guddler said:I get that with common anode, where +ve is the common no connection would be on, not off, but why then is analogWrite(rOUT, 255); (and 0 for blue and green), full red?
trm said:A few people have already worked on decoding the remote. This is one dude's efforts:
http://www.instructables.com/id/Reverse-Engineering-RGB-LED-Bulb-with-IR-remote/
guddler said:Oh, during the "working" period where I was actually quite positive about it, I had a think about what I wanted out of it and really, all I want is what the stock remote gives me.
(SNIP)
just needs a few shortcomings sorting out (only being able to fade down so far for one, bias adjustment for another).
So my test pixel strip turned up. And they are AMAZING! I decided to go for a prewired strip in the end; when I got a ruler out and looked at the pitch of the WS2812B ICs I figured SMT was getting a bit too small for me to happily solder. Which turned out to be very true because I needed to attach the Vss, Vcc and MOSI line and f**ked the first LED with my Red Bull-shakes induced soldering I knocked up a bit of code to learn the FastLED API and will ultimately be snipping these down to make a proto pixel panel, at some point to be replaced with a much larger one that I'll make XY addressable. I really want to code up an old-style Amiga demo plasma for one of these and hang it on the wall. Couple of vids to show my tests. First is a Williams style rainbow blur bar - http://www.youtube.com/watch?v=8AWKC9hHPh0 And the second is a rainbow Cylon I've pretty much chucked my own code out of the window and am working with FastLED (aka FastSPI2). It's got hand crafted assembly to keep things flying along, has reimplemented my clamping colour routines and includes RGB & HSV colourspace functionality. Definitely worth a look IMO https://github.com/FastLED/FastLED/wiki I'm not sure how clear it is from the video but these seem to be 5050 RGB equivalent and even better resistor balanced than my good 5050 strip. I'm using a 30ms delay between updates in the videos. With 60 LEDs on the 1m strip and no delay I can update the strip faster than the LEDs can keep up using a 16MHz AVRMega 1280. So I reckon I'm probably good for about 300-ish WS2812Bs at wire-speed. My desired panel would ideally be about 6-700 pixels so I think I'll have to move to a Teensy 3.1 for that, but we'll see. The strip was £19.99 from eBay and sucks about 1.5A@5V when I've got the full strip cycling rainbows. Need to start looking at building a 20A PSU... Yeah, cat in the first vid. This is the last update I'll post to this OT thread. Follow-ups to http://www.ukvac.com/forum/ot-arduino-and-web-control-for-rgb-lightstrip_topic336496.html |