Just Another Simple Spectroscope...

[baffe]

Just Another Simple Spectroscope:

 

First I read some articles on the web about spectrometer construction.

 

http://saisa.eu/blogs/Guidance/?p=1065

 

http://www.popularme...meter-14908361/

 

 

https://www.flinnsci...roscope__1_.pdf

 

http://saisa.eu/blogs/Guidance/?p=1065

 

They show how to prepare the CD to get a grating:

 

http://www.euhou.net...de-spectroscope

 

Then I did a lot of tests to determine distances, measurings, angles and so on. I tested CD's (big range but bad resolution) DVD's (good resolution but small Range and bluray's (almost impossible to handle). I decided to use a DVD for the spectroscope shown here.

 

The operating range ends at red, where the performance appears to be very poor. It shows no infrared but reaches into uv as far as possible but limited by the lens and the webcam's chip. And limited of course by the uv abilities of the used DVD grid.

 

The whole instrument is built into and around a 7/16" professional RF-Connector, it's male and it's female part. After disassembling them removed the center pins with insulation first.

I took the cable intake part, soldered a brass plate onto it to cover the cable entrance and drilled a deep 1,5mm hole into brass plate and housing to get the visor.

 

http://up.picr.de/21948321on.jpg

 

Then I drilled a 2,5mm hole into the plug, cut a M3 thread in to have a possibility to secure the installed and aligned visor.

The circuit board of the USB webcam was glued onto the thread of the female connector. The front element of the lens was removed (to increase focal length) and the filter removed from the lenses back side. The DVD grid was put into the connector thread, aligned and fixed by screwing the plug in and tightening it.

Then the visor was mounted, adjusted and secured with the M3 screw. And the whole unit disassembled some times to adjust and test focus alternating.

After finding the correct setting the whole units inside was painted black and covered with black rubber foam to reduce the stray light. This is the finished spectroscope unit.

 

http://up.picr.de/21948322dp.jpg

 

http://up.picr.de/21948323hs.jpg

 

http://up.picr.de/21948324ij.jpg

 

The module was installed into a blue painted aluminium cast housing with some 1/4" UNC threads for mountig a tripod and acessories.

 

Here are the parts:

 

http://up.picr.de/21948325qt.jpg

 

http://up.picr.de/21948326dj.jpg

 

And the finished instrument:

 

http://up.picr.de/21948327tu.jpg

 

http://up.picr.de/21948328yc.jpg

 

 

And this is a file straight out of the cam (daylight through insulating glass):

 

http://up.picr.de/21948329tp.jpg

 

Later I calibrated the unit and wrote a small software in "lazarus" wich is very close to borlands Turbo Pascal but runs in Windows 7.0. The frame shows the analyzed area in the file:

 

http://up.picr.de/21948395ma.jpg

 

Intended to prepare the files automaticly in size and put them into a calibrated grid:

 

http://up.picr.de/21948396sx.jpg

 

Explanation about the markers:

 

http://up.picr.de/21948397to.jpg

 

Finally here are some samples.

 

Noname "uv" lamp:

 

http://up.picr.de/21948392cv.jpg

 

http://up.picr.de/21948393bs.jpg

 

White LED:

 

http://up.picr.de/21948394mm.jpg

 

MTE301 365nm lamp:

 

http://up.picr.de/21948398uf.jpg

 

http://up.picr.de/21948399wx.jpg

 

I want to thank my friend Dr. Klaus Dieter Schmitt for his help! Especially in measuring my homemade LED lightsource (thermic stabilized to 52°) and giving me documentation about it. It was very important to calibrate the wavelength in the software according to the instrument. And he gave me some helpful hints improving my english.

 

baffe

[colinbm]

Good on you Stefan, keep at it, it is great to see you doing & learning all this, wonderful

Col

[nfoto]

"Simple" takes on an extended connotation .....

 

Nice work. And the tiny tripod is cute, too.

[baffe]

Oh i like this small tripods! I have some of them.

 

Otherwise you find yourself with one hand on the spectroscope, the other on your lightsource. Both pressed together to hold a filtersheet between.

 

Everything is ok but you haven't got a third hand to click a mousebutton.

 

 

So don't worry, call batman.

 

 

-or use that funny tripod!

[Alaun]

Hello Stefan, very nice work!

 

Seeing your post, I experimented a little bit and

a few questions come to mind.

 

You have to focus on the gap (here hole), so I assume the web cam -with your design-has to have a very short minimal focusing distance. A digital microscope camera perhaps?

(I have no clue about web cams).

 

To get the full picture of the spectrum on the sensor, you have to move the grating (DVD) rather close to the gap. So your "gap size” )=drilled hole diameter) is rather large compared to your picture /sensor size. So even a single peak signal would be “smeared” as a kind of steep hill I assume. Though the hole might be nice to bring a light source via a fiver to your spectromter, did you try to “make your hole/gap more look like a narrow gap” to improve the resolution? (The peaks of your UV-lamp curves seem to show this very nicely).

 

Can you control the exposure with the web cam (something you might need to do with a narrower gap)?

 

The “missing” IR response, is that due to a -still in place- IR-filter in the camera? (Might be helpful in blocking out visually interfering higher degree spectrums?)

 

Werner

[baffe]

Hi Werner!

 

Thanks!

 

The lens used in my unit ist not it's own lens bt the lens of different cam and that even modified. But every webcam that I ever tested has a very close minimal focus distance.

 

Of course the cam can not show a spike, but the MTE301 look like that (1:1 crop from software delivered bmp):

http://up.picr.de/21966618ov.jpg

 

I have calibrated gaps with about 0.5mm and one adjusteable currently ajusted to 30µm:

http://up.picr.de/21966721yj.jpg

 

http://up.picr.de/21966722xr.jpg

 

This is the picture with a gap.

 

http://up.picr.de/21966723by.jpg

 

I came to the drilled hole to get some more low light performance. Even if the files are not so sharp because of better overall performance.

 

The dvd grid is not as "passive" as it should be. At wavelength much shorter than 350nm it doesn't work at all.

 

There are some reasons for the missing infrared abilities. First the grid from dvd-r is blue. And the red comes out of the grid more flat than blue.

 

I removed the coated front elemt from the lens but there is still something shimmering in it. But it works useful to 350nm.

 

No I do not have a fiber that is useful for that. And I don't need that fpr my puposes too.

[Alaun]

Stefan, thanks for the information.

 

It seems, further improvements are on track ;-)

 

(with all your housings, are a "Funkamateur" ?)

[baffe]

Oh I'm not finished with that theme yet.

 

So that is a already working narrow band spectrometer with electric servo adjustment for different wavelength.

 

http://up.picr.de/21967571ra.jpg

 

No I'm not a "Funkamateur" (in english Radio amateur) :-)

[colinbm]

Hi Stefan

You don't need a lens, the slit becomes a 'pinhole' & you are effectively focused at infinity.

Col

 

Sorry Stefan

I was thinking of the Amisi Prism Spectroscope that I use.

Col

[baffe]

The slit makes pictures "looking better to the eyes".

 

The round hole produces some kind of corona around the image of a spike if overexposed. This enables the software to detect a maximum even if the image (I don'n use a line but an area!) is totally overexposed.

 

If underexposed then you get a significant "area of increased noise" where the software's statistic functions also can detect a maximum and this works very well too.

 

But of course you loose some resolution...

 

Best would probarbly be a slit with the edges not in parallel. I'm thinking of a slit in delta shape. This will produce both in the image, lines with high resolution because they are correctly exposed and also lines with lower exposure to find maxima at high levels and of course lines with higher exposure to work at lower resolution at lower light levels.

 

There's only one problem: I don't have a drill that drills triangular holes. ggg

 

The blubox is not a precision instrument, but it shows a good overview...

 

The black one shows better results but first one has to calibrate it before every use (the wavelength scale), adjust the range by reading the multimeter display according to a spreadsheet. And of course focus referring to that has to be adjusted afterwards and the exposure has to be measured out very carefully.

 

I'm going to disassemble that black thing anyway because the plastic housing is too floppy and bends if you don't use a trigger delay on the cam. The blubox doesn't suffer from that problems at all! No moving parts and a very compact bulletproof housing...

[baffe]

Following Werners hint with the fiber i took a 62,5/125 fiber of 1m length to cover both, light input and narrow gap.

 

I used my Samsung NV7 (unmodified) fired the flash straight into the fiber at no distance. I put a Grid onto the lens, focussed onto the other end of the fiber and made a shot.

 

http://up.picr.de/21986261fx.jpg

 

The bluspec software was used to analyze the shot. The lambda calibration oft the bluspec spectroscope is not valid!

 

http://up.picr.de/21986262ts.jpg

 

Resolution is really good but the loss of light is tremendous. Remember the fiber is held direct onto the lighting surface of the cams flash.

[Alaun]

But it seems you still got a signal. The entrance of the fiber might have an influence? Is it like a mirror surface, then you might need to take care of the inclination so the light from the flash is not reflected.

 

Can you expose for a longer time and ignite the flash several times?

 

 

Werner

[baffe]

Oh this little test was ok. But sometimes I have sources a little darker than a flash.

 

I'm not talking about the 1 or 2 decibels a better fiber alignment will save...

[Andrea B.]

I've been reading as the work continued and thought I should say hello. Excellent experiment, Baffe !! ;)