Spectrometer

[scratchie]

Has any member current knowledge of gem spectrometers. I don't mean the spectroscope commonly used by gemmos but the apparatus that  connects to a computer and creates a wave length pattern from the stone being examined. I saw that there was a model being sold at the Tucson gem show recently for $1295. Now that's a lot of cappucinos but my aging eyes can't cope with the spectroscope.'
I remember having read that it is possible to make one from the chip in a digital camera.
Any info. will be of interest.
Also ideas about how to win the lottery will be welcome!!!!!!!!
John

[Plutonium]

I don't have particular knowledge of gem spectrometers.
I own an optical spectrometer which I just did a quick check is the same thing that gem collectors use.
I didn't buy it specifically for gems, but there isn't anything different about the way I use it to what gem collectors would do.

There are a couple of other different spectrometers, so I've done a quick search to see if I can find one of the units you mention to see how they work.
I've found two quick examples.
http://www.cigem.ca/research-technology/gl-gem-spectrometer
and
http://www.mightexsystems.com/family_info.php?cPath=102_103&categories_id=103&gclid=CPXiuN2A2rUCFQRapQodPUUALw

Both indicate that they use CCD (charge coupled devices) which is the fancy name for a type of digital camera sensor.
We're off to a good start. I had secretly hoped that the mention of spectrometers might be with reference to "Mass Spectrometers" which at $1500 would be a cheap setup and something I would love to have for the same reasons gem collectors would.
I guess I'll have to find an SEM on ebay.

OK... so the toys you saw at Tuscon were obviously optical CCD devices.
They work the same way as the hand held devices, but having a computer do the work of looking at the spectrum is much more accurate than looking at absorption bands.

So a bit of a techno guess I would say that an ordinary digital CCD just the same as those used in normal digital cammeras, but with the IR and UV filters removed.  The light source looks like it is LED's, either a broad spectrum white or more likely a couple of different LED's to cover the entire spectrum.

That part is simple, the hardware doesn't seem that complicated or special.
It's what is done with the data from the camera. That's where the money is.  I'd guess that if you can get hold of the software, you might try an experiment of just taking a picture of light passing through a gemstone and then running that image through to see what happens.  It's crude, but I would expect to see some results if there is not one more piece of trickery in the units*.

I don't know so much about how the colour information is processed. If I think of a standard uncompressed BMP (bit map) it only contains basically two bytes of information. Colour and Luminance. This doesn't seem to be enough information to resolve a spectrum.
It's digital white light in three channels RGB.  What is required is hundreds preferably thousands of channels. (from an MCA - Multi Channel Analyser).

* I don't actually expect a simple camera image to work for this reason.
The light from the target needs to be split into channels. (the spectrum).
Rather than an ordinary lens of a camera, a prism is used in it's place. That way different colours are scattered to different parts of the CCD, which itself doesn't need to be a colour device. A simple black and white CCD element is all that is required.
Then as the device data is read off, the peaks can be translated and calibrated to the physical position of the prism and CCD.

Even cheap CCD's a quite sensitive and there isn't a lot of software number crunching to be done with this method.
To sum it up you could make one yourself on the cheap without too much difficulty.

What is needed is a good light source which you can use to calibrate the CCD.
Some old black and white CCD TV cameras with the lens removed.
A choice prism that scatters the light onto your CCD
Some method of capturing that information, or even displaying it real time on a TV screen or monitor.
If you can capture that information as digital information it can be ported into something as simple as an Excel spreadsheet and a graph generated to give you the result.

It's not going to be as nice or as simple as the $1500 unit, but for $50 in parts and some of your time, you can at least have a bit of fun.

I can think of another cool type of spectrometer to make using a rotating prism.
I know I couldn't build a rotating prism, but I think I've seen some in old barcode scanners which could used with a very fast optical PIN diode to recover the spectrum.

Sorry it's that's a bit too technical, but somebody might pick up on the idea and have some fun with it.
I'll be having a look through some of my prisms (so to speak) tonight to see what I can find.
 

[Rockman]

Just thought I`d mention that the light source on all spectrometers must be a pure form, like quartz halogen or an ordinary tungsten filament light.(unless you`re
targeting a specific area (wavelength) of the spectrum.
So if you decide to build one of these, you need to use a pure light source, otherwise there will be areas of the spectrum that are already absorbed before you even start, and ID becomes very difficult.
This is why sunlight, fluoros, leds`etc are not used.

Rob.

[Aussie Sapphire]

You can sometimes get suitable halogen lighting on Ebay as for microscope work for those interested in exploring this type of setup - sounds like a great idea to use something more objective to measure the spectrum.

I remember some years ago when I used to work in agronomy research, we had a test that assessed HCN content in various types of pasture - worked very much a litmus paper test. Our lab manager who was very computer geeky worked out a way of scanning in the test papers and then measuring the colour intensity in a way that was calibrated for the actual trait we were trying to measure. More objective than assessing by eye.

Not quite the same thing but does indicate that you can make it more objective and efficient with a bit of lateral thinking.

good luck,
cheers
Leah

[Plutonium]

It should be possible to use a variety of light sources without issue.
All that is required is that the sensor be calibrated before use.

Thinking about it, "I`d mention that the light source on all spectrometers must be a pure form ".
This is not the correct way to describe it.
It is the "unpure" or broad spectrum light that is preferred.
I would normally consider light like the monochromatic light from a laser to be pure (one colour) as opposed to the light from
an ordinary light bulb which is just a broad spectrum blur.

The newer high brightness LED's (GaAs) have very narrow spectrum so they aren't of use. Combinations of the older LEDs would work if you only needed to look at a narrow spectrum. White LED's I dare say could also be used, I've forgotten the spectrum on them.
I 'think' it was a narrow blue line from the driving LED and a broad yellow spectrum from the phosphor.

Fluoros have bright green mercury bands and not a lot of else. Some of the phosphor's in fluoros do have a broad spectrum response but the mercury lines still dominate the picture.

Sunlight.. oh that's a tricky one. Sunlight is nice and broad band white light. It's perfect except that it has absorption lines of Hydrogen and Helium. Which aren't much of a problem since we're not looking for them. It's the other elements absorption lines that cause a problem. The advantage is to a hand held spectrometer these lines are hard to see, they don't really cause a problem. To a sensitive CCD they would be noticeable. However, with a gem in close proximity to the prism, the absorption lines will be overwhelmingly obvious compared to the background lines in the sun light.

I'd sooner hold a gem and spectrometer up to sunlight than a fluoro or an incandescent.
If I get a bit of spare time I'll try and capture some spectrum examples.
I've never tried to capture them before so it's something new to play with. It might be a bit tricky with with my small spectrometer.

[scratchie]

Wow ! Thanks for the responses. I recall enough about physics to get a handle on what you're saying. When I did gemmology I was never able to get a reliable reading of the emission and absorption lines. We used a diffraction grating spectroscope and a prism type but I couldn't master either.I have spoken since with other ex gemmo students who tell me that they couldn't do it either. There is a strong feeling that older people may have lost the ability to use these instruments due to loss of something or other in their sight. It occurs to me that if a number of gemmos can't use the standard spectroscope there should be a market for using digital technology. From what I understand of the machine being sold in the States they have created a database of many gemstones and the computer software compares the stone being tested with the examples in their database. So the suggestion that the software is the key has some merit because the light source and the values for the gemstones has been calibrated and stored.
I have dug out my OPL guide to spectroscopy  and will read back through that to refesh my memory.
Any advice from anyone who has tried to make their own spectroscope would be helpful.
Thanks
John

[scratchie]

Sorry , the last line in my post should read "spectrometer "
John

[Aussie Sapphire]

It occurs to me that if a number of gemmos can't use the standard spectroscope there should be a market for using digital technology.
John

Would definitely be a market for something easier to use - hopefully someone is working on a consumer level version.

cheers
Leah

[Rockman]

Yes, of course I should have said " pure WHITE light source.
There`s nothing terribly complicated about the functions of the GL Gem Spectrometer. Put simply, like Plutonium says, it does work similarly to
a digital camera, where the sensor converts an optical signal into a digital one! 
I`ve owned a GL Gem Spec for several years now, and it was bought with a halogen light source, which is light
with no transmission or absorption across the visible spectrum at all. (700 - 400 nm)

Anything other than pure white light makes it confusing and difficult to determine whether the absorption or transmission in the spectrum
you are observing is caused by the gemstone or is missing from the light itself.
Other light can be used, such as UV or near- infrared , but only for these particular areas of the spectrum.  eg . your not going to see iron
absorption at 450 nm in sapphire if this part of the spectrum is already missing.

I also have my hand held spectroscopes that work very well and use quite often, but I find the digital much faster for doing a number of stones in a hurry,  but the spectroscope/meter is only one instrument necessary to ID gem material, and is certainly not sufficient on it`s own!

Rob. 

 

[Plutonium]

And I happy to say, I'm a failure

Now I've never applied my talents to gems, and of course this has been a very wise choice.
Cubic Zirconia ... I know I can pay $5 and I'm not getting ripped off and it looks better than a diamond
Ignorance is bliss. 

As for terms of science, I applied what I do know and what I am good at.
Starting with sunlight. Yes... what a good source of broad spectrum light.
I've not really paid close attention to it other than some obvious Hydrogen and Helium lines, bot of which are not common in gems,
so I was safe to start. But then I looked closer at some of the other elements.  They've always just been fine lines to me and never of much more than a passing interest. Then I looked closer at them, and closer and damn ! I'd forgotten how much STUFF there is in the sun's atmosphere. There is Aluminium, Carbon, Nitrogen, Oxygen, Sulphur to name a few. Now normally I would think in gems I'd be looking for things like Chromium in Rubies, Titanium in Saphires and Iron in Emeralds etc.
But it's the lighter elements like Aluminium, Carbon and Beryllium that could be masked in sunlight's spectrum.
The lighter elements may be a problem and there's just lots of them so if say you were looking for spectrum lines for something heavy, unless you specifically knew what you were looking for, there is a lot of lines from everything else causing confusion.
But being held up close to the spectroscope, those lines may be more pronounced.

I did take some photos of the spectrum in the scope, but they're basically washed out. Only the mercury spectrum from the fluro's is obvious. The ordinary white fluoro .... big green spectrum from the phosphor.

LED's in my shed, yep. Big blue spectrum from the LED and a big Yellow spectum from the phosphor. It could be used because there's no obvious lines in the spectrum, but don't expect to fine any red lines

Incandescent light was quite good, just not a lot of blue in it.

Halogen down lights were surprisingly good, better than incandescents.

A red LED, yep, just as expected, a nice broad red band but nothing else.

A xenon flashtube .... These are wonderfully white. I figured Xenon lines would be obvious.
DAMN ! There's Xenon, Argon, Krypton and its like a zoo of narrow bands, lots of greens and blues.
How we see this as white I'm not sure

Yes, of course I should have said " pure WHITE light source.
There`s nothing terribly complicated about the functions of the GL Gem Spectrometer. Put simply, like Plutonium says, it does work similarly to
a digital camera, where the sensor converts an optical signal into a digital one! 
I`ve owned a GL Gem Spec for several years now, and it was bought with a halogen light source, which is light
with no transmission or absorption across the visible spectrum at all. (700 - 400 nm)

Thanks Rob, I had to have a look at a couple of halogen down lights to make sure I wasn't having a loan of myself.
I was expecting to see lots of reds and orange, but they were all quite broad, it was really surprising.

I had some misc rubbish gemstones lying around and had a look at them with the scope.
Yep... confirmed. I have no idea what gem is what even if I can work the scope

I had to laugh, I was reading a passing article of microwave spectroscopy. Hmmmm.... grind up your gemstones, inject them into your microwave oven and tune in your FM radio to see what stations are missing.
It requires the destruction of some material, but I can't help but wonder if it might work without doing that and just mounting a gemstone in a waveguide.  I'll have to ask some important people some time if they know of anybody who are tried it.
I currently don't have a particular piece of gear to try it.