DIY spectrometer

ON_

Newly Enlightened
Joined
Nov 3, 2008
Messages
1
Hello everyone,

I would like to be able to compare the color output of several light sources (mainly LEDs) but an off-the-shelf spectrophotometer is (for now) out of the question due to its price.
So the project is to build something that would have a similar result but with a lower cost.

The plan is to use a slit and a diffraction grating, and then to focus the resulting spectre onto a CCD sensor. The basic idea is to use a webcam or a digital camera and then to transform the "jpg" image into a spreadsheed with values for each pixel. The wavelength calibrations and the intensity calibrations can be done by comparing to a known source (the sun for example).

My problem lies in two areas:

1. I am looking for a simple way to transform a jpg file into a spredsheed which can be used by a software like Radiant Imaging's "Color Calculator"

2. Instead of using my digital camera (which I still need for its main purpose - taking photos), or a webcam (sensor might be too small) I thought of using the CCD sensor from an office scanner. I haven't found any way of controlling that sensor without the motor and light that are included in a scanner

so here are my questions: Has anyone here any thoughts about all this? any experience in the CCD-part, or in the picture conversion part?

Thank you in advance

Oliver
 

SemiMan

Banned
Joined
Jan 13, 2005
Messages
3,899
Enough experience with CCDs, spectrometers and the like that I could never justify the time it would take to build my own. A fun project perhaps, but so many issues to doing it right.

Semiman
 

Magic Matt

Enlightened
Joined
Dec 22, 2009
Messages
444
Location
Near to Portsmouth, Hampshire in the UK
HAving seen numerous people attempt to do spectroscopy of the sun using several thousands of dollars worth of CCD camera, and fail to get good readings, all I can say is it's a lot harder than you imagine. :(

The best results were apparently obtained with a monochrome CCD.

My personal opinion of what you're trying to do is that you'd probably find it easier and more accurate to get the manufacturers data on the bin etc. of the LED than to try and determine that data yourself.

If you succeed then I would be extremely interested, as would quite a few others at the observatory, as we would like to get spectroscopy working with our CCD cameras!
 

SemiMan

Banned
Joined
Jan 13, 2005
Messages
3,899
http://www.edmundoptics.com/onlinecatalog/displayproduct.cfm?productID=2306

$2,000 for a spectrometer. I know it may seem like a lot of money, but what is your time worth?

From a principle standpoint, it is not difficult:

- diffraction grating or prism
- some sort of CCD -- generally a rectangular one and definitely monochrome
- Interface hardware
- software
- precision optical alignment
- more software

It would be a fun project... just not easy and I would guess ultimate not that cheap.

Semiman
 

gallonoffuel

Enlightened
Joined
Mar 11, 2008
Messages
570
Location
Northern MD
I know nothing about building a spectrometer, but to answer your first question, i would try to get the output format to bitmap, then read it into Matlab and process whatever you need. You can then either write out to a tab delineated file or write directly to an excel format using xlswrite.
 

PeterRamish

Newly Enlightened
Joined
Aug 7, 2016
Messages
32
Hello everyone,

I would like to be able to compare the color output of several light sources (mainly LEDs) but an off-the-shelf spectrophotometer is (for now) out of the question due to its price.
So the project is to build something that would have a similar result but with a lower cost.

The plan is to use a slit and a diffraction grating, and then to focus the resulting spectre onto a CCD sensor. The basic idea is to use a webcam or a digital camera and then to transform the "jpg" image into a spreadsheed with values for each pixel. The wavelength calibrations and the intensity calibrations can be done by comparing to a known source (the sun for example).

My problem lies in two areas:

1. I am looking for a simple way to transform a jpg file into a spredsheed which can be used by a software like Radiant Imaging's "Color Calculator"

2. Instead of using my digital camera (which I still need for its main purpose - taking photos), or a webcam (sensor might be too small) I thought of using the CCD sensor from an office scanner. I haven't found any way of controlling that sensor without the motor and light that are included in a scanner

so here are my questions: Has anyone here any thoughts about all this? any experience in the CCD-part, or in the picture conversion part?

Thank you in advance

Oliver

try this:

https://spectralworkbench.org

and

http://store.publiclab.org/products/desktop-spectrometry-kit-3-0
 

CuriousOne

Enlightened
Joined
Oct 14, 2012
Messages
813
If you have enough skills in electronics, you can built your own with ease, You will only need high quality prism and precision movement for it. I have built such one and it works just fine. Only drawback is, it can'be used for measuring constantly changing light - it needs about 2 seconds to sweep whole visible spectrum.
 
Top