Re: ultra small spectrometer yields the power of a 1000 times bigger device

Oct 25, 2024 Last reply: 1 year ago 6 Replies
M


Ultra-small spectrometer yields the power of a 1,000 times bigger device


> The tiny, relatively inexpensive devices could be used for customized astronomy research
>
formatting link
Summary:
> Researchers are designing new ways to make spectrometers that are ultra-small but still very powerful, to be used for anything from detecting disease to observing stars in distant galaxies.
>
> Seems a clever way to do, see picture of it here:
>
formatting link

It is certainly a novelty and may work in some research where a very compact high resolution robust portable spectrometer might be useful. Identifying pigments on fake grand master paintings for instance. The kit is down to suitcase size (from half a room full of kit).


It strikes me that calibration of the waveguide is everything and trusting AI to learn the patterns for each wavelength may not work for continuum spectroscopy even if it works well for emission line spectra.


It sounds just a bit too good to be true...


The last really cute trick was high resolution grating one way and prism at not quite right angles so as to map a line spectrum onto a normal rectangular CCD array. PE had a nice example of that in the 90's.

Vote
Twitter Facebook LinkedIn
J

I want the opposite, a super broadband spectrometer, maybe 1800 to 300 nm, to test laser diodes and LEDs. The spectrometer people fight for picometers of resolution over narrow bandwidths.

Vote
Twitter Facebook LinkedIn
J

I cringe at the meaningless and incongruous comparison of a resolution figure with the width of a hair. Anyway, better resolution and smaller size inevitably go at the cost of sensitivity. I don't see these things being of any use in astronomy, although they may be useful in setups where the light intensity is plenty.

Jeroen Belleman

Vote
Twitter Facebook LinkedIn
M

Resolution and signal to noise is everything in deep sky spectroscopy. You can do a lot of physics based on the observed shape of a known to be very narrow band emission (or absorption) line.

Much of the evidence for black holes comes from seeing temporal changes in the spectrum of a hot spot on the accretion disk and the jet hotspots.

You will need at least two sensors for that wavelength range the best back thinned CCDs can do 300nm-1100nm but you will need something else beyond that.

formatting link
Hyperspectral cameras will go out to longer wavelengths than you want but cannot do the UV comfortably (fall off around 400nm).

formatting link
A rough and ready method is that emitters also make useable detectors if you really don't care about resolution and just want a rough idea of whether the LEDs are working and at what approximate wavelength.

Vote
Twitter Facebook LinkedIn
J

We did buy a fiberoptic 3-way WDM splitter, so we can tell 850 from

1310 from 1550 laser diodes, make sure we got the right ones.

You can actually see 850 if you poke the fiber directly in front of your eyeball.

It would be cool to make a wideband spectrometer. We have some ideas, but optics isn't really our business.

Vote
Twitter Facebook LinkedIn
J

Pointing a smartphone camera at the light source can be useful. They see 850nm but not 1310. That doesn't help with distinguishing 1310 from 1550 of course. John

Vote
Twitter Facebook LinkedIn
J

I had an ancient floppy-disk Sony camera that could see 1050.

formatting link
The laser was on loan from NIF, so it may have been pretty powerful.

It would be cool to have a phone that can see IR, and a little set of filters to tell what wavelenght a laser is.

I want a circular bandpass or lowpass filter that I can rotate to slice wavelengths.

Vote
Twitter Facebook LinkedIn

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required