• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 28 subscribers
  • Views 409 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

DLPC150: NIR returns incorrect spectrum

Renan Santos Adriano
Mastermind 18255 points
Part Number: DLPC150

Hello Team,

I'm posting on behalf of my customer. Please see below inquiry:

This NIR was purchased by the Northeastern University Mars Rover Team in 2019. That team has since graduated, and the NIR has been picked up for use again. However, when tested on water, it returns an unexpected, noisy spectrum (attached). How can we fix this? Thank you.

Water Unexpected Spectrum Data.csv

Regards,

Renan

  • 0
    TI__Mastermind 23070 points

    Hello Renan,

    Thank you for uploading sharing the details of the result. This is very helpful.

    I believe you are using a transmissive head with Cuvette configuration. I noticed that PGA gain in the CSV file was 64.  It is very likely that the sensor is getting statured and the noise is because of that.

    Please reduce the PGA gain to lower values like 16 and 8. Hopefully that should resolve this problem.

    regards,

    Vivek

  • 0 in reply to Vivek Thakur
    TI__Mastermind 18255 points

    Hello Vivek,

    Attached are screenshots of spectra and data of scans when the PGA gain is lowered to 16 and 8, but the spectrum is still just as noisy. Are there any other possible solutions?

    result.zip

    Regards,

    Renan

  • 0 in reply to Renan Santos Adriano
    TI__Genius 12785 points

    Hello Renan,

    Please allow us to consider other ways of lowering the noise. We will reply back by the end of the week.

    Kind regards,

    Austin

  • 0 in reply to Austin Snyder
    TI__Mastermind 23070 points

    Hello Renan,

    On a second look at the uploaded CSV file, I noticed that raw signal strength is bellow 1% for all wavelengths even at PGA Gain 64 compared to the reference lamp intensity (Reference Signal).

    The SNR will be very high at such low signal.  We need to review how the sample is presented for scanning. That could be root cause of such a high noise.

    Could you please request customer to post a picture of the system stem and also describe the system configuration?

    regards,

    Vivek

  • 0 in reply to Vivek Thakur
    TI__Mastermind 18255 points

    Hello Vivek,

    Please see my customer response below:

    On the rover, the NIR is pointed directly at soil. During testing (the scans sent to you), the NIR was pointed directly at water in a cup, about an inch away from the surface. I recently watched a video of the system from previous years and noticed that students fixed some form of lens to the NIR (image of that setup attached), but we have not been using this during testing. Is a lens (the gray piece attached to the NIR in the image) necessary to increase the signal? Thank you.

    Regards,

    Renan

  • 0 in reply to Renan Santos Adriano
    TI__Mastermind 23070 points

    Hello Renan,

    They need to modify the system design/set-up to improve reflected signal strength.  NIR Scan Nano EVM with reflective head requires that sample be placed against the window. An inch of distance too much and sample is not getting even illumination and also a lower collection of reflected light. Please refer to page 14 of the NIRScan Nano user guide for details:

    DLP NIRscan Nano EVM (Rev. G) (ti.com)

    They may want to work with a third-party design house like Optech to design and build custom illumination for such application.

    Home (optecks.com)

    Some of the improvements to consider:

    1. increased number of lamp or power of lamp - Please refer to the application note on stand-off head design for improvement ideas

    App Note Template (ti.com)

    2. Probe connected to EVM with optical fiber so that it can be placed as close as possible to sample.

    3. The sample are coarse, and it will have lot of diffraction and reflection is wide angle. Can you prepare/crush sample before scan

    My guess is that attachment is picture is not just lens, but an illumination head designed to increase distance between sample window and EVM slit.

    Regards,

    Vivek