A White-Light Laser, On The Cheap

Lasers are known for the monochromatic nature of their light, so much so that you might never have thought there could be such a thing as a white laser. But in the weird world of physics, a lot of things that seem impossible aren’t really, as demonstrated by this dirt-cheap supercontinuum laser.

Of course, we’re not experts on lasers, and certainly not on non-linear optics, so we’ll rely on [Les Wright]’s video below to explain what’s going on here. Basically, a “supercontinuum” is just the conversion of a monochromatic source to a broader spectral bandwidth. It’s a non-linear optical process that’s usually accomplished with expensive bits of kit, like photonic crystal fibers, which are optical fibers with an array of tiny air-filled holes running down their lengths. Blast a high-intensity monochromatic laser down one end, and white light comes out the other end.

Such fibers are obviously fantastically expensive, so [Les] looked back in the literature and found that a simple silica glass single-mode fiber could be used to produce a supercontinuum. As luck would have it, he had been experimenting with telecom fibers recently, so along with a nitrogen laser he recovered from a Dumpster, he had pretty much everything he needed. The final setup uses the UV laser to pump a stilbene dye laser, which shoots a powerful pulse of 426 nanometer light into about 200 meters of fiber, and produces a gorgeous supercontinuum containing light from 430 nm to 670 nm — pretty much the entire visible spectrum.

It’s great to see projects like this that leverage low-cost, easy-to-source equipment to explore esoteric physics concepts.

51 thoughts on “A White-Light Laser, On The Cheap

    1. I agree the fiber is probably not single mode there (for shorter visible wavelength). But why is the center dark even though it’s multi mode? Could the visible light be actually guided in the cladding and not the core of the fiber?

        1. I’m not an optics guy, so grain of salt, and have mercy on my soul, but could it be a culmination of photons at the center of the fiber due to internal reflection of multiple wavelengths of light? I mean, the photons aren’t all reflecting at precisely the same angle right? Again, just an uneducated guess.

        2. Forgive my ignorance, but could the core beinging too large coupled with multispectum wavelengths going down a decently long length of fiber cause culmination at the center of the fiber? I mean the photons are traveling at different relative speeds because they arent all reflecting at the exact same angle down the length right? (Not saying the speed of light isnt constant, just that its not going in a straight line.). Forgive me if thats daft, im not an optics guy.

        1. The dumpster behind a corporate cabinet shop was a good source for various species and sizes of wood for me.
          That is, until they put up a sign that said something like,
          “Private Property, Removal of items from dumpsters Prohibited”
          I feel kind of special that they went through the effort making a sign “Just for me”

          1. When they realized that stuff was magically going away before the usual truck arrived, they probably got worried about liability. What if someone gets hurt on our turf and sues us. What if someone takes something home and then hurts themselves and then sues us. So they try to cover their keesters by putting up signs like that. Too bad they don’t just post “all items as-is” or “on your own head be it”

          2. “When they realized that stuff was magically going away before the usual truck arrived”

            Or stuff was magically going away before they were able to retrieve it for personal use.

      1. Up until recently writing style guides called for capitalizing “Dumpster” because it’s a trademark of the Dempster Dumpster company, and therefore a proper noun.

        I only know this because I had an editor correct it on a short story I wrote.

    1. Strangely the dumpster in the yard of my home is somewhat in between. It’s unbelieveable, what people throw out. from simple tools over pieces of furniture, an electric kettle and a bag of computer games + a tablet (needed charge calibration) and a netbook to a working stereo microscope. Which I upgraded to LED lamps. Unfortunately no laser but optics

  1. Hey, stupid guys! STOP publishing articles with youtube videos ONLY. I can make video I’m invented time machine. Publish ONLY articles that have some practical info like project page, github and similar. With YouTube videos only you loose our time and shadow your reputation as a serious site. If continue that way, you become reddit nr. II and get same laugh as on reddit.

    1. I’ve long complained about HAD depending on Youtube videos so heavily. I want to read articles, not waste time on videos of questionable worth.

      Quote: “It’s great to see projects like this that leverage low-cost, easy-to-source equipment to explore esoteric physics concepts.”

      Your concept of “low-cost, easy-to-source equipment” is much different than most of the rest of the real world.

      1. ‘Low cost’ and ‘easy to source’ are always in relation to what the project is doing, as it has to be. Otherwise every project using anything remotely real laser would be ‘expensive’ to some, as even the cheapest homebuilt lasers are not dirt cheap…

        As for depending on video only projects – HAD can only link to and discuss whatever the other folks out there have made, if they only make a video for youtube it is disappointing for most projects would benefit from better explanations. But at the same time I’d rather see this x I may not otherwise have seen… Wouldn’t have discovered this method of white laser existed otherwise – mostly as I have no known need for one I can think of yet. But still its good nuggets of information that you have in the back of your head for if you ever do come across a problem for which this may be a solution.

      2. After you click on the link to the YT videos, click on the area under the vid that presents info about vid. Scroll up a bit and there should (might?) be a link for second-by-second transcription of the video.

        It will present one second’s worth of video speech in text form on one line and even move to then next in sync with the video. You can scroll text ahead or back but it will sync back up if you read too long.

        If you read something in the transcript that you want the video to skip up or back to, just tap on the timestamp link at the beginning of the text line and it moves the video to that timestamp.

        I’ve only used this a few times, so I’m not sure how widespread this ‘feature’ is or if I have grasped it’s ‘benefits’. Like you, just give me some plain old text instead of some dynamic fluff

    2. The documentation is coming, and I might write an academic paper on it as well. One of my projects shown on here from last year just made it into the IEEE, you can’t get more serious than that. I am sure other on here do the same as well, real science and engineering happens on hackaday.

    3. This is why I mostly gave up on the Adafruit blog, for example. Lots of really interesting stuff there, I admire the hell out their operation, but 9 times out of 10 (if not more) the links just go to video.

      What’s wrong with videos? A good fraction of the time, having seen the article title, I just want to know, “did they use X in that?”. With a decent write-up or github repo, I can figure that out in a few seconds. This then decides whether I’m really interested or not, whether it’s worth investing some minutes digging deeper. With video, there’s nothing for it but to sit through someone’s presentation, and with the way YT economics work, there’s incentive to stre-e-e-etch things out, so often times they don’t cut to the chase until several minutes in. Also, it’s linear as heck. The information comes in the order the creator chooses, as opposed to being able to skip around and take the info in the order that works best for me.

      1. Video, like other media, can be used well or poorly.

        Les’ vids are very to-the-point, IMO, and worth every second if you’re into lasers. And here, you _really_ want to see the supercontinuum laser doing its thing. I mean, I’d still rather see it in real life, but a video is probably the next best.

        We only run videos that we liked, and we write them up so that you can figure out if you would too beforehand. YouTube has these perverse incentives that they want to keep you looped in 24×7, but we don’t at all.

        But we also live in reality, and so we cover more videos when there are more good projects being described via video. We go straight to the code / writeup whenever that’s an option, but when it’s not…

        Anyway, we try hard to be a good filter for y’all.

      2. A suggestion when watching YT: You can increase the speed of videos (click on the gear), and you can skip forwards/back with the arrow keys. This way, you can save a lot of time by skipping intros and irrelevant (for you) bits.

    4. “I’m invented time machine”
      “YouTube videos only you loose our time”

      I would love to loose more time, it seems YT often wastes my time instead.
      Does this have something to do with your time machine?

    5. Aka complaining about a free site that posts free content. Don’t like something, no one is holding a gun to your head and forcing you to read/watch posts. HAD pays for the bandwidth so they get to decide the content, if you don’t like it then start your own hack/mod site.

  2. Hey Les (et al.): Is this effect not similar to that exploited in an Optical Parametric Oscillator, only without a resonator & phase matching to select the wavelength?
    Ergo, would it not be more effective to use some explicitly non-linear medium (KTP, LiNbO3 or similar) instead of depending on the accidental non-linearity of an optical fiber?

    1. Only if you want monochromatic output. The phase-matching does not allow for broad-band output in those crystals. Here the main cause of spectral broadening is cascaded stimulated Raman scattering, and the broad input pulse spectrum nicely fills out the gaps between the Raman steps.

    2. I think it depends on the peak power available in the pulse. For the ns pulses used here, Raman scattering and 4-wave mixing are the dominant processes, though keep in mind that material dispersion also helps with broadening.

      If you were to use femtosecond pulses, then self-phase modulation would be the dominant process – because of the insanely high peak powers with these pulses you’d need much less interaction length to get the same broadening since the chi-2 nonlinearities are dependent on peak power.

      1. Ah. OK, so if I’m understanding you correctly, the only reason a conventional OPO works (with the usual Nd:YAG Q-switched 6 ns pulses), is *because* of the resonator and phase matching, to (effectively) increase the power available to the non-linear process in the crystal (or, conversely, increase the effective interaction length).

        In an OPO I used in the past, it put 400 mJ of 532 nm pump energy into a 5 mm spot on the OPO crystal in 6 ns, so a peak power density of 333 MW/cm^2. That’s not far off Les’ estimate of 1 GW/cm^2 he’s putting in the end of that fiber, but his interaction length (200m) is a heck of a lot longer than the 1 cm of OPO crystal.

        Interesting stuff.

        1. That’s my understanding anyway, but yes.. I forgot about the cavity, which really helps in an OPO. I was gonna comment that 400 mJ sounds like really a lot, but then if you’re using frequency-doubled, flash-lamp pumped YAG then this is about right. I always forget how powerful a YAG laser is.

          Most of my experience is in femtosecond lasers (mostly materials modification), but I started in the days when free-space fs lasers were the nrom, and the task laser alignment meant that you had to have a pretty good grasp of the operating principles of the cavity, otherwise the alignment was near impossible to get right. Nonlinear interaction was always my favorite, hence why I got into the field in the first place.

          Kids these days with their turn-key fiber lasers don’t know how good they have it! /s

      2. My recollection of femtosecond lasers was that wavelength broadening was inherent to them, isn’t it? Like wavelength spread is inversely related to the pulse width?

        1. You’re on the right track, but what you’re thinking of is the time-bandwidth product, i.e. the duration of the pulse is inversely related to the available spectral width. The time-bandwidth product basically says that for a given spectral width there is a minimum theoretical pulse width attainable. The reality is that there are lots of small factors that will give non-transform limited pulses.

          What you’re thinking of is the effect of dispersion when the pulse propagates through various media. Since refractive index is wavelength-dependent and the pulse is inherently broadband, each ‘discrete’ wavelength in the pulse will have a different velocity. This of course only assumes the linear component of the refractive index, so you can imagine that the effects become much more complex for nonlinear interactions (which are much more likely when considering very very short pulses with peak powers in the terawatt range or higher).

  3. Is it truly a laser? One of the characteristics of a laser is coherent output, right? It seems unlikely that each of the wavelengths is coherent by itself, and it’s impossible for all different wavelengths to be coherent.

    1. Les says he sees speckle in the spectrum, and that’s a pretty sure sign of coherence.

      And sure, it’s certainly possible for white light to be coherent — it’s just a question of over what distance (or time). A monochromatic laser has narrow linewidth, so it’s coherent over a long distance (can be many meters, or many kilometers in the case of LIGO). A laser pointer with its short multimode cavity, not so much. A LED with 10 nm linewidth, much shorter still. Wideband ‘white’ light, the coherence length will be on the order of microns, like a femtosecond laser… :-)

        1. It’s done all the time. Even with a cheap green laser pointer: Internally it’s a IR diode laser at something like 850 nm, pumping a Nd:YAG laser, emitting at 1064 nm. That infrared goes into a frequency doubler crystal to yield 532 nm green light.

          On a laser I have, 900 mW input electrical power yields almost 50 mW in green light, so about 10% efficient. Not so bad for a laser. Darned good considering it includes a doubler.

          If you’re asking if you can use a thermal infrared source to pump something to make visible light, that’s pretty tough: almost impossible to get the required pump power density until you’re into visible thermal radiation wavelengths already.

    2. The beam is spatially coherent and can be collimated into a nice tight beam, something you cannot do with other light sources. It is not temporally coherent given the broadness of the spectrum. That said Laser speckle is very prominent.
      I am working on a new version that produces narrow discrete lines across the spectrum (looks like a multiline argon laser when it runs)

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.