DLPNIRNANOEVM: SNR and sample distance

Marcello Vanzulli

Part Number: DLPNIRNANOEVM

Hi everyone,

Has anyone studied the best window-sample distance in order to maximize the SNR of this spectrometer?
Thank you!

Marcello

over 4 years ago

0 Vivek Thakur over 4 years ago

TI__Mastermind 22830 points

Hi Marcello,
Welcome to DLP forum and thank you fro your interest in DLP technology.

The samples needs to be placed against the window of DLP NIRScan Nano EVM to capture maximum light/signal that shuld provide the best SNR. Please refer to figure 1.4 (page 14) of "DLP® NIRscan™ Nano EVM User's Guide" for laps and collection lens orientation.
www.ti.com/.../dlpu030g.pdf

Is there a unique problem or concern you want to address?

regards,
Vivek

0 Marcello Vanzulli over 4 years ago in reply to Vivek Thakur

Prodigy 190 points

Hi Vivek,
Well, I was thinking about the distance between the sapphire window and the sample in order to obtain the best SNR ratio.
I know other similar NIR spectrometers need different distance (up to 3 mm); is the best distance between 0.5 and 1 mm? (i.e. 0.75 ± 0.25 mm)
Thank you.
BR,

Marcello

0 Vivek Thakur over 4 years ago in reply to Marcello Vanzulli

TI__Mastermind 22830 points

Hi Marcello,
That is correct. Please make sure that surface clean and consistency of sample presentation is critical for repeat ability of result .

regards,
Vivek

0 Marcello Vanzulli over 4 years ago in reply to Vivek Thakur

Prodigy 190 points

Thank you Vivek!
Just another question: do you think that distance (0.75 ± 0.25 mm) is ensured by only putting the spectrometer on the surface? Is there no need of some sort of a spacer (maybe, a very little spacer)?
BR,

Marcello

0 Vivek Thakur over 4 years ago in reply to Marcello Vanzulli

TI__Mastermind 22830 points

Hi Marcello,
Typically our customer have placed the sample flush against the window and have not used spacer. I would focus more on consistence of sample presentation (distance, orientation etc) for reputability rathen than exact location.

Often the flatness of the sample surface will have higher variation than 1 mm variation.

regards,
Vivek

0 Marcello Vanzulli over 4 years ago in reply to Vivek Thakur

Prodigy 190 points

Hi Vivek,

Thank you for your kind response, I think you have been loud and clear!
I was thinking, also, about the thickness of the sample: is there a suggested minimum thickness, in order to not cross the sample itself and scan the support (for example, a table) below?
Thank you very much once again!
BR,

Marcello

0 Vivek Thakur over 4 years ago in reply to Marcello Vanzulli

TI__Mastermind 22830 points

Hi Marcello,
It is very difficult to suggest a minimum thickness for the sample. It will depend on the material of the sample. Is it opaque or transmissive for NIR wavelength. If the sample is thin or transmissive, you may want to consider place a NIR reflector plate behind the sample.

regards,
Vivek

0 Marcello Vanzulli over 4 years ago in reply to Vivek Thakur

Prodigy 190 points

Hi Vivek,

I understood what you said about opaque and transmissive material; I am working, principally, with agrifood matrices (cheese, milk, fruits, juices, ...); I think they are slightly more opaque than transmissive; what do you think about?

About the reflective plate: if I place it behind my samples, am I changing the mode from reflectance to transflectance? Could it be a problem for the consistency of the measurement? (i.e. Is there something I need to consider in order to build predictive models?)
Thank you very much!
BR,

Marcello

0 Vivek Thakur over 4 years ago in reply to Marcello Vanzulli

TI__Mastermind 22830 points

Hi Marcello,
I have found the transmissive head to be most consistent in sample presentation. It works well with liquid and minimizes variation. You can find one such transmissive head at following partners:
www.ti.com/.../thirdpartydevtoolfolder.tsp
www.ti.com/.../thirdpartydevtoolfolder.tsp

Putting a NIR reflective plate behind will change it from pure reflectance, however you prediction model will work well if training data were taken in the same configuration.

regards,
Vivek

DLP®︎ products

DLP products forum

DLPNIRNANOEVM: SNR and sample distance