New here, and have my first PixelBlaze on it’s way!
Have a bit of a project that has been in the works for a while; started it off with an Arduino Uno and HC-05 Bluetooth module, though since I’m quite new to this hobby, I think that the PixelBlaze may be more suited to watch I want to achieve, in terms of ease of setup, and use.
End goal is to power 2x 24 WS2812 LED Neopixel Rings, and 2x 8 WS2812 LED Neopixel strips in series, with the layout as follows:
power/data in > - > O > O > -
(- = Strips)
(O = Rings)
Photo for reference with position of the LEDs that will be behind the dials.
Now, I have also got a 12V to 5V 5A DC-DC converter, as I plan to run this in my car off the battery.
My main question here is whether I will need the step down converter, or whether the PixelBlaze will power all 64 LEDs? And if I will need the use the step down converter to power them, should I just use it to power the LEDs & and PB?
Also, on the coding side of things, I am ultimately wanting the rings, and the strips to have their own ‘sections’ so I can give them individual patterns/colours, though as the most important part is having a single, solid colour for driving, and then worry about patterns, sections, etc later for the likes of car shows. But I’ll be sure to take any thoughts or advice into consideration for when the time comes.
Video of what I managed to achieve with a single ring, on an Arduino Uno.
And here’s the first iteration of the project, which was just a strip of standard non-addressable RGB LEDs, controlled via a pre-made app using bluetooth.
As far as coding the rings and strips separately, I haven’t tried with the pixeblaze. I do have the serial expander that allows you to connect up to 8 strips, but it treats them all as one big strip. If Pixelblaze allows you to set up 4 separate arrays or to treat on array as 4 (using a 4 starting positions and math that uses mod (% for short) to prevent drawing outside the array).
With such short strips of LEDs you probably can get away with using 3.3v powered arduinos, which opens you up to using other boards, but if you ask me the Pixelblaze is a great combination of customizable patterns that has plenty of cool patterns already, and the ease of use of your phone or computer to change patterns. I have been able to do this with bluetooth low energy boards (Adafruit’s feather series) and with Particle Photons and a great app called Blynk, or Particle or Adafruit’s apps. Pixelblaze needs no app, and can run long strips (up to 240 per strip). I have 4x300 led strips connected using the Pixelblaze serial expander board. Only 240 out of the 300 are usable per strip, but a new board is coming out that can do a lot more. If you wanted to do long strips without the Pixelblaze, you would need an Arduino, or a more powerful 5V board, but they don’t have connectivity (BLE or wifi). I like the Adafruit Itsy Bitsy or the pjrc.com Teensy 3.5 for that. Pixelblaze is fit to task better, and I have to tap into that web side of things more, like that iOT post I just saw: https://forum.electromage.com/t/hubitat-elevation-home-automation-integration/231
Have fun!
-Odor Decoder
@Dr.Acula,
Welcome! That looks like an awesome project!
+1 on the buck converter. You can probably run all of it from that.
Theres a thread talking about some methods for doing different things for different sections here:
I wonder what kind of signal the original dash dimming control has (if it has them), maybe there’s a way to interface it with the ADC, and you could keep dimming using the original controls.
There’s also a few folks that are using Pixelblaze for turn signals and that sort of thing, it would be really cool to have everything networked up for shows!
At this stage, I’m not worrying too much about the coding, as I will just get it all running first, then figure it out as I go. I did see another post about lighting up separate sections, so will be keeping an eye on that.
Have seen quite a few different options in terms of controlling the strips, though purely for ease of use, the PixelBlaze is by far my favourite. At least I’ve still got my spare UNO and BT board for my next project, whatever that might be.
Cheers, Ben! Incredible product you have here, and while I haven’t quite received mine yet(it’s in the country, so shouldn’t be more than a day or two away) I can already tell that I’ll be putting a couple more orders through, for other projects!
Thanks for the link, have done a quick skim over the thread, and will definitely give some of the examples that people have posted a go, and work out what works best for me.
Have thought about wiring it through the original dash to try and get dimming to work, though I’m not really fussed at this stage, as I’ll just dim it myself through the code.
Once I’ve finished this project, I’ll be looking into creating halo lights in my headlights with this, that will also allow for turn signals and whatnot to be run through it.
Decided to (terribly)draw up sort of what I’m trying to achieve, and get any feedback on what’s good, what’s not, suggestions, etc.
Now I have read that it’s worth chucking a cap that’s between 100µF and 1000µF going from power to ground to smooth out the power, as well as putting a resistor that’s between 220Ω and 470Ω on the line between DAT out on the Pixelblaze, and the DI(data in) on the first strip, so as to protect the data pin.
The buck converter I have to power the lot is a 12V to 5V 5A, so if my calculations are correct, with the LEDs using at most 60mA each(at full brightness) the LEDs could use up to 3840mA, leaving what looks to be enough headroom for the Pixelblaze, though I’m unsure exactly how much power that uses by itself. I of course don’t plan on having them run anywhere near full brightness, as even in my first test with a single ring, 10% brightness was probably enough.
Sorry about the shit drawing, it’s a very quick, rough sketch, ok
Caps can help in some cases, but I wouldn’t add 1000uF right off the bat, your buck converter probably has a decent cap internally, and you aren’t running wire a great distance (in other words, you aren’t fighting massive inductance or resistance).
Pixelblaze has a 100ohm resistor on the output already. The recommendation for resistors is for impedance matching for longer runs, and to ensure that you don’t accidentally power the LEDs via the data pin if power isn’t connected for some reason. You should be OK without.
Keep in mind many buck converters will experience voltage drop before their rated current. You should be OK though.
Only minor thing I noticed, probably just an oops on the drawing, both rings don’t have a line for ground. I’m not really sure what the second G on your strips are, maybe a redundant ground? If so, you are probably fine connecting one or both, whatever’s easiest.
Caps can help in some cases, but I wouldn’t add 1000uF right off the bat, your buck converter probably has a decent cap internally, and you aren’t running wire a great distance (in other words, you aren’t fighting massive inductance or resistance).
Pixelblaze has a 100ohm resistor on the output already. The recommendation for resistors is for impedance matching for longer runs, and to ensure that you don’t accidentally power the LEDs via the data pin if power isn’t connected for some reason. You should be OK without.
That’s good to hear that the Pixelblaze has it’s own resistor built in already, saves me a bit of hassle. And I think you’re right about my buck converter already having a cap.
Only minor thing I noticed, probably just an oops on the drawing, both rings don’t have a line for ground. I’m not really sure what the second G on your strips are, maybe a redundant ground? If so, you are probably fine connecting one or both, whatever’s easiest.
Oh, yes, that’s just an oops on my part, the rings should both have a line for a ground. Also, the second ground line on the first strip was an oops too, though by the looks of it, it may be more for redundancy anyway; they’re cheap enough as it is, and I have a local seller for them, so I’ll run it with 1 for now and see how it goes.
A quick question you, or someone else in the forum may be able to answer; due to space constraints with the area I’m working on(behind the dash) I’m looking into having the Buck Converter and Pixelblaze in a remote location, and have a USB cable to connect them and the cluster LEDs together.
Now, my question is, what would be the max amount of power that a standard USB Micro cable could carry safely? I’ve worked out that at full white output, the LEDs can draw 3840mA, though I’ll probably cap the brightness at 50%(1920mA?) and I’d probably never run full white anyway, so it’ll be another percentage drop from that 50%.
I’ll mostly be running either 1 or 2 static colours, and have the occasional pattern going that would obviously draw a lot less power.
I’ll also be mounting a potentiometer to my dash somewhere to control brightness(min 5%, max 50%) which I’ll wire in separately at some point.
Here’s another crappy drawing to hopefully get my idea across
The USB connector on Pixelblaze is rated for 1.8A. USB isn’t designed for high current applications. It would probably work, especially if you never run white, but since this is in your car and wiring isn’t easy accessible, I’d advise keeping things within specs for worse case scenarios to be safe. There are circumstances that could cause the LEDs to go all white, like wrong settings, or a loose data wire, and that sort of thing.
There are other pluggable connectors that can be used as well, just be sure to check the current ratings. There are fairly inexpensive 3+ wire automotive connectors that would work (like for trailers) that are usually rated for a lot more than you need.
Another thing to consider is the wire size. I consult the AWG chart to figure out what wire size I need (using chassis wiring column):
I wouldn’t be using the USB connector on the Pixelblaze itself, just a separate breakout board wired to the buck connector for 5v/ground, and wired to the Pixelblaze for data.
I have read a bit of info about USB only being rated up to around 1.5-2A, though thinking about it, is that due to the individual wire size? As I just had a thought that I could perhaps split the 5v at the first breakout board, and essentially use 2 of the wires in the USB cable to carry the power, then just have 2 separately powered LED sections(1 ring & 1 strip per wire) which if it would work out, would basically be able to supply 960mA per section of LEDs; or is the actual 5v wire in a USB cable a different size, and specific for carrying that 5v?
So this is the Type A breakout board I planned on using on the Pixelblaze, buck converter side, so was thinking if I wire the positive to the VBUS & the D- terminal(then data to D+ & ground to GND) and continue the split on the cluster side for the 2 sections of LEDs. Would this work? Or is it the entirety of the USB cable only rated up to that 1.8A?
A quick read of that Wire Size link you post, and some more research points towards the individual wires in a USB cable being around 24-28AWG each, which even a data wire, that could be as small as 28AWG can carry 1.4A, so splitting it, in theory, would be safe to supply my approximated draw of 960mA.
Please forgive my ignorance in all of this, I don’t have the most knowledge when it comes to these sorts of things; it’s mostly just research & tinkering in my spare time haha.
Right, I read that then forgot when I started writing about USB stuff But in general USB connectors aren’t rated for high current.
If you do borrow d+ for 5v to increase current capacity, the current has to return, so if you don’t also use d- for GND, GND will have the sum of both of the other wires (weakest link, a circuit includes return paths).
It will work of course, and normally it won’t be a problem, but if you pull theoretical max things will be in a state of overload, which could mean heat, which can sometimes mean fire. I don’t want to recommend anything that has even a remote possibility of fire.
Thats why I’d recommend a different kind of connector. There are many pluggable connectors, and a screw terminal and some wire can work as well, though installing it would be less easy in tight spaces. The automotive stuff is usually pretty solid, high current, designed to handle vibration, heat, and be reliable. Something like this might work well:
