this signal?

You can try to tweak the above to make the spectral spread near the detection is closer to the noise floor.

Regards,

Kaushik

0 ?? ? over 2 years ago in reply to Kaushik Gowda

Prodigy 100 points

Hi,Kaushik

Thanks for your reply. With a stronger signal you can see a small increase in surrounding noise as well. But I think a stronger signal shouldn't increase all doppler line noise. This makes me a little confused, and for stronger signal detection, the SNR maybe not effected, but for a small signal detection which has the same range with stronger detection, it will affected by the noise increased.

Thanks

0 ?? ? over 2 years ago in reply to Kaushik Gowda

Prodigy 100 points

Hi Kaushik

1. Yes, we test different windowing for range and doppler, and it has same performance:

This is the hamming window result, previous one is the hanning window result

2.We simulate a static target by target generator, and the static target's range is 80m, velocity is zero,this is the spectrum of this signal

0 Kaushik Gowda over 2 years ago in reply to ?? ?

TI__Mastermind 20605 points

Hi,

Can you determine the frequency of this raised noise floor using the bin number and the mapping to the sampling rate post 2D FFT? I would like to verify that this is not any of the known spurs that have previously been observed. (You can refer to the errata document for this)

Regards,

Kaushik

0 ?? ? over 2 years ago in reply to Kaushik Gowda

Prodigy 100 points

Hi Kaushik

Thanks for your reply, but I think this phenomenon didn't have some connect with frequency, Any rangebin has strong target will increase the noise floor. We try both rangebin 30、136 & 77G、78G, the phenomenon is same.

Thanks

Patrick

0 ?? ? over 2 years ago in reply to Kaushik Gowda

Prodigy 100 points

Hi Kaushik

We found this docs, but I think they are not same problem.

0 Kaushik Gowda over 2 years ago in reply to ?? ?

TI__Mastermind 20605 points

Hi,

I understand you are not facing an idle channel spur, but could you please try with a higher gain?

Regards,

Kaushik

0 ?? ? over 2 years ago in reply to Kaushik Gowda

Prodigy 100 points

Hi Kaushik

do you mean to increase the RX gain? we tried RX gain to 36, and has the same problem.

Thanks

Patrick.

0 Kaushik Gowda over 2 years ago in reply to ?? ?

TI__Mastermind 20605 points

Hi Patrick,

Would it be possible for you to share the following items?

ADC data in both cases
Script to process this ADC data.
Your chirp and profile configuration used to obtain this data.

With this, I will have an internal discussion and get back to you with an explanation.

Regards,

Kaushik

0 ?? ? over 2 years ago in reply to Kaushik Gowda

Prodigy 100 points

Thanks for your support, we create another question for this

Sensors

Sensors forum

AWR2944: AWR 2944 Question about strong detection will increase the doppler bin's noise.