Good practice and setup for spectrometer measurements?

[Cadmium]

Steve, he just got a Zeiss Hasselblad 105mm f4.3 UV Sonnar (!) so I don’t think he has to worry about the LENS.

 

Andy, not very many people have such lenses as a UV-Sonnar (!) or a UV-Nikkor, do you? He may not need to "worry", but it is something you might want to consider, if you don't have such a lens.

So I point this out for the usual situations. Don't go buying such filters if you only have more common UV-friendly lenses, you will probably not like the results.

I can't say how the blue channel will look using such filters that remove the upper UV-A with UV-Sonnar or a UV-Nikkor lenses.

Just a word of warning.

[Andy Perrin]

The exclamation mark is because it is an AMAZING lens. I'm jealous.

[Cadmium]

Certainly, I just feel it is important to make clear for most people,

that they will probably not have a good result with such filters that suppress the upper UV-A range if they don't have a special purpose quartz fluoride lens with an (!).

[Shane]

Yes, I know the baseline is dropping below zero which isn't good,

There is a trade off when using these systems, a longer total collection time is typically better however too long and the source intensity drifts.

Are you using the mini D2/W light source or the larger (expensive) source? and is it the OO unit or Avantes? Past testing revealed the Avantes D2/W large source to perform better than the OO unit, as did the Avantes integrating sphere.

Also, best to let the light source have a long warm up to help stabilize it. Then once data collection starts make frequent source and dark current calibration/corrections.

[Andy Perrin]

Certainly, I just feel it is important to make clear for most people, that they will probably not have a good result with such filters if they don't have a special purpose quartz fluoride lens with an (!).

:D

[JMC]

Shane, I have an OO FX spectrometer from a previous job, and the light source is an OO DH-2000-BAL light source. I haven't used an integrating sphere for my work here, I have collimating lenses on the incoming and outgoing fibre. Warm up time for both the light source and spectrometer is definitely something I need to pay more attention to. The graphs shared here were done between 30mins and 1 hour after switch on, which I think is rushing it a bit.

 

Steve, Andy, I often get wrapped up in the minutiae of my work and research, and forget that the main reason to have this kit is to take photos. So I'll sometimes post things because I find them interesting and help me to understand more about how it works, not necessarily that they'd be usable or practical photographically. To be honest I look at a lot of the pictures on the forum in awe, and realise I still have a lot to learn.

[Shane]

30 to 60 mins should be plenty, I think I used to wait 15 to 30 mins. I was actually fortunate enough to be able to use the OO DH-2000BAL and the equivalent Avantes source, as well as comparing and using the Halogen and pulsed xenon sources from several manufacturers. The Avantes was by far the better DH source. The OO unit seemed weaker in output and was less stable so became an expensive bookend. I realize you are not using an integrating sphere for your tests but was just making a point for other potential users that the Avantes sphere was also a much better designed unit. However, I haven't checked to see if OO has redesigned the sphere or not and perhaps the newer DH-2000BAL performs better.

 

Frequent drift calibration/correction was required to eliminate baseline below or above zero. On important spectra I would do this immediately before and after each scan to be sure but that was not normally required. However, (trying to think back) I think about 10 minutes was the limit before drift became an issue.

[Andy Perrin]

Shane, out of curiousity, why couldn’t you just fix the baseline as a post processing step? Does fixing the drift in the spectrometer itself provide better accuracy or something?

[Shane]

Does fixing the drift in the spectrometer itself provide better accuracy or something?

Yes, particularly when post mathematical processing is required to identify subtle differences between identical looking samples.