21st Century Light Bulbs Using 3D Printer And Chemistry Equipment


[Andreas Hölldorfer] brings his light fixtures into this century by using a couple of modern technologies. The fixtures combine LED modules, 3D printed pieces, and laboratory glassware to give his room a unique look.

The glass enclosure is something he’s had on hand for quite some time but they never actually got used. There is an opening at one end which is meant to receive a stopper. He modeled one including holes for the wires and printed the piece with a 3D printer. Also fabricated in the same way is a bracket that is used for mounting the fixture to the wall. The blossom of components inside the glass are each made up of five LED modules. There’s no word on what he’s using for a power supply or how he managed the cable runs, but he did post an image of two of the fixtures installed in his living room.

20 thoughts on “21st Century Light Bulbs Using 3D Printer And Chemistry Equipment

  1. Or he could have just gone to a surplus chemistry supply website and ordered a rubber stopper.

    Also: with the modules clustered like that with no heatsink, and especially inside a glass container, those LEDs are going to have a lifespan measured in hundreds, not tens of thousands, of hours. Thermal design with LEDs is VERY important.

    1. “Or he could have just gone to a surplus chemistry supply website and ordered a rubber stopper…”
      …to grind up and use as a feedstock to make a 3d printed rubber stopper!!!!

      With that amount of heat, he is getting set up to do some nice steam extraction with that glass.

    2. No one is claiming he couldn’t purchase the part. No one even said there is anything wrong with purchasing all your stuff instead of doing it yourself.

      Speaking of which, I think you should return your expectations for a DIY website back to the store, they seem to be defective :P

  2. I really like the look of the LED “bulbs”. They look like they give off a decent amount of light from the pictures on the source link. My question is, those LED cubes will generate a lot of heat, so I wonder how the author is getting rid of it. If they were liquid filled that would ddo it, otherwise those bulbs won’t last very long. Say we’re talking 5 LED’s per stalk, 1W per LED. Try to hold a 5W resistor with 5W running through it – it’ll get hot enough to scorch paper….

    1. You would use a constant current LED driver instead of a voltage drop resistor in this case. Something like a CAT4101 should work perfect in this case. Then if you drop say 70% the max current across them they will still produce a lot of light but with much less heat. Should keep you from needing a sink for 1W leds that badly.

      1. The problem isn’t the heat created by the driver. I believe Micro was refering to the resistor so that people could relate to how much heat is generated with 5 watts.

        The light will most certainly work, but the primary failure mode for white LEDs is blue-shift. The phosphor will burn out over time and shift the LED towards the blue spectrum. Adding ventiliation and/or running the LEDs at a lower duty cycle would mitigate the issue, but they generally do need a decent amount of surface area to dissipate heat. (I spent a year designing form-factors for products incorporating 10W LEDs a few years ago.)

      1. Actually, you’re the one who’s wrong. LEDs still dissipate most of their input power as heat—googling for ‘led percentage of power dissipated as heat’ finds a reference that claims 90%. So, yeah, a 1W LED is probably emitting 900mW of burning pain.

        Even the most efficient thing we know of (blackbody radiators at 6600K, aka stars) still dissipate 40% of the total EM radiation as non-visible (heat and UV). The next closest are low pressure sodium lamps, which emit 60% of their input power as heat.

  3. Shame about the heat, cos it’s quite attractive, shame it’s not gonna work for long.

    A bit more work with the 3D printer might let him interface some tubes and a small pump to do liquid cooling. Can you run high-power LEDs in water? Maybe de-ionised water? Then put a heat sensor in with the LED bundle to keep an eye on temperature in case the pump fails, or leaks.

    I can think of 2 related fields, one is overclocking PC components, the other is water-cooled HID lights sometimes used in indoor hydroponic gardening.

      1. Oil’s a lot more viscous though, more strain on a pump, and simply more weight for the glass to support. How’s it’s heat capacity compare to water? That’s why I said deionised water. Doesn’t conduct, especially not at 5V or whatever. I was just wondering if the LEDs would mind it.

  4. Okay, so maybe it isn’t’ the most PRACTICAL setup for those LEDs, but if one ran the numbers and could figure just how hard they CAN be run in that configuration, then simply set THAT as the maximum brightness, call it a decorative bulb and you’ve eliminated the issue!
    I have an older IKEA lamp that I modified with a cool neon bulb that doesn’t do squat for lighting, but I love how it glows.

  5. Nice work Andreas, I’ve been playing around with LED’s in different configurations since the mid 70’s and it’s nice to see someone thinking outside the box in a world of people who can only see whats ‘wrong’ with a project instead of praising the hacker/maker, and like so many other do’ers here, you deserve the *YDRDG (You Done Real Darn Good) Award*.

    And I’m fairly confident that when the LED’s eventually do fail, that you’ll have a good time replacing them with whatever improved hack you come up with.

    Hack On, Brother!

  6. Thanks for the compliments.
    Some infos about the LEDs. They consume only 20mA per Cube, so no thermal Problem. And got them for 0,70€ per Cube, you can’t make them for that.
    And the print material is so cheap that often buying something and modding is more expensive. Both parts cost less then 1€.

    1. 20mA for a 12 volt part is still 240 mW of power. If I was you, I would check the temperatures because you’re insulating them inside a glass sphere.

      LEDs really really don’t like heat. If the ambient temperature goes up to 50 C you can expect dimming in roughly a year.

      1. sorry are you kidding ?
        The Glass is not sealed so the air can circulate. The bulb has a volume of about 1 l
        and not all Power goes in Heat (LEDs produce Light about 30%). I’m sure it’s cooling faster than heating.
        I do have a lot of experience with LED Lights, and none of them died by overheating.
        Another thing is the copper wire leads to the outside and provide additional cooling.

        But I will run them tomorrow for a view hours and measure the temperature….

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.