Hi! It's a good question. CFLs have multiple very strong and stable lines, so it's hard to find something like that. If we had infrared we might use the missing lines from the atmospheric spectrum. But they're not in the visible range so not everyone can use them.

We could use 2 colored lasers but again it's not that easy to find.

What about, maybe, burning a little salt, and using the sodium line? But again it's only one reference, so then we'd have to burn 2 things. Is there something we could heat up in a candle flame that would have 2 strong emission lines?

Some of the early lamps for hplc detectors used mercury lamps, which could cover both visible and uv wavelengths- granted only specific ones. But those have gone the way of the west. Agilent has an app note"phar maceutical analysis with USP<857..... " That was a quick glance, so you may have to dig a little more. You should be able to find it on their web page. But it probably won't help. They talk about didymium solutions, holmium chloride solutions, holmium oxide glass, etc. Some of these can only be used in UV and from past experience, they were all expensive. How about using the leds that glow in 3 colors? You need to switch voltages ( polarity and AC, if memory serves me right). But could these be used for calibration?

Amazon has tricolor, four lead, LEDs. So all of the different color LEDs are in the same package. I'm this case, they give off red, green and blue light.

I'm not sure the Amazon LEDs are the ones to use. Some of the remarks make you wonder. But at least it proves that multicolor LEDs still exist.

You want a sharp, distinct, and stable line for calibration. Which is why the lasers are sought after. They are also hard to find. Please check any LED you get to make sure it fits those characteristics.

Another option is a pack of three lasers marketed as kitty chase toys. They run about $20. Again, quality is an issue. The colors are listed as red, green, and violet. The run time for each, with AAA batteries, is listed as about 10 minutes. One of the problems mentioned was the violet laser was weak. The lasers had the standard warning tag on them, which normally includes wavelength. But from the picture, it wasn't clear if the warning label showed the wavelength. Maybe a better description would.

I checked the pictures of the 3 lasers in one of the marketing shots. The red, green, and the violet laser all showed a laser warning tag with 405 nm as the wavelength. Looks like someone did some photoshopping.

You've got to be careful with lasers. The visible lines are often not their main emission lines. For instance a green laser (532 nm) puts out more energy in the infrared with lines at 808 nm and around 1200 nm. LEDs often have a fairly broad output of about 30 nm and the peak wavelength has some variability (about 5 nm) within the same batch. I normally use didymium glass for calibration (it has several sharp peaks through UV - vis-IR ). It is somewhat expensive, but maybe you could pick up some a little cheaper by looking for "glass blowing spectacles" which are made of didymium glass.

Thanks for the alternate source of didymium glass. Do you know of any less expensive sources of holmium oxide glass?

Thank you.

Checked on amazon. Found three spectrometers, similar in function to the spectrum lab. They are priced at $7, $9, and $37. The $37 is the best. If interested, please look over the data on it, if nothing else. The calibration lines (for example, the sun) are listed on the front. The accuracy is an issue, judging on the one previous comment. The data sheets from the manufacturer are interesting reading, if nothing else.

Sorry, got a phone call. The same line listed for this spectrometer should be usable for the spectrum lab instrumenent. The lines are only listed on the $37 instrument. And DO NOT point the spectrometer directly at the sun. Point it at a price of paper with the sun's image on it.

One of the alternate calibration methods, posted on a different thread, was to use photographic filters that were made of didymium. It sounds like fiction, but they actually exist. Their spectrum is very similar to the usp standard and can be found attached to many of the filter specs. If you choose the right filter, it will be very similar to a standard used for instrument calibration. Not NIST tracable, but you can't have everything. And the cost is much less. Some of the online sources have them for under $20. Choose your poison. Good luck.

Look up didymium glass filters,PNB586 or BG20. That will be a start. Usually, if you find a good vendor, they will have a copy of the UV/Vis spectrum in the specification. There are a number of different configurations (glass size and shape to choose from). More on possible light sources later.

A possible light source for a spectrometer--the light on the cell phone. But what is the spectrum like? I didn't know. Here is an answer from quora;

Quora.com/Can-a-mobile-phone-s-flashlight-produce-uv

Looks like the spectrum is pretty good over most of the visible range.