72 LED Persistence Of Vision Globe


[Ben] told us about his POV globe yesterday. We took a look and saw just one photo and the code with no real explanation of his project. He certainly set to work over night and now we see all the goodies we look for in a great build log. He even threw the Hackaday logo up for our enjoyment. His build is well executed and he found some creative ways around the common problems in these projects. We take a closer look after the break.


[Ben’s] design looks much like an actual globe, with a base, a frame, and a rotating ring mounted with its axis tilted (just like the earth). 72 surface mount LEDs are used for the display, a PC fan with the blades removed provides the rotation, and a reed switch in conjunction with a magnet is used to sync the rotation with the display parsing.


Surface mount components are meant to be placed and reflowed on a circuit board. They are usually considered too small for free-form circuits. [Ben] made this process work for him by lining up all 72 LEDs face down on the sticky side of some masking tape. This made it much easier for him to solder together the multiplex needed to interface them with the microcontroller. You can see he is using varnished wire that can be soldered directly without the need for stripping. [Ben’s] using a slice of a clear DVD-R container cover as the rotating ring for the display. On the right side of the picture above you can see the completed LED multiplex installed in this transparent ring.


Getting power to the rotating part of a POV display is always an issue to consider. [Ben] tried feeding the connection through a brush motor but had problems with power loss. His next attempt involves using a headphone jack and connector as the pivot point. Regulated power and ground are passed through two of the connections and he’s had great success with this system. Above, you can see the connector before it is fully inserted in the jack on the protoboard.


A reed switch is glued to the transparent ring and is actuated when it passes a bar magnet on the frame. This allows the microcontroller to measure the rotation of the ring, and sync the display output.

[Ben’s] done a great job here. He salvaged a lot of the parts, including the LEDs. He’s chosen an Atmel AVR ATmega8 for the microcontroller. This is an inexpensive and readily available chip that is pin compatible with the more powerful ATmega168 so there is a possibility of future upgrade if animations or other functionality are desired, requiring more programming space. We would suggest a decoupling capacitor on the power pins of the uC in order to help filter out any noise on the lines, especially considering the rotating connection used to provide the regulated power.


34 thoughts on “72 LED Persistence Of Vision Globe

  1. Wonderful idea that audio jack power xfer point.
    I was forced to abandon my POV design (just a line of LEDs going in a circle) because of power transfer difficulties to the graphics driver circuit.
    It seems now this fantastic design has reinvigorated my desire to take another look, in fact this design seems to indicate a globe, instead of a simple ring, is more feasible!
    The best ideas are those that you see are incredibly simple but still ingenious!
    But isn’t there a certain amount of ‘rubbing’ which generates a squeak of metal on metal, or will eventually result in the connector wearing out? If this is an issue, would a spot of WD40 prevent the power crossing over to the protoboard?

    Thanks Ben!

  2. Stu: There will be a certain amount of rubbing, and yes the connector will rub out eventually. Eventually being after we are all dead…

    Slip rings (which do the same thing) are very similar in their design… This is a really neat implementation.

  3. Hi, thanks gys, yes I did put a bit of WD40 on it to “loosen it up” I also slackened of the sping connection points inside the socket (the pc fan is quite weak, and not up to rotating with much resistance).

    Amazingly the hole thing produces VERY little sound. probably less than the PC fan used to.

  4. Thanks Ben-
    Loosen off the spring connection in the socket – Gotcha, good to know.
    In fact I’ll probably be trying it out with a Tamiya RC car motor instead, I have an old PC case fan I could use but its best not to rip these things apart if you dont need to.
    Yes indeedy, would love to see a video of it in action, but invariably a lot of cams dont catch the effect too well, all those vids on YT are testament to that. Still, worth a look.

  5. Wow, this is super goodness, some great problem solving in this project. I can’t wait for this effect/tech to evolve a bit though, then maybe we’ll see an RGB version w/ head tracking :P.

  6. I did try making a few video’s but with minimal results they were very dark and the frame rates caused real problems, I will try again and update the project pages when I get a good one done.

  7. As I understand it the LEDs are not individually addressable at all times and the display is multiplexed in banks of 8. Is that right? Doesn’t this multiplexion introduce delays which skew the image?

  8. I’m with Andrew, WD-40 is a temporary solution when there’s no actual lube around. His suggestions work, and so does silicon lube as well, it’s sold as “household” everything lube.

    That is clever though, the whole thing is.
    Won’t the fan motor heat up quite a bit with no moving air?

  9. I absolutely love this. It shows really well what can be achieved using brains instead of money.

    I was surprised to see that a simple reed contact was fast enough for synchronisation. I thought a hall sensor was mandatory.

    Now, were is this old CD container…

  10. That’s a mighty fine project there, Ben!

    My $0.02 worth of concerns/questions:

    1. Reed switch will quickly become magnetized and always “on”. Expect to replace it quite soon. Maybe an optical solution is in order? (Look for the HaD article about scavenging old ball-mice.)

    2. Glad to hear WD40 isn’t your perm. lube solution. I third Andrew and _matt’s suggestions of graphite/petrolatum/silicone.

    3. Any plans on sending real-time graphics data over the third pin of the phono plug?

    4. You might think about using thumbnails for those 2 MP images (for those with slower connections/PCs).

  11. the audiojack power line is so clever, why havnt anyone thought of that before? i wonder how long it will last though. but as said in the article, a cap might be a good idea, especially when the audio jack starts to get worn.

  12. Two things to add. In your orbiting ring why not build in an inductor coil with your existing copper? Stationary mags on the out side providing alternating current. Pop in a wifi controller with arduino and bam, wireless connection to device while also maintaining wireless power. No friction on the power induction surface (that is, physcial friction from surface to surface).

    1. …And I’m dumb because I didn’t see the mag part. Good plan, thinking the wireless arduino will set you up. Other wise if you are thinking unconventional, I’d suggest ball or pin style bearings. They help reduce friction and can also carry current though physical connection. Could always lube with ferrofluid.

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.