Hi all! I’ve worked with LED trips in the past, but always through programs like Touchdesigner controlling the strips via DMX
I recently got recruited to work on a geodesic dome for Burning Man, and after a bit of searching I’m thinking PB is the best controller set up for what I’m hoping to accomplish. Bear with me as I’m still wrapping my head around some of this ins and outs of this project
The goal is to string LEDs on about 120 struts inside the dome, with the total pixel count around 1500; currently I’m planning on using 5v WS2811 “fairy light” style LEDs you can find on Alibaba, with each pixel at .1w
I’m planning on using multiple PBs and possible some Output Expanders; Would I be better served with 4 PBs or 2 PBs and two Output Expanders?
If I’m, trying to create one cohesive array to display patterns on, does the Sync function have any major advantages over Firestorm or vice versa?
Last, I don’t own a PB yet, but how tedious is it to map out 1500ish pixels in a dome shape? Any advice on the most efficient way of accomplishing that sort of a map?
Any pointers or advice is welcome; appreciations in advance!
With 4 PBs you’d have to wire each 1/4th into one long string, if thats easy enough then go with 4. Expanders really help with running multiple outputs, and slightly with data rates, but not as much as another PB. With 4 you’d be limited to 88 FPS due to data rate limits of those LEDs, assuming you spread the 1500 out evenly. Thats pretty high actually, and effects would be really nice and smooth.
For just about anything, the new Sync function is superior to Firestorm. I should add something to the repo mentioning that! There’s way less pattern state management, that is to say zero with the leader/follower sync groups, and it can share sensor board data, and you don’t need a computer somewhere.
The cohesive array/display would be mostly setup in the mapper, which would be the same either way.
If you have some math, or can model it and export coordinates, then its super easy. Otherwise post your dome structure and we can help!
I wanted to pop in to suggest 4 PBs with 4 output expanders. The gained convenience in the wiring for a star network like a geodesic dome is well worth the extra $38, keeping runs short and reliable, and you’ll gain some output speed too.
Excellent that’s what I suspected based on what I’ve read on the documentation and in the forums!
Yes, that’s the ideal setup! I’m doing a little budgeting at the moment to see how much money is available for the project; the camp I’ll be with is pitching in most of the funds, but I agree the expander boards would make wiring the much easier.
The current thinking on wiring is to have a string of 11 to 13 pixels per strut, and connect the various lengths of LEDs with xconnectors or something similar. It’s a lot of soldering, I know, but part of the goal is to try and make the LED setup simple enough to install for someone else if I can’t make it to a future Burn.
Perfect, that’s what I was hoping for!
I don’t have a visual worth sharing of the structure at the moment, but I should have something tonight
That would be a lot easier with expanders I think. Then you can have more components that are identical and assemble in a pattern, making installation easier. Either way you’d need a ton of labeling and documentation.
The blue represents a 13 pixel string, the red represents an 11 pixel string. The two green struts are the door, which should have about 15 pixels on each strut
Have a wiring plan, like how the segments would snake into strings? This 2d view would work quite well. Z could be generated with some math if you want 3d.
I think there must be some folks that have already come up with a strategy for these but I don’t know of any offhand.
Not yet; still awaiting budget confirmation from my camp so I know what hardware we are purchasing; once I have the number of controllers & expanders confirmed, the wiring plan will follow…
I love that thinking! I had been considering wiring based on the pentagons & hexagons, but I’ll map out both once I have the hardware setup confirmed and see what they both look like
Also wanted to add here, since I’ve been down the path several times known as, “let’s have a camp work day where everybody learns how to use lead free solder and cut cut LED strips and solder to the half copper pads left and seal them back up with silicone– it’ll be great community building!”…. More than a few times… and I’ve seen who actually pays the price in the long term (or they become eWaste) – I just have to say I strongly recommend having a contract manufacturer in China execute on a precise cut sheet spreadsheet that assigns a part number to every segment, lists out how many feet of leader cable to your Xconnects, and also has them print the part number on some heat shrink. If you need me to get on the horn with any of your camp leaders about how this extra $400 is the best money they’ll ever ever spend, let me know!
woah, now that is a game changing idea! Heck, I’d pay for this service personally if camp wasn’t interested; $400 is very much worth not having either organize a bunch of soldering parties or having to do it all myself… not to mention the production quality is going to be significantly better start to finish.
@jeff, I’m very curious as to how to start looking and arranging this sort of thing; any pointers?
My favorite vendor is Shiji - I can send you the docs I used for Titanic’s End showing examples and make an intro to a sales rep there. Ray Wu and Gree are other popular options. PM me an email you used with Google Docs!
+1 to @sorceror’s layout (4 latitudinal circles filled in with zig-zag lines). Especially with 3V domes, shared edges make planning pentagon- and hexagon-based layouts really difficult. Gets into this really weird and annoying space of “ok, these strips will all group as a single hexagon, minus this one edge because it’s already accounted for, and plus this other edge because the next hexagon will share it,” etc. etc. etc.
for what it’s worth, I am planning on doing some experimental interactive sensor work! Probably with IR, MMwave, and/or possibly some OpenCV. I played around with OpenCV in grad school about 10 years ago, and poking around now, it looks like it’s come quite a ways!
This year I’m mostly focusing on getting the dome lit up and running with some fun, relaxed patterns (most of the time), with room to play for those who can code. The theme camp I’ll be with is historically an interactive chill hang out camp. 7 Sirens Cove, we’re at 6:30/A for anyone on playa this year
Several years ago at a regional burn, a crew put together a dome with an interior entirely covered in sound reactive LEDs. It was a small/short dome to begin with, plus lots of billowy hanging fabric for diffusion that meant you pretty much just crawled inside and sat/laid down.
Effect was pretty amazing. You enter, everyone is hushed but alert, lots of shushing… then one by one, people would make loud, short, pitched vocalizations to trigger little fireworks above them, until the noise was just goofy and everyone started laughing, and the whole ceiling would erupt in color and light, following by shushing and giggling and the whole dome would go dark and the cycle began again.
A woman brought her cello in and played for us, it was a beautiful moment. Believe the dome made it out to the big burn the following year, haven’t seen it since.
