After lots of trial and error with buck converters and 3.3 / 5 volt data issues I have a functional circuit for the playa bikes! Thanks for all the help from folks on the fastLED list!! Hope to meet some of you at BM!
anyone have any good recommendations for gluing led strips like the neopixel weatherproof strands to metal? Ive tried jb weld, e6000, and high temperature hot glue all with poor results. I'm thinking of trying gorilla or superglue next.
Very few glues will adhere to the weatherproof silicone sleeve. Using zip ties for a “mechanical” bond is usually faster and easier. The only two reliable glues we’ve found are Permatex 66B Clear RTV Silicone (not all silicone glues will work!) and Loctite Plastics Bonding System, a 2-part cyanoacrylate glue.
+Christopher Arnold I'm running into the same assembly challenges with my recent LED circles project. I actually put it aside for awhile and will hopefully come back to it with a fresh perspective before too long!
Some of you who follow my music production posts will probably know that I've been really excited about this new Blend.io service for music production and online collaboration.
So here's my first Ableton live project, shared out with parts for resampling and remixing. I think this service will help foster more collaborative electronic music production in the same way GitHub has facilitated online code branching and forking of software projects.
I have been struggling with some strange behavior using a teensy 3.1 board and ws2812b 60 LED/meter strips (approx 300 pixels) for a bike project. After much back and forth I've discovered the following behavior. If I'm powering a simple example FastLED rainbow sketch off of USB power from my laptop I can run the sketch for a strand of 30 LEDs fine. When I attempt to use a 5v battery to power the Teensy 3.1 and the LEDs I get no animation or color changes. If I slowly scale the number of LEDs addressed in the sketch to 20 or fewer everything appears to work correctly using USB power or an external battery or power supply.
Finally after many hours of wondering what sort of hardware issues I may be running up against and swapping out several LED strands and teensy 3.1 boards I tried a different sketch that uses the Adafruit_NeoPixel library and went through the same battery of tests. So far I am unable to reproduce this problem with that library so I think it may be a FastLED issue. Has anyone seen behavior like this? I've been working off of the FastLED2.1 branch.
I will also try to finally get this nailed down to work around the voltage issue - this and some intermittent problems people are having with the WS2801 on some platforms, are the only things really keeping me from calling 2.1 done.
Thanks +Robert Atkins. I ordered a bunch of those just in case. I'll try one out next week. I'm still futzing around trying to understand the issue. Doing the 3.3v => 5v logic conversion using the HTC didn't fix the issue. I think it must have something to do with amperage on longer strands. I don't understand why but adding a delay(2000) in the setup() at the beginning of one of my simple hue sketches fixed that sketch up to 255 pixels. Adding delay() calls in various places to slow things down also has some positive effects on a couple of test sketches.
Enclosure for my custom LED Bike controller. Teensy 3.1, Power on/off, momentary press for pattern changes, mic input for audio based patterns, potentiometer for hue/brightness control. The bike will have a strand of 300 ws2812b leds running various fastLED based patterns I've culled from the group.
$16.30 for 5000mah and just add a y-cable and a buck converter for a couple of bucks (or make your own loom). Adafruit have a design using the Tracopower DC::DC converter - I have a couple of them, but for $15, I can buy 7 cheapo transistor-based buck converters from China - or again, make your own with a 25c transistor from TI.
Has anyone else had issues with certain led chipsets when using a teensy 3.1 audio shield? Noticing my fft object is all zeros after instantiating a P9813 strand but if I comment out the addLeds call I see data. I was going to try a 2811 strand tonight to see if it's specific to that type of strand.