DLP®︎ NIRscanTm Nano EVM showing noisy spectra

Paula Dalla Veccha

Hello,

We are having a problem with two DLP® NIRscan Nano EVM devices. They were working normally but in the latest analysis they are showing noisy spectra. When starting the spectra acquisition, we read a reflectance pattern in "new" mode.

We've already checked aspects like PGA Gain we're using in automatic mode. In addition, we use Hadamard mode with 228 wavelengths and 10 scans to average. Settings are attached. We made a series of acquisitions that way, however, only the last ones became very noisy. Is it a hardware issue? What tests could we do to verify?

Also, we use devices connected to a portable battery as instructed by the datasheet, could this interfere with the acquisition?

Waiting for your answer,

Paula

over 3 years ago

0 Cody Smith over 3 years ago

TI__Expert 3360 points

Hi Paula,

Welcome to the TI E2E forums and thank you for your interest in DLP technology. We will look into this further and get back to you by middle of next week.

Thank you,

Chris

0 Vivek Thakur over 3 years ago in reply to Cody Smith

TI__Mastermind 23070 points

Hi Paula,

Could you please upload CSV file of the scans where you are seeing the problem? This will help us analyze.

regards,

Vivek

0 Paula Dalla Veccha over 3 years ago in reply to Vivek Thakur

Prodigy 10 points

Hey Vivek,

The 3 spectra of the screenshot are attached.

Best regards,

PaulaHadamard App_121003_20211028_151936.csv Hadamard App_121005_20211028_152008.csv Hadamard App_121006_20211028_152020.csv

0 Vivek Thakur over 3 years ago in reply to Paula Dalla Veccha

TI__Mastermind 23070 points

Hello Paula,

Thank you. Could you please also upload CSV file with Column scan method? Please let me know if you seeing similar problem in Column mode too?

regards,

Vivek

0 Paula Dalla Veccha over 3 years ago in reply to Vivek Thakur

Prodigy 10 points

Hello Vivek,

The files in both column mode and hadarmard mode are attached.

Column 1_124006_20211103_091306.csv Hadamard 1_124008_20211103_091913.csv

0 Paula Dalla Veccha over 3 years ago in reply to Vivek Thakur

Prodigy 10 points

Hello Vivek,

The files in both column mode and hadarmard mode are attached.

1104.Column 1_124006_20211103_091306.csv 8802.Hadamard 1_124008_20211103_091913.csv

0 Vivek Thakur over 3 years ago in reply to Paula Dalla Veccha

TI__Mastermind 23070 points

Hi Paula,

I will get back to you in couple of days after analyzing the data. Meanwhile - How does the noise level in column scan compare to Hadamard?

regards,

Vivek

0 Paula Dalla Veccha over 3 years ago in reply to Vivek Thakur

Prodigy 10 points

246 / 5000

Hello Vivek,

As you can see from the spectra I uploaded, spectra using column are less noisy than using haddamard mode. However, we made more than 200 acquisitions using haddamard mode and only from now on we started having this kind of problem.

Best regards,
Paula

0 Vivek Thakur over 3 years ago in reply to Paula Dalla Veccha

TI__Mastermind 23070 points

Paula,

Thank you for update.

regards,

Vivek

0 Paula Dalla Veccha over 3 years ago in reply to Vivek Thakur

Prodigy 10 points

Hello,

I would like to know if you have any updates on this case. Also, I take this opportunity to ask you if you have any idea what might have happened on the device (it's another one) so that we've gotten a spectrum like the one I'm sending you in the attachment. This behavior has been repeated a few times during acquisition, but most of the time the spectra are normal.

Best regards,
Andressa387_2.csv

0 Lori Wang over 3 years ago in reply to Paula Dalla Veccha

TI__Genius 11565 points

Hi Paula,

Thank you for your patience. We will need a few days and get back to you.

Regards,

Lori

0 Vivek Thakur over 3 years ago in reply to Lori Wang

TI__Mastermind 23070 points

Hi Paula,

Thank you for your patience. I noticed the starting at 12th value the intensity spikes at interval of 24. I have reached out to internal team and try to find possible reasons.

Meanwhile ,could you please scan using another PC/ laptop? Would like to rule out if there is system configuration or static variables in the PC firmware(GUI) which is causing this problem.

regards,

Vivek

0 Paula Dalla Veccha over 3 years ago in reply to Vivek Thakur

Prodigy 10 points

Hello Vivek,

Thanks for your response.

The files you requested are attached, I did the scans using another PC.

Best regards,

Paula

TEST_1.csv TEST_2.csv TEST_3.csv

0 Cody Smith over 3 years ago in reply to Paula Dalla Veccha

TI__Expert 3360 points

Hi Paula,

Thank you for sending these. We are still waiting for response from our internal teams to respond appropriately.

Thank you for your patience,

Chris

0 Vivek Thakur over 3 years ago in reply to Cody Smith

TI__Mastermind 23070 points

Hello Paula,

Thank you for your patience. It has been very hectic and it took longer to get back than I originally planned.

Couple of observations:

Unit serial # 6100208 - The first unit

When you plot the observance value from CSV file, it looks much smooth than the GUI picture you shared in your first post. I believe there is no issue with the captured scanned data. The EVM is working fine.

The display graph in the GUI seems to have lot more noise. The very likely reason could be

- it is configured to display plot of single scan rather than final data (average of 10 scan). A single scan will have higher noise and when you average over multiple scans, it is smooth.

Unit serial # 6110046 - The second unit

Vivek Thakur said:
I noticed the starting at 12th value the intensity spikes at interval of 24.

I observed above behavior in this unit. When number of digital resolution does not equally divide the number of mapped DMD columns, some wavelength (mapped digital point) step may have one extra DMD mirror column. You will notice a bump in intensity for those specific values of wavelength. The reference value should also have bump for same wavelength. This common factor neutralizes the observance and the graph should be smooth.

Unfortunately for this unit , the reference value don't show same behavior. I still suspect some abnormal behavior for the software. Could you please try following experiments with this unit:

1. Take scan using "previous" not "factory" scan under "Scan Reference Select" options. For this you need to first take "white point reference" using a 99% NIR standard. This reference, you should take using a "new" option un "Scan Reference Select" . After this you you can samples using "previous" option.

Hopefully this should resolve the issue.

2. If the problem is resolved in step 1 then please go ahead and overwrite the factory reference.

3. If issue is not resolved in step 1 then change a laptop and scan. Please check if issue is still present. If issue is resolved then it was software issue on laptop.

4. If the problem is still present then please re-flash the EVM with latest firmware from the TI website.

Please let me know if this address the problem. If not then very likely hardware issue.

regards,

Vivek

DLP®︎ products

DLP products forum

DLP®︎ NIRscanTm Nano EVM showing noisy spectra

Hello Vivek,