I’ve worked with a few different addressable LED strips, but never the quantity of a project suggested by our school’s track coach: making a “rabbit” for a track. That’s a strip of LEDs all of the way around, with a few LEDs lighting up and traveling around the track at a user-designated speed, for pacing. Lighting them all would certainly be a problem with power, but this would only light a few at a time. Teensy is probably needed to keep up with the memory requirements (is flash memory the right spec to look at?) - say 30 LED/m, for 1600 m = close to 50k LEDs. Is this remotely possible?
Better choice IHO… alternate 1m present with 2m missing. 1ea pro mini with rf trigger … you could make each seperate and battery (solar recharged) driver…
This is worth reading. Might also be worth exploring the Teensy 3.5/3.6 but they’re not officially supported here yet. Another viable alternative (in my opinion) is a number of ESP8266s.
I’d seriously consider the spacing on the LEDs. 10 per meter would bring the number required down to 4000 and make the whole thing more manageable. I’d lay some LEDs out and see what different spacings look like (easily done on software), get them as far apart as possible, even if you use led pixels. There’s loads of waterproof (probably) ones on AliExpress, and they’re often 12v which is better over distances.
I agree with some of the thoughts and ideas Jeremy is putting forth. I was wondering if a cluster of pixels every 1/2 meter or 1 meter would be an acceptable spacing. At every 1/2 meter that would be 800 “points” around the track so way more manageable. If there were 4 pixels at every 1/2 meter point then 3200 pixels. I’m thinking not a flat strip, but something with wire between points.
If you were buying 400 meters then one of the led manufactures on http://aliexpress.com can probably do some sort of custom run for you.
Power will certainly be something you’ll need to think about. Going with 12V could make things a easier.
Would these leds be periodically rolled out/rolled up, or something that would be left out for months at a time or permanently installed? Cable management, fixturing, water proofing, and durability will be a challenge.
Other thoughts, will they be bright enough if track practice takes place during the bright day time. Does it need to be RGB (obviously that’s the coolest option of course!), but would a mono-color work/keep costs lower?
This sounds like an interesting/challenging project. Let us know how things go.
I think that, for pacing alone, every 50 cm or so would be perfectly fine, but I walk also thinking about applying it to striding, since he says that over-striding is an issue for lots of kids. If they’re maybe 10 cm apart, then we could show individual footfalls at the right pace and location. I only picked 30/meter because that’s what I could find available online.
It would be a permanent installation; there’s power available, but waterproofing is something to look at. It’d be on the inside rail, so it shouldn’t be stepped on regularly, but definitely would need a rigid clear guard.
I haven’t looked at them outside yet; I know that the strips that I’ve used are uncomfortably bright at full intensity, but I’d have to try it out.
With the 30LED/m 12V WS2811 variety, you’re addressing groups of three RGB modules as a single pixel – in other words, you only have 10px/m. This reduces your addressable pixels for a 400m length to just 4,000px, which is easily handled by a Teensy 3.1 with its 64kB of SRAM.
The real problem you’re going to face is communicating to those LEDs. If you just use the entire string of LEDs as one big chain, a failure anywhere will prevent the rest of the lights from working! The 5V signalling level used for most addressable LEDs doesnt work well over long distances (“long” here being more than around 10m using the SN74HCT245 buffer in the Teensy OctoWS2811 board to push the signal over Cat5). Some solutions you could look at are:
Using a differential signalling method like RS-422 or RS-485 to transmit the LED signal most of the way, then converting back to the 5V level at the start of each LED strip,
Designing a bunch of small controllers, each of which handles only a few hundred LEDs, and sending a radio pulse to synchronise them (this is complicated, but fun if you’re up for the challenge)
Use a boatload of repeater buffers (probably one every 7-8m to be safe) to reinterpret and boost the signal as it propagates along the wire (the signal lines will then require power for the repeaters).
@Jeremy_Spencer Just to add, I’ve been using Teensy 3.6 with great success with FastLED (and parallel output PORTDC 12x and PORTB 8x) using the kcouck fork :
