Since I discovered that no AVR boards will ever work with WS2811/WS2812B LEDs[1], where shall I address my RGB adressable need? My project consist in merging MPU-6050 IMU (which is communicating with I2C and is using interrupts), MSGEQ7 for audio spectrum analysis and LED animation representing the two different data
If you already have the WS2812’s, you can use OctoWS2811 as a driver for streaming the LED output without blocking interrupts, and still enjoy all the awesomeness that is FastLED. OctoWS2811 only runs on Teensy 3.1 [full disclosure - I’m the guy behind it], but if you’ve already bought a lot of LEDs, that’s a cheap way to make use of them. If you haven’t bought LEDs yet, APA102 gives you a lot more options. But if your project has a lot of LEDs and you don’t need persistence-of-vision PWM speed inside the LEDs, the extra expense for a Teensy & Octo driver board might be worth the savings for less expensive LEDs. Options are good.
Hi Paul (I know who you are :)).
Options are always good.
I appreciate your comment, and I may follow your advice and finally give Teensy 3.1 a try. The point is that I want this project to run on any strip Fastled is compliant to. The project is sorta of a kit.
So I guess I’m taking the MPU-6050 out and use a simple gyro.
I just ordered some APA102’s from Ray Wu on Ali Express based on positive feedback I read on here. I haven’t received them but tracking says they left hong kong today. The price was amazing, $19 or $20 per meter IIRC.
What a coincidence: I have a couple 6050s, 1 msgeq7, and a bunch of LEDs sitting in a project box right here!
I’m designing a handheld ‘light-show’ device for a techno/club-loving friend. It will push light patterns through a 10" fiber-optic ‘tuft’ (like a floppy paint brush) and change in relation to its position/velocity/accel, combined with the timbre/tempo of music (that’s where the eq7 comes in). I experimented with the idea last fall, but since then other projects have taken priority. It’s good to know someone else is working on the exact same set of hardware - I’ll have to make time to work on this project now!
For my prototype I used NeoPixels, simply because they made it easy to run on 1 LIPO. Now that I’m more experienced, I may switch to APA102s - they’re better for POV effects. On the other hand, my device needs to fit in a small flashlight, and a NeoPxiel ring makes this easy.
The hardware side of development is relatively new to me, so soldering up a handheld device (the only way to get real-world data from the 6050) that won’t fall apart and can emit usable debug-data has been the biggest challenge… so far I think I’ve just scraped the surface of what the 6050 can do and I haven’t wired up the msgeq7 at all. Nonetheless, I’d love to touch base with you and share tips, tricks, or ideas.
Keep up the good work!
-Matt
PS: For a long time, I thought the timing requirements of the 2812s were causing 6050 buffer overruns. Recently, I learned that some 6050 libs are built with incorrect interrupt bitmasks - I thought I was responding to ‘data ready’ but I was actually responding to ‘overrun!’