LED P.O.V. Globe

Ok, Right in with the basics.

A ring of 64 leds, 32 on each side. The leds are spaced approximately 1 led apart with one side being slightly offset.

When the ring spins, the second side of leds will pass through the gaps from the first side, providing an interlaced image and hopefully a resolution of 64 leds high.

The leds are arranged in an 8×8 matrix. The anodes are connected in blocks of 8 with 8wires on ribbon cable, and the cathodes are also connected in different blocks of 8 for the second ribbon cable.

A Max 7221 controller IC connects to the matrix to light the leds in any pattern required via a simple serial interface from the PIC controller.

6 5 4 3 2 1

These first pictures show the plastic ring that forms the base for the globe. This was part of an Airconditioner pipe! I just cut a band off the end. Some holes were drilled around, 32 on each side with a pair for the central shaft, 66 in total.

Next I started to place the leds. The drilled holes were the size of the main part of the led, so they pushed snugly through and stopped at the wider lip on the bottom of them. The Legs bent round to hold them firmly in place while I added more. After 32 were in place a band of hot melt glue was poured around the middle to glue them firmly in place before cutting short the legs on each one.

Each bank of 8 had the anodes connected together, making 8 banks of 8 anodes, and 64 cathodes. Later I wire up the cathodes in banks of 8 too. A ribbon cable was cut to length to allow the connections from the leds and up to the controller that will sit in the middle of the ring. Again hot melt glue was poured over the fixed wires to complete the anode side and protect it from being disturbed during construction or flying apart during use.

113 112 110 109 108 107 104 103 102

There is of course a video of the controller working here


I dropped the idea of the 7221 controller as the PIC can easily control just 64 leds with its own output ports. Only 8 current limiting resistors were needed, I chose surface mount and cleverly hid them on the rear of the PCB. Just cut the track on the board and then bridge the cut with the resistor.

The software is working and addressing all the leds correctly, I only have to mount the ring in the spinner stand now.

The shaft is 2 sections, both conductive metal seperated (and joined as it were) by a plastic spacer. This means I can provide power to the pic via the metal shaft. + on one end and – on the other. Two more plastic spacers glued to the ring hold it all in place securely.

There are lots more photos to go here! No time to upload them all at the moment.
The software has been written to show a test pattern and a simple rig for spinning it to verify operation. Videos of it here, and working in the dark, here.

Finally, I got around to adding the IR pair that allows for detection of spin rate. Each revolution the IR beam is received that allows the pic to sync the image with the start of rotation as well as time the spin so that the image fills the full rotation.

The test pattery is rubbish, but at least its stable and holds its size/position even during variation on the rotation speed. Video here.