Homemade Silicon Carbide LED

Here’s an LED indicator which was made at home out of a Silicon Carbide (SiC) crystal. The concept is simple, but a bit of trial and error goes into getting that tiny amber spot to light up.

The guesswork comes in finding the right piece of crystal. First [KOS] broke it into tiny pieces, then he started poking the chunks with electrified probes to see if he could get some light out of them. Once an active area was found he needed a base for the crystal. The image above shows the two nails which he used. This provides a large mounting area that also acts as a heat sink to make sure the LED won’t burn itself out. There’s a solder blob which he kept molten with his iron until the crystal could be pushed into place. That holds it securely as the pin which serves as the cathode is positioned.

The whole setup is soldered to some protoboard and is ready to use. This is the second time we remember seeing this technique used to fabricate LEDS. The first time was an accident.


  1. Dan B says:

    Reminds me of the old crystal radios I used to play with.

    • PI says:

      This SHOULD remind you of crystal radio.

      Back in old days, before the advent of packaged semiconductor diodes, your typical radio detector (diode) consisted of some kind of a metallic crystal (often galena) and a metallic probe or “whisker.” The point of the whisker against certain spots on the crystal formed a PN junction.

      Galena makes for a good radio detector, put the whisker pressure on the crystal is very slight, which means that the detector as a whole is sensitive to mechanical vibration. This is a disadvantage on shipboard or in a portable radio set, for example.

      Silicon carbide is not quite as touchy as galena, and promised to be a better detector for shipboard use, except for one thing: The forward voltage drop was much higher than that of galena. In radio terms this meant that the crystal was far less sensitive to weak radio signals.

      To combat this, experimenters used a battery and a pot to apply a forward bias to the junction…in order to push it to the almost-but-not-quite-on condition. A weak signal superimposed on this bias could then be detected.

      The fact that a DC voltage was applied a silicon carbide crystal meant that the discovery of the first LED was inevitable. In fact, there are articles from the the early 1900′s in which experimenters report seeing glowing specks of light on their carbide crystals. I think lab-documented electroluminescence goes back as far as 1907.

      Of course, nobody had yet conceived of a light-emitting diode, so there was considerable debate as to what mechanism was responsible for the light. Some thought the light was the product of microscopic arcs. Others, rightfully so, suggested that the light was being generated by an as-yet unknown process.

  2. giacomo says:

    This would make an awesome indicator for a steampunkish project.

  3. NewCommentor1283 says:

    other then “buy anything and everything on ebay”,
    where might one get ahold of such SiC ?

    i mean, i think ive heard of SiC cutting blades
    or something else using SiC in stores.
    im not sure. possible? expensive?

    anything broken i can get for free that might have suitable SiC in it???

    • giacomo says:

      Hm… well a quick search for ‘silicon carbide’ on Home Depot turns up a wide range of grinding bits and sandpaper. Not sure if you’ll find anything of the purity required. Perhaps individual grains from the sandpaper? As for blades, it seems those are tungsten carbide.

      • NewCommentor1283 says:

        JUST NOW i finished my googling… yes you are correct!

        still thanks for the reply.
        it was all a along read and TOTALLY worth it btw

        PS: “cat’s whisker” as is can light up…
        hmm “PRE-1960″

    • hboy007 says:

      black sand blasting material should also be SiC, you can buy it in buckets of several kilograms.

  4. barsmonster says:

    Apparently, it was done before by other guys, for example http://www.dlip.de/?p=99

  5. Charles says:

    I can’t seem to find anything about which side is the positive end and which one is the negative end… Is it just the location that’s probed?

  6. conundrum says:

    I have a few known glowing pieces here if anyone is interested, mailto mandoline at cwgsy dot net
    Took me ages to locate a big crystal, try the “new age” shops and ask for carborundum.

  7. Harvie.CZ says:

    I have some SiC and i want try to hotglue everything together (instead of using second nail with spring) to make compact and robust LED :-)

    • SV3ORA says:

      Well, this was my initial thought, but you have to ensure that the two suports are insulated or else you will have a shortcircuit.
      If one had an expensive wire bonding machine he could evem make it inside a transistor package as well :)

  8. Harvie.CZ says:

    BTW… On wikipedia i’ve found that SiC can be doped by aluminium… Anyone tried DIY SiC doping using aluminium foil? :-)

    • Tate says:

      not exactly , what you do is add the aluminium while producing as SiC has a very high melting point,
      just add clay ( aluninium silicate) to graphite (coke, charcoal,diamonds or any other carbon source), abd heat them in a steel crucible. Although i dont have enough time to explain this chemically, it should work

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s