A few weeks ago, we featured this water-based LED graffiti art installation that allows anyone to paint in light using only a bottle of water. When one of [Chris]’ friends saw the video of this build, he immediately asked him how it worked. One thing led to another, and now [Chris] and a few other members at the BUILDS hackerspace at Boston University are building their own water LED installation.
The basic premise of this build is allowing water to serve as a conductor between the anode and cathode of a LED. Without spraying or painting water on the circuit [Chris] whipped up, there is an infinite resistance between the two pins of the LED and current cannot flow. After applying water to the anode and cathode pads, a small amount of current is conducted through the water and the LED lights up.
Right now, [Chris] is working on a test board with different sizes of pads and spacing to get the best water graffiti LED effect for his future build. The plan is to build a single one-meter panel out of one hundred 10 cm x 10 cm boards connected together with jumpers.
All of [Chris]’ work is up on GitHub, and even though [Chris] hasn’t begun designing the production boards, it’s more than enough to get you started if you’d like your own water LED painting panel.
Not really…. The water doesn’t bridge the anode/cathode of the LED (the junction takes care of that). The water acts as a large-value resistor, connecting power to the LED.
In their diagram, the water bridges the Cathode of the LED to ground. (the anode is permanently connected to power).
Dan is correct.
The water does not bridge from the Anode to the Cathode. It should never do this in fact as there should be a water tight seal so that water can’t get onto the back panel.
Bridging the Anode to Power is functionally identical to bridging the cathode to ground with water (a large value resistor).
I think that a bit of water behind the panel shouldn’t be the end of the world (depending on the spacing and orientation of your design) because water should easily bridge the front pads while it should be much harder to bridge the leads behind everything (especially for any amount of time, and even if you do there’ll be much more resistance between those pins than the path to ground up front so it should still work)
Yep, a tad of water shorting VCC to GND or across the diode should have no ill effects. It would essentially be like adding a very large resistor, which would essentially have no real effect.
I would still take precautions and fuse your power supply if you can.
Actually, a bit of water between the anode and cathode of the LEDs can give rise to electromigration of the tin plating on the leads of the LED – eventually shorting the LED out.
I have experienced that with a 8×8(x5 LEDs) matrix made from cheap yellow xmas lights.
After a couple of days in the rain some of the LEDs went dark, but the other in the string worked fine. When I later debugged this problem thinking the LEDs were burnt out and shorted, I found that there was formed a conductive path between the leads of the defect LEDs – after removing that with a piece of tissue paper, the matrix worked again.
The milliamp currents required to light a LED will quickly corrode the pads. Adding a transistor per LED would make for more sensitive input, and less electrolytic corrosion.
There’s also the real possibility that dirt in the water, or the dissolved copper will bridge a pad and cause a short, which will burn the circuit.
A smarter solution would be to coat the pads in laquer, and use high frequency AC which will jump the gap once water is introduced, amplified by a transistor.
Needs more current limiting than just the water as a resistor, too easy for someone experimenting with how it works to pop an LED (for example by holding a key across the contact pads)
or you could have a million current regulators instead of built in resistors, same current regardless of white led, red led, or short-cuircuit! :)
cool and all, but personally i’d rather it redesigned for a rain-“art” installation, turn the thing flat on the ground and turn each sensor into a small pyrymid and have mini “gutters” flowing between rows and colums of sensors… make the whole thing (semi-)transparent and voila! raindrop == flash of light :)
I too thought it might be nice for a garden thing when it rains, although your thoughts were a bit more elaborate though.
And it might also work nice on advertising signs actually, put it in britain where it rains a lot.
I am a big fan of this idea! I have been trying to figure out some more ways of doing it such that it was not just a clone. That sounds really cool. I might actually talk to the gardening club at BU about installing some water sensors and maybe doing some art in their garden/green house now…
Argh that boolean expression drives me nuts :P
I’m thinking of what might be best for the contacts, because copper will not last long of course, so I think aluminium might be more up to the task, rather than copper, even if you tin it.
But how to bind it to a waterproof surface? You could use the sticky airco aluminium tape I suppose, but then how do you reliably connect it to the LED?
What are your thoughts on that people?
Polish it once in awhile?
Unfortunately I think you will find a layer of non conductive aluminum oxide will form within seconds with contact to air.. best option here would be a thin layer of gold plating I think, most PCB places would be able to do this I think.
Have you seen he gold prices? Even plating (which is extremely thin) is now not done as much anymore.
Also – I think oxidized aluminium isn’t that extreme in its non-conductiveness
I just tested an aluminium bar that is attached to my bicycle for several years now, and it’s been in the rain and snow plenty of times and I can report conductivity is excellent. It’s a bog standard bit of aluminium from the DiY store that I also polished before installing it way back to get some scratches off, so there is no layer of any kind on it apart from the natural oxidation layer aluminium has, pure aluminium.
Is copper oxyde conductive or not? if it is that’s not really an issue, it’ll just looks green!
Copper slowly erodes, especially in the rain.
Bud copper oxyde is protective, just look at the eiffel tower.
The issue would be to know if the copper oxyde is conductive enough to keep the system working and that’s not easy to find :-/
Solution would be coating with gold/silver; there’s some electric pen coating kits that could do the job; with silver it stays “reasonable” for a ultra slim coating
You can’t compare thick copper to microns thick copper on PCB’s though.
And I think you mean the statue of liberty, since that’s copper while the eifel tower is puddle-iron
https://en.wikipedia.org/wiki/Eifel_Tower#Material
I also looked up the statue of liberty and notice something interesting about the thickness of the copper, it’s only 094 inches (2.4 mm) thick!
Now all we need is a new-yorker with a multimeter to go measure conductivity :)
Hello,
I can’t find how this circuit can work :
– on the back you have the + VCC linked to all the leds
– on the front you have the ground
– wate allows ground between main front circuit and the “central ring” of each led
But how do you connect the – from the led to that “central ring” ?
I can’t find any picture showing that, even on the release video of the french project; leds are all soldered on back with led being “flat” & “sealed” but where do you make that damn link from front ground to led cathode ? :'(
When you zoom in on the back of the PCB you’ll see that the copper where the – of the LED is soldered has a trace towards the centerhole where the LED goes through, and I guess that just connects to the front central ring over the edge of the hole through the PCB.
It’s a bit unclear since the purple lacquer layer is over it and it’s a small trace
That’s looks quite a pain to do far all part and will be an issue if you drill holes to the size of the led?
On the french project, they seemed (on earlier pcb when they put led from the back) to solder the led trough the pcb using another hole.
I would love to try such project, once my 16×32 rgb led matrix is done, it seems like i’ll have a new challenge ^^
for oxydation, i found some company to sells “electric plating pens” with solution for gold/silver plating.
Putting a very thin layer would be enough and could protect the circuit for a low extra cost
I don’t understand what you write but i know for a great doc video .Is possible look instruction video or photo for construction a little wall WATER GRAFFITI LIGHT ?
Sorry for my English,
Thank you ,I hope you can help me.