Lux Capacitor is an industrial-scale solar light, designed to illuminate art. V1.1 shown below.
Designed by Glass House Arts, Lux Capacitor won a Regional Civic Activation grant to refine and open source the design in December, 2025.
To collaborate, be an alpha tester, or get updates about this project send us an email at art@deldiosglasshouse.com
For our art, check out https://www.deldiosglasshouse.com/
This is a copy of the working doc Colin is writing to solicit feedback! Feel free to add comments / edit / suggest alternatives big or small.
This seems like a great idea, but I've never seen them around (commercially or custom), which gives me pause. Would love anything from detailed feedback (lighting controllers, material thickness, etc.) to broad ideas (different shape, buy vs build, etc).
Flood lighting in one direction.
Reusable, repairable, designed for the long haul.
Off grid, able to do 8 hours at full brightness and handle one cloudy day.
Typically warm white but more features ok.
Safe to be walked / biked on.
Ground mounted.
Robust to dust / rain / heat / transport.
Does not need to be opened on site.
Specifically for art illumination, though also safety and construction
Now that we have a working version, the next step is to open source it. Targeting CC BY 4.0 for docs/CAD, MIT for any code to be permissive with attribution. CAD will be Onshape (browser) parametric model + exported DXFs and printable templates.
Two build paths will be supported:
Hand tools: shear/snips + 5″ vise brake + rivets/screws
Service: SendCutSend + rivets/screws
Light module: designed around the off-the-shelf Hampton Bay 120 lm spotlight with replaceable 18500s, but with support for other lights also.
While we could just put the current files and instructions online, we think that widespread adoption depends on completing the following steps:
Design for simpler fabrication: Lux Capacitor V1 was designed around tools at the Glass House Arts studio. To facilitate fabrication of Lux Capacitor in a garage with under $100 in specialty tools we need to make some changes:
Expand cutting options to include CNC plasma, water jet, fiber laser, plus printable template with hand tools and also services (SendCutSend or similar).
Modify the bends to make it easy to make with a 5” vice brake, which is widely available for under $50.
Redesign the parts to use tabs with pop rivets or machine screws rather than welding.
Parametric CAD: Currently designed around one light option, but a parametric CAD file would allow folks to modify the design to fit whatever solar light fits their needs best. Could be done in free, browser-based CAD so people with essentially zero CAD experience could make the changes.
Testing: We should do a run of prototypes, including cutting by hand, using a service, and testing different materials (different thicknesses of steel, maybe experimenting with aluminum).
Documentation: In addition to the needed files, we will set up a website with detailed step-by-step instructions and also a Discord for discussion with other Lux Capacitor builders.
Alpha rollout: Glass House Arts co-founder MJ is on the production staff for Youtopia and LoveBurn, and is a volunteer for SOAK and Saguaro Man. We will use these connections to reach out to artists that might benefit from Lux Capacitor, and offer a modest subsidy for being part of the alpha rollout, adding to the documentation, and answering questions that come later.
October 2025 we are submitting to the Burners Without Borders Regional Event Grant. https://burnerswithoutborders.org/uncategorized/2025-regional-event-grant/
Lux Capacitor seems like a good fit, since it is open source experimental infrastructure, focused on sustainability, aimed at regional burns but suitable for other events (even neighborhoods!).
We asked them for $500 for the following:
$50 for four different solar lights to test out the parametric design.
$100 for more steel and aluminum for doing test runs with tabs, parametric designs, different materials and thicknesses (we buy used metal at wholesale prices).
$150 for small run with SendCutSend or similar online cutting service, to validate that it is a suitable approach (I’ve never used them, but hear good things, and this could be ideal for folks with more money than time).
$200 for four $50 grants to alpha testers. These will be given to artists bringing art to a regional burn or similar event, who Glass House Arts does not have a relationship with. The reason for the money is to de-risk the up-front costs, and encourage documentation and social proof.
All labor, shop time, hardware, web hosting, etc. will be donated. We could also do this without the alpha grants, but it might just be slower to get started.
V1 was pretty great, but making a few tweaks would really improve the design.
Smaller flanges on sides
Move anchors back (tipping moment is forward)
Move bolts closer to panel (backing plate bowed)
Narrow front opening to restrict side-to-side motion of light.
Also used an oil-based paint with an automotive sprayer.
We made 4 of V1.1, and used them to light our art “Uplift” at Burning Man 2025. Zero breakage, good light all night. Fuck yeah.
Compared to the rest of Burning Man the lights were somewhat dim, but totally sufficient for safety and for folks to enjoy the art. The rain + playa dust did cause issues with the oil-based paint, and we should explore other surface finishes.
Solar light: $13
4 ¼-20 bolts and nuts for housing: $1
Steel: $6 (from our local scrap yard, $1.07/lb, price accounts for waste in nesting)
Paint, welding wire, etc: couple of dollars?
Two lag screws: $10
TOTAL: $30-$35
If we had to buy a few tools (pop riveter, vice brake), and pro-rated the cost over the parts, plus used a service for cutting, the cost might be more like $60 to $75 per light if we made 5 (including prorated tool costs).
Lux Capacitors were deployed to light two projects at Youtopia, the San Diego regional burn. Both Nipplidoscope and Family Can Be Murder used them as safety lighting to supplement the interactive lighting, to great success. We did find that if used in the white setting on the longer nights/shorter days we ran out of power around 4 AM on the lights where the panel faced north. Solutions were to face them south or change to a color setting that uses much less power.
Starting with a solar light from Home Depot is way cheaper and easier! Then we just design an enclosure.
Hampton Bay makes a $13 light that does warm white and a few colors, has replaceable 18500 cells (x2) and 120 lumens. Link
Designed for the Glass House Arts plasma cutter, and welded together. Testing indicates it can run 10+ hours on the white setting with a full day of sun (yay!).
Decided to go a different way.
This is the best I could get AI to do - ignore proportions/details
Links are to Amazon because it is easy - will shop later.
Custom, plasma cut and welded - maybe in aluminum? Roughly 18” x 18” x 3”, with all sides at about 45 degrees or a little less steep.
Solar panel mount is tough because of heat, fragility, dust. Options considered: plastic cover, potting cells, grate, vertical,
LEDs are covered in clear plastic (⅜ polycarbonate?).
Holes on the corners (or tabs) so it can be lagged down (or screwed for other orientations). Weather stripping all the way around + closing hatch on bottom so it is sealed from the elements inside.
Adding some slope (even five degrees) to the panel will help with water and dust.
LEDs: Maybe just strips of WS2811 with ip6x rating (example), stuck to an aluminum sheet inside. The COB ones look rad, but I can’t find weather proof ones. Aiming for 5-10 Watts at full bright white.
Controller: ESP32 with WLED seems like the right approach. I’m currently leaning toward an integrated version so we don’t have to deal with an enclosure, level shifter, buck down, capacitor and resistor on LED data line, etc.
Getting them to coordinate is a neat idea, and I think we can do it if we have a WiFi router (no internet needed, just network).
Alternatives: Could just go warm white if we don’t want future flexibility, but I think it is worth it. Some folks balk at working under RGB lights, they are a bit less pleasant and can cause visual distortion (multiple shadows, etc) and odd color rendering.
Maybe 12 V 10 Ah LiFePO4 battery as the core (example). A flat pack of 18650 cells would fit better perhaps.
The desired 18x18 max size limits us to 20-30 Watts of solar panel.
For the solar controller I really want MPPT, waterproof, load (capable with day/night), and LiFePO4 compatible. Haven’t found one yet!
Victron Energy SmartSolar is the best I have found (ip32 which is ok). But it is $62! It does have bluetooth, which might make debugging easier (don’t have to open the box to see charge).
Tempted to say “fuck it” and get a $27 20W panel with PWM controller - we can always upgrade later.
There’s a line of very low cost MPPT bare-board controllers from China like this one. Would need housing but could be sorted out in the integration.
Inline fuses will be added where needed - probably ought to store a spare onboard.
Solar controller: Is there an affordable solar controller that is MPTT, waterproof, LiFePO4 compatible, and has day/night load?
LED controller: What’s the best controller for the LEDs? Affordable, robust, waterproof, sync between devices, easy to update while camping.
Thermal management: Operating in full sun on the playa means internal temps could get dangerously high.
Mitigation ideas under consideration: Elevate solar panel slightly for airflow underneath, add heat sink fins, couple to earth.
Pretty worried about the plastic over solar. Maybe pot raw cells in epoxy? Go more vertical? This one is hard.
Wi-Fi sync reliability: Can WLED sync robustly in a dusty, low-signal, multi-node environment with a 12 V router?
Lighting brightness vs runtime: 5–10 W may be too high for all-night operation with one cloudy day. Need testing.
Ingress protection: Full sealing is planned, but fine playa dust may still creep in. Any need for breathable membrane or pressure relief?
Light positioning: The higher the light is, the better it illuminates things. But then it can’t be walk-over-able. Maybe two physical modes? Something that rises up optionally? Lights that are spring-loaded so they fold down when they are run over?
How deep to design for? If we require digging down 2 inches (maybe with a template) then that opens up a lot of design space.
Victron stuff is mostly though dealers. Might be able to work with https://www.infinitervmarine.com/victron-energy in San Marcos, maybe get panels and controllers?
ECO-WORTHY 25 W might be a good alternative panel.
This is a hard design space - the heat, size, price, etc. Alternatives:
Smaller, off the shelf: smaller, combo of driveway lights (or these) and yard solar lights (or these, this package is a little weird, but great reviews). The very low cost of these means we could do 10+ easily - more robust to failure.
Design: simple folded sheet that has both a yard light (flood facing art) and one of the driveway lights (ambient / safety / secondary). At $10 each we can get these super cheap. Make it laggable, open sides for ventilation (since already waterproof).
Could also do an upwash version with these. Just put 3 or 7 in a plate, lag it down.
Bigger: 100 watt panel is more economical (fewer controllers, panels are common), and also much easier for heat. Maybe a bench with lights on either end, facing the art? Solar as seat (under plex / grate), maybe some soft down lighting too.
No solar: batteries are getting way cheap. We could just go with 100 Ah of LifePO4 and have it last ten+ days. Much simpler, about the same price, more weight and volume, much less fragile and location-dependent.
