So, I bought a science surplus spectrometer a few weeks ago and finally got to calibrating it. It has a 1800 l/mm grating installed. For calibration I used a used a mercury vapor black light lamp, 175W mercury vapor lamp, 532nm green laser and a red HeNe.
After recording the pixel number of light sources listed above I fit the numbers to a 3rd degree polynomial. This equation generates the coefficients the software needs to convert the ccd pixel number to wavelength.
Everything looks good so far. I tested the spectrometer with the 532 laser and the HeNe. Dead on. Next I checked with the blacklight bulb. The Uv spectrum is just slightly off a few nanometers. Close enough to get started I thought. First I checked the WL of a cheap red pointer... 660nm. Good! I'm thinking I must be close. Next, I checked one of the new 465 diodes. This came in at around 446nm. Certainly not right. The lower end must be off enough to cause this I though. Moved on to a g71 diode. This thing measured around 624nm. Thats not right! hooked up a oclaro 700mw red. Same thing 624nm. This should be higher than the HeNe. Hmmm..
Ok, so next I reset the software to read pixel number again. I measured the HeNe as a known baseline then shone the Oclaro diode on the fiber input at the same time. It is showing a lower pixel number than the HeNe. How is this? The cheap red pointer also shows a lower pixel number.
At this point I am confused. I haven't opened the unit up yet. Maybe another diffracted order is bouncing off something. But then the CCD would show two lines at once.
Help....