Safe, cheap, and colorful?

[Ben Goren]

First, I'm new here. If this post deserves a clue-by-four, do please feel free to apply it...but hopefully by pointing me in the right direction...?

I'm looking for a reliable method of calibrating a spectroscope. I've got a generic laser pointer with both red and green lasers, and that's likely sufficient for what I need...but it'd be nice to have more than just two points of reference.

I most emphatically don't want high power. I should be able to shine the laser on a white piece of paper and not have to worry about blinding myself or my cat, or burning an hole through the table or setting passing aircraft on fire. If it's no brighter than a flashlight, that's awesome.

It doesn't have to be particularly well collimated; indeed, something with a wide beam spread would be ideal (but I can work with a small point, too). Even novelty shapes are fine.

And cheap is also very, very good. These will see very little usage, so I don't need anything built like a tank; something that serious users consider disposable at best will almost certainly last me a lifetime. Nor do I care about looks or size or ergonomics or battery life or power cord length or anything else like that.

What I do need is well-defined monochromatic light. The more (discreet) colors, the merrier, and spaced throughout the spectrum, too. At the very least, a single laser at the blue / violet end of the spectrum (~400 - 420 nm) would give me the most bang for the buck. Something slightly different from but not overlapping the standard red and green (presumably, 650 nm and 532 nm?) that I already have would be okay if especially cheap. One or two more in the middle (yellow / orange) would be icing on the cake. And an half dozen or so every ~50 nm would be an embarrassment of riches (and, I'm sure, cost waaaaaaaay more than is worth spending on this project).

Also important is stability of wavelength, since this is for an exercise in calibration. If the color is going to shift depending on battery charge or temperature or the like, it's probably not a good choice unless there's a reliable method of calibrating the laser. (I should note that I don't even know if it's physically possible for laser frequencies to drift, short of relativistic doppler effects; apologies if this whole paragraph is nonsense.)

And...I'm not an electrical engineer. Off-the-shelf compete units are ideal. I'm willing to put a bit of sweat equity into the project, but I'd need idiot-proof instructions of exactly what bits to put together where and how; improvising my own voltage regulators or whatever is right out of the question.

So...can anybody offer any suggestions for where to look?

Thanks much!

b&

P.S. As a bit more information, I'm going to be using the spectroscope to create ICC color profiles for cameras. Some of the exposures I'll be making will use sunlight, and Fraunhofer lines are ideal for calibration. But I'll also be making exposures with incandescent and other light sources, which is where the lasers come in. In all cases, I'll be pointing the spectroscope at a piece of white paper or similar reflective diffuser, and never directly at the light source. While I could get away with that for an incandescent bulb, doing that with the Sun -- or, obviously, a laser -- is a good way to ruin cameras and / or eyes. b&

 

[icecruncher]

For stability with wavelength you are way better off with HeNe lasers than diodes, because as you said, diodes are going to be affected much more by temperature and battery.

Online there are a few guys that have calibrated their own spectroscopes for laser, to determine exact wavelength.
Below is the one I've seen it done with, but I personally have no experience with it. you might check the website out and see if it has any other information you can use.

Spectrometers - Science-Surplus

 

[Ben Goren]

icecruncher and BubonicCronic, thanks for the replies. I'm sure you're right about the superiority of gas lasers for this type of application...but I now realize I should have been more specific about the "cheap" part.

The spectroscope in question is a box I made from foamcore, lined with black flock velvet, put a pair of razor scrapers at one end for the aperture and some plastic diffraction grating film at the other, all held together with gaffer's tape.

Apparently, the gold standard for spectral calibration is an helium argon discharge lamp...but those seem to cost as much as a gas laser. There're some vague mentions of using cheap compact fluorescent bulbs as "good enough," which I'll also try, though I'm a bit skeptical.

What I'm leaning towards now...WarnLaser has a banner on these pages, and they have a $40 5mW 405 nm laser pointer.

5mW Violet Laser Pointer of Tiger-VX Series

If I also got similarly-cheap red and green laser pointers, all powered with AAA or AA batteries, I could calibrate them with fully-charged batteries against Fraunhofer lines from the Sun, which should be "good enough" for my precision needs (so long as I remember to always use fully-charged batteries, or periodically re-check the lasers against the Sun).

Sooooo...is that WarnLaser pointer likely to be reasonable for what it is? WarnLaser also has a $25 5mW 532 nm pointer; similar question. They don't seem to have any AAA-powered red lasers...any recommendations for one of those?

Thanks again,

b&

 

[Wolfman29]

The problem with diode lasers is that they drift as much as 5-10nm in either diirection. My personal 405nm runs at about 411nm (spectro'd in the laser lab at the university). Bionic was wrong - for wavelength stability, DPSS is the way to go. I would suggest using a 532nm laser most definitely. Furthermore, you can usually get a cheap small 632.8nm HeNe for around $40 (it won't be new, but it should be easy enough to hook up - just attach the power supply to a wall outlet - or a 12V supply, depending - and they connect the power supply to the HeNe tube), which is far more stable than any diode or DPSS laser. Finally, you can probably use a fluorescent tube for the lower end of the spectrum. Or you could even use a 405nm laser.

 

[Wolfman29]

I understood what you meant, but I think it wasn't clear in your original post, so I clarified. Because he is looking for wavelength stability because he wants to calibrate a spectrometer, power means almost nothing.

 

[Ben Goren]

Wolfman29, thanks for the clarification. Yeah, a 10 nm wavelength shift would more than defeat the whole purpose. As you note, power is irrelevant, so long as it's bright enough to make a visible (even if barely) dot on a white page in direct sunlight and not so bright as to need safety precautions. Where would one find the type of cheap DPSS laser you mention?

BubonicCronic, thanks for the warning and the reference. I'll check them out.

Cheers,

b&

 

[Wolfman29]

Fortunately, any green laser you see for under $20 will be the 532nm DPSS laser. So, just pick one up on the eBay. Because of the nature of the physical process used to produce the 532nm light, it doesn't matter the "quality" of the laser. So if you can find a green laser for under $20, that's the laser you need.

 

[Cyparagon]

You don't need to use lasers as your only references. Mercury lamps for example have precise and narrow spectral lines at 365.4, 404.7, 438.5, and 546.1 nm. Neon lamps could cover your yellows, oranges and reds.

 

[Ben Goren]

Thanks, Wolfman29 and Cyparagon.

I already have a generic red / green combination laser pointer of unknown origin. If I can determine that the green one is ~532 nm, is it safe to assume that it's DPSS and exactly 532 nm, or is there a chance that some other type of laser could also be (almost?) the same wavelength?

I've also done a bit more research on fluorescent bulbs, and my skepticism might be overblown. At the least, they're certainly cheap enough to warrant an experiment. I should be able to compare emission lines from the bulb with Fraunhofer absorption lines from the Sun; ones that are a perfect match should be as reliable as the Sun itself. I'm thinking that a generic fluorescent bulb plus a cheap "UV" fluorescent might be good for the overall scale, plus the green laser to exactly pin down at least one point very precisely.

Any chance there might be one more guaranteed-stable wavelength laser cheaply available? Any of the reds, for example?

Cheers,

b&

 

[Wolfman29]

The green in that generic laser is most definitely 532nm. So you're good there.

 

[Ben Goren]

The green in that generic laser is most definitely 532nm. So you're good there.

Perfect -- thanks!

I'm hoping to finish building the spectroscope tomorrow, which would mean doing the first real exposures with it on Friday...or maybe Saturday or sometime next week. Whenever it happens, I'll post an update on how it all works out.

Cheers,

b&

 

[ScottW]

Fortunately, any green laser you see for under $20 will be the 532nm DPSS laser. So, just pick one up on the eBay. Because of the nature of the physical process used to produce the 532nm light, it doesn't matter the "quality" of the laser. So if you can find a green laser for under $20, that's the laser you need.

Most the cheap 532's will also be throwing out some amount (possibly quite a bit) of IR, like 808nm in addition to the 532nm. Is that going to throw off a spectrometer calibration? Or is it actually helpful by giving two different wavelengths to use in calibration?

 

[Ben Goren]

Most the cheap 532's will also be throwing out some amount (possibly quite a bit) of IR, like 808nm in addition to the 532nm. Is that going to throw off a spectrometer calibration? Or is it actually helpful by giving two different wavelengths to use in calibration?

Scott,

That's an excellent question. The short answer is that it's not a problem for me.

First, this is an exercise in building ICC color profiles for digital cameras. Except for certain models and / or modifications for certain types of astrophotography, all have IR cutoff filters -- a piece of glass opaque to infrared covering the sensor. Its effects are easily detectable in a photo of a spectrum after you play around with image brightness and the like. I don't remember where the cutoff is, but it's much short of 800 nm.

Second, the goal for this particular exercise is to be able to have fixed reference points for different wavelengths. So, when the light gets spread out through the diffraction grating, there'll be a very bright green dot in the green part of the spectrum, and I'll know that that dot has a wavelength of 532 nm. If I also have a few more similar references, say of emission lines from a fluorescent lamp or Fraunhofer absorption lines from sunlight, then I can map out where in the photograph of the spectrum each of the wavelengths is.

What I'll be doing is making an image with that sort of reference information in it, and then, without moving the camera, make another image with an illuminant more suited to photography. I'll then be able to use Photoshop to precisely lay a reference grid on top of the reference photo, and then swap the "real" one out for the reference photo. Then I can feed that image (with the reference grid) to the ICC color profiling software, which will (in theory) be able to figure out that such-and-such an RGB value in an image corresponds to such-and-such a real-world color. It then does a bit of math and creates an ICC color profile that can be used to accurately correct the color of photographs.

So, even if there was contamination in some other part of the visible spectrum from the laser...well, if it was reliably monochromatic, that'd be awesome because it'd mean another reference point. If not, it'd only be a problem if it was of a similar wavelength and intensity of one of the other reference points I was using such that I couldn't tell (by eyeball) the two apart.

Cheers,

b&

 

[hakzaw1]

PM coming about a good deal for you

 

[The Lightning Stalker]

DPSS is 'good enough' unless it's a high resolution spectrometer. Neon and mercury bulbs are the way to
go though as far as stability and multiple wavelengths. It really depends on the sensitivity of your
spectrometer both in terms of resolution and light gathering power. You won't know until it's all put
together weather you can measure a tiny low power neon source.

 

[Ben Goren]

DPSS is 'good enough' unless it's a high resolution spectrometer. Neon and mercury bulbs are the way to
go though as far as stability and multiple wavelengths. It really depends on the sensitivity of your
spectrometer both in terms of resolution and light gathering power. You won't know until it's all put
together weather you can measure a tiny low power neon source.

Spectrometers in the graphic arts industry typically only have 10 nm resolution, with 3 nm being "high resolution," so I'm pretty sure that fixing the spectrum to within 1 nm should be plenty.

I've been looking for cheap mercury, neon, and sodium lamps. Cheap bulbs aren't hard to find, but they all seem to be for odd fixtures -- street lights and the like. If anybody might have any suggestions in that direction, I'd sure appreciate it....

Cheers,

b&