https://www.euro-online.org/enog/inoc2007/Papers/mac-slots.html https://www.euro-online.org/enog/inoc2007/Papers/m https://www.euro-online.org/enog/inoc2007/Papers/mac-slots.html

Strange Noise Spectrum

Opik Taupik Akbar's picture
Forums: 

Dear all,

I have problem with noise psectrum in diagnostic test (performance test). Strange peak shown in 1100-2500nm. Based the reference, that because two things. First, because of order sorter misalignment or didn't syncrone and second because of Torque motor and Encoder no syncrone anymore. I already checked the instrument and there is no misalignment with "order sorter". But i don't know how to check Synchronize between Torque motor and Encoder.

If any of you ever have the same problem and already fix it, I need your suggestion what should i do or give advice about someone who should I contact.

Thank you for your time.

 

Best regards,

Opik

Uploaded Images: 
sabki's picture

Dear Opik,

Please contact staff of PT. Haes Brothers in Jakarta (Specialist of NIRs FOSS Product Indonesia Representative).

Contact Person : Mr.V.Ronaldy Suyanto/Mr.Johanes.

Telp (+62) 021 3846165

email : [email protected]

Thank & Regards,

Sabki

 

Opik Taupik Akbar's picture

Dear Mr. SAbki,

Thanks for your information. So this is impossible for me to fix by my self?

 

Best regards,

Opik.

td's picture

Dear Optik,

Welcome to the Forum! 

I think I have seen something similar but I cannot remember what it is! I'm sure Karl or Phil would be able to tell you.

My only thought is that the spectromer is very short of energy. This might be a lamp or electronic problem?

Last year Karl and I offered to carry out a servey of instrument noise but the response was too small to make it viable. Should we try again?

Best wishes,

Tony

td's picture

Dear Optik,

When I saw your spectra, I was sure I had seen something similar and, by chance, I have found it This was an idea I had about some robust standards for NIR spectrometers. They were robust but the results were difficult to interpret and we went on with different ideas.

No one could explain why the noise produced by GS4 was so different and much higher than from the others.

Hope you will get it sorted.

 

Best wishes,

Tony

Uploaded Files: 
hlmark's picture

Tony - both sets of spectra, Opiks' and yours, are rendered in such poor quality that it's hard to make anything out, much less the wavelength scale markings. Nevertheless, I can tell that there's sinusoid running through the spectra (again in both Opik's and yours), which indicates a source of anomalous spectral variation, although I wouldn't call it "noise" (I reserve that term for cases where the vairability is random). I can think of two possble sources of such a signal being imposed on a spectrum: extraneous room light, or electrical pickup of the power-line signal running around in the instrument. Both causes could give an anomalous signal at either the powerline frequency or at double the powerline frequency. From the rate of scan it should be possible to tell what the equivalent frequency is, of the anomalous signal being imposed on the spectrum. That might provide a clue as to where it's coming from.Both of those problems are known to engineers and are normally accounted for and prevented by the instrument design. But if something breaks down it's possible for the signal to get through.

Howard

\o/

/_\

 

hlmark's picture

I just thought of another possible cause: mechanical vibration. This would have to be synchronized to the scan in order for it to not be rejected by the multiple scans normally used to collect spectral data.

 

\o/

/_\

 

td's picture

Hi Howard,

If you have JNIRS volume 6, you can see the original (and the PCA weights) but they are from 1100 to 2500 nm. If a high level of "noise" is detected by the noise program it seems resonable to call it "Noise" irrispective of its source. I agree with your sources of non-random noise. I have never seen room-light noise but I thought that it would be very regular. I have seen electrical noise in another instrument (not a spectrometer).which had some rapid high amplitue spikes separated by an almost flat baseline. The loadings plot from the PCA on spectra from one of the instruments does show evidence of regular frequncies.

Hi Opic,

Have you made any progress with your spectrometer?

Best wishes,

 

Tony

PS (for Howard) Did you receive my e-mail?

hlmark's picture

Opic - no, either I didn't receive it or unwittingly deleted it as spam - I get so much spam (300 a day or more) that all I can do is CNTL-delete through the list if I don't immediately recognize the sender. Can you try sending again and put something noticeable in the Subject line (definitly include your name), and I'll try to screen more carefully.

I'm pretty sure I have JNIRS Vol 6, but I'll have to look at it a bit later, not at 6 AM here!

Howard

\o/

/_\