I am getting ready to build a new deck and the builders are going to install some lights (mainly so I don’t have to do all the labor!). I am specifying individually addressable LEDs to have some fun with patterns and colors when I’m not planning on doing just normal white light.
The length of the deck railing is about 130ft (40m). At 60 LED/m that’s roughly 2400 LEDs. I did some research on WS2812b/WS2015 and some of the software such as WLED to control them. It seems like I won’t be able to drive more than 800-1000 LEDs at a time with this package with any appreciable framerate, requiring several WLED controllers that are sync’d together. Unfortunately in this configuration, they will all so the same pattern, but you can’t mimic the behavior of one continuous LED strip.
I recently discovered Pixelblaze and the APA102 LED strips. Am I reading the Pixelblaze’s product page correctly in that it can drive 5000 APA102 LEDs in a single continuous strip with a single data line? Or do they need to be broken up into multiple ‘channels’ similar to what I would have to do with the WLED/WS2812b configuration?
A single data wire would be ideal since the installers don’t have much familiarity with digitally-controlled LEDs (they have done several analog projects in the past), and although they know about connecting the strips to power ever couple of meters, dealing with multiple data wires could lead to some complexity.
I’m super excited to have stumbled upon this product, and am excited to have some awesome lights on my new deck!
With any controller, you’ll hit data rate limitations at some point. The WS2812 LEDs and family only work at a data rate around 800kbs, or about 33K pixels a second. If you are driving 2400 pixels of WS2812 on a single line, thats a max frame rate of 13.75. Thats ok for slow moving patterns and fades, but will look choppy for even moderately fast animation.
APA102 type LEDs can run at a much higher data rate, though some can’t run at a high data rate AND long distances, to further complicate things.
Many controllers do cap the total number per channel. Pixelblaze lets you have up to 2500 WS2812 type or 5000 APA102 type on the built in output as a single strip.
There’s also an output expander that can be used instead, which provides 8 channels and up to 800 pixels per channel (and up to 8 expanders can be added for a ton of pixels). Those are mapped back to a linear pixel address space, so you can set it up like a single strip.
Pixelblaze itself can only calculate so many pixels a second as well. Figure on somewhere between 12k-45k depending on pattern complexity.
If you mirror/fork the data, and put the controller in the middle of two strips going outward, you only need half as many “pixels” and can drive the same number of LEDs. The patterns will appear mirrored symmetrically at the center point, but sometimes that can be the right look.
If you want to simplify wiring and get faster FPS, you could install multiple Pixelblaze and then synchronize and control them with Firestorm. It synchronizes the animation of all Pixelblaze on a network, and lets you switch patterns simultaneously. Firestorm runs on a computer, or a raspberry pi and can be access over the network as a web app.
Thanks for the detailed reply!
It sounds like 2400 pixels might be a bit much for anything fun and fast moving. Although in reality I probably wouldn’t do anything beyond some subtle ambient animations, so a lack of a high frame rate might not be necessary. I could also drop down from 2400pixels at 60p/meter to 1200pixels at 30p/meter and that seems like it would be the best of both worlds. A single continuous strand powered from a single controller.
I do like the possibility of splitting thjnfs across two controllers. This would let me split up the power supplies too. Approximating about 750W total, and the 1kW 5Vdc power supplies are over $1000. But the 500W equivalent is less than $200.
With firestorm and two pixel blaze controllers would this behave similar as the earlier scenario you described with the expansion module? Essentially two separate systems that would be mirrored?
Lastly, if the LEDs are going to be mounted under the top railing of the deck and covered by a translucent diffuser, do you think I would notice an appreciable difference between 1200 and 2400 pixels, especially with the higher frame rate for the lower LED count? I’ve heard incredible things about the brightness of the APA102 and equivalent SK-something, so from a lighting standpoint I may be okay with 30p/meter. I realize I won’t light up my back yard, but may be sufficient to cast some light onto the deck surface at night.
Looks like things are (mostly) a done deal! I settled on the 30LED/m of the APA102, resulting in roughly 1200 pixels for my deck lights. I did some rough math and even with the lower density strips the power consumption will be quite large, would be double for 60LED/m!
I just ordered 2 PixelBlazes with expansions (may only need one, but the other will be a backup), and soon will order the LED strips from your store as well. I’m getting excited!