Hi!
So I am totally new here and quite frankly a bit overwhelmed after stumbling on this site while searching out materials for a project I need to build.
In detail - I need to make 4 backdrops ( for a marching band show) that can mimic a lightening effect (show theme is Into the Storm) , be battery powered, and be controllable via an app or remote. And hopefully all without breaking the bank or requiring an electrical engineering degree.
Any direction or suggestion you might have would be incredibly appreciated.
While this set up would be essentially for just a single season use, I could see many of the components being re used down the road for other shows.
My thought was 4 led strips with a lightening effect.
Wow! Yes!! That would be even more perfect than what I had been thinking. Do you know what kind of lights those are? If I bought 4 pixel blazes , can I control them all simultaneously ? Oh… and these could be battery powered? Thank you!
You can do this effect with any of the popular chipsets. 5V options include the beautiful SK9822, or the cheap SK6812, and in 12V I really like the GS8208. 12V systems require less additional wiring, known as “power injection”. You’ll need to decide on a density (common are 30/m, 60/m, and 144/m) and overall geometry/size, then you can figure out how your power will need to work. For most choices except 144/m, the 12V pixels would probably let you power all strips from just one end.
Yes, you can control them all simultaneously if you use 4 Pixelblazes. Just configure sync mode.
FYI the LED wiring topology in Ben’s lightning video linked above is pretty rare - this is a special case where he splits the segments with his own soldered Y connections. Everything after a Y displays the same data. In other words, you can’t control the first pixel after the Y on the left and right independently. This probably works great for lightning that splits; if you want independent control of every pixel, use an output expander to wire things with a “star” topology.
Battery powered is possible but, like power injection, this is another deep field that takes a lot to explain all the decision points. Similar to power injection, if you spend the time to read prior threads in these forums, you’ll find everything you need to know to make this project happen and how to choose everything from the wire size to LiPo charge controllers. If you have a midsize budget or better, off-the-shelf large rechargeable power packs with build in 110 VAC (wall sockets) will be easiest.
Here’s my advice: First commit to a budget, then the overall geometric form which will drive what density you can use and what wiring topology to use. Everything else will follow from there!
Okay, thank you! I am, I think, starting to grasp a little bit more of what this might entail.
I have around $500-600 to spend ( though with some arm twisting might be able to get more if needed.)
I was looking to get 4 pixelblazes - 4 rolls of 5 m (adhesive) of 30/m SK9822 (APA102) RGB LEDs and I found some 12v led battery packs on amazon to power each strip. Am I reaching here as far as power goes? The wiring and other basics of the app I either understand or am confident I can learn quickly. But the power calculation is where I am struggling. Is it a pipe dream to power these with a single battery pack ( per roll) ?
This is the battery pack I was looking at: 12V Battery Pack Rechargeable 5200mAh Lithium Ion Battery for LED Strip/CCTV Camera/Electronic Organ/Optical Network Unit/Router,Portable 12 Volt Battery DC5521 Interface (Blue)
Be careful — the SK9822/APA102 are 5V. This is ok, though! I like to stick to 5V so everything can be USB powered, whether it’s from a phone charger, higher-powered wall wart, computer, or a power bank.
By my accounting you’re well within budget, ordering most parts from https://shop.electromage.com/ . I would recommend the IP67 waterproof LED strips (+$2 each) since the silicone tube protects the electronics from a lot more than just water: dust, impacts, etc. It’s definitely worth it for your marching band which I presume would be outdoors at least some of the time.
Also … with a sensor expansion board, you could easily make the lightning audio-triggered, in sync with the music. You only need one because the Pixelblaze with the sensor board attached can share audio data with the others over WiFi!
4 Pixelblazes: $156
4 5m SK9822 IP67 strips: $196
1 Sensor Expansion board: $29
That brings you to $381, leaving $219 (if you max out your budget) for 4 of something like: PowerCore 20100 - Anker US … each of these would have more than enough capacity to keep 150 LEDs flashing for the length of any marching band performance.
@sorceror’s reccomendation for a 5V system is super solid.
Here’s a little math on the max brightness you can use for such a 5V system. I like to set my brightness limit the max current in an accidental all-white pattern to not melt the connectors.
Micro USB current limit through a PB: 1.8 A
Current: 53 mA per SK9822, so 5 * 30 * .053 = 7.95 A
Brightness limit to use: 1.8 A / 7.95 A = 23%.
Runtime: Since the USB power brick Sorcerer mentioned is 20,100mAh, you’d be able to run an all-white pattern for 20.1/(1.8 + .3 for Pixelblaze) = 9.5 hours before recharging. Most patterns consume half of full-white power or less, so perhaps as high as 20 hours.
You’ll probably be pleasantly surprised by how bright 23% is - it’s about half as bright to human perception as full 100% brightness.
(I can’t officially recommend this until you’re a little more experienced and either use an ammeter or are willing to risk blowing a Pixelblaze or install a fuse, BUT): if you’re careful to only run certain patterns like Lightning ZAP, you’ll probably find you can turn it up to about 50% safely (run it for a while while measuring current or feeling the heat through the connectors). Fire is at your own risk, please be careful.
Current: 12.3 mA per GS8208, so 5 * 30 * .0123 = 1.84 A
Brightness limit to use: 100%
You could even buy the 60/m density (double the number of LEDs for only 35% more $) and run that at 100% brightness.
While the all-white brightness of an SK9822 is slightly more than a GS8208, the fact that you’re running 2X the pixels, and running them at 100% intensity instead of 23% means I’d expect this to be about 4X brighter.
Runtime: Using the 5.2 Ah packs you found: 5.2 / (2*1.84 + .3) = 1.3 hours at full white, about double that for 30/m, and another doubling for average patterns that aren’t full-white.
You’d also need to buy and solder a $4 12V → 5V converter to power each Pixelblaze from the 12V batteries you linked.
Hope this gives you an intuition for LED power math!
This through a PB part is a very important point. I never do that, because I want my PBs to last. I always connect my power source to the LED strip, and feed the PB from that side. The power bank I mentioned can supply up to 4.8 A, so that would allow ~60% full-on power for ~4h. Plenty of room to run things under spec for reliability!
I suppose for a flashy pattern like Lightning ZAP, a capacitor would let the LEDs flash to 100% even with the limited A from such a power bank, and I could probably work out the capacitance required based on this detail from another thread:
All excellent recommendations here! I just want to add that from personal experience, I wouldn’t use less than 60/m density for lightning.
What you can’t see in my IG post is that just off screen is a PB powered over USB from a USB battery pack, all running at 5V. It’s possible, but with some caveats.
The lightning ZAP pattern tends to randomly light up large portions of the strip, which can really spike power demand. In my case, I was running a limited number of pixels, and not at 100% brightness.
These LEDs are now running on my shop ceiling, off of a weaker AC adapter USB power supply, and I have to keep brightness lower on that power supply than I was when I used the USB pack. Your milage may vary, and a lot depends on the equipment!
I would build one set to test before building the other 3 if your timeline allows it.
The 12V + battery route may have less power issues, and you could probably run it at 100%. Throw in a fuse between the battery and the rest of the system, just so you don’t end up with a fire should something go wrong, especially when using something like that 12V lipo battery pack.
The silicone tube LEDs are nice to prevent shorts, which is especially important if you are powering it all from a LiPO pack. They will be a little more of a pain to cut to make many splits, and you’ll want to seal it with RTV silicone. This will give you a much more robust setup than exposed strips.
So I have everything set up and have been able to get all four set of lights/pixelblazes to work … Which is AWESOME!
Appreciate the help from above on this and getting me on the right track with what I needed. The soldering and such was quite straightforward.
QUESTION - since these are supposed to be lightening bolts… Do you recommend the use of 4pin T and L solderless connectors? Or is there a better route to go? Just keep in mind I have 4 sets to make , so soldering individual strip segments might not be feasible.
Also, while I am at it… Is there a way to turn everything ON/OFF with the push of a button? Or am I relegated to using the Brightness Slider?
Those solderless “teeth” style connectors are generally speaking a risky bet for something that will get hard use, be moved around, etc. If you were doing an install that was firmly attached to something that won’t ever be moved / flexed / etc they’re fine, but for your situation, unfortunately, you probably just have to solder on wires for connectors. The problem is the solderless connectors you’re describing will work when you test it stationary on a table, and will fail when it’s moved around by a bunch of people for a marching band show.
To your second question: Setting the brightness to 0% will turn it off visually. You can also attach a button (pads are on the back) which duplicates the functionality of the onboard button – where a single press advances to the next pattern in the sequencer. This can be used to cheat and turn it off visually, by just setting the pattern time in the sequencer to “a very long time” and putting a pattern that’s 100% black (doesn’t light up any pixels) alternating with the pattern that does what you want, so one press swaps to the black pattern, then the next swaps to the real pattern.