h a l f b a k e r yMay contain nuts.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Driving down the road looking at the decorative lights I've noticed that some people have theirs blinking. I don't like that. It annoys me and I don't know why, until now. All the bulbs either blink at once and make it look like a power failure or they do a little dance where every other bulb goes
out. It's not uniform, it's not pretty, and its annoying. What I thought of to make it better was to make the string of lights blink vertically or spell something out or something interesting. How to do that though? Well, have the base of each light 'report' back to the master-plug (the one in the wall) and report it's basic position according to where it's at on the tree/bush/house/person. The master plug can then report to a computer a 3-D model of where the lights are and the savvy computer user can design what they want. Now the master-plug should know where all the bulbs are so it can line them all up according to predefined user settings. Then, the master-plug will tell each bulb when it should blink. Now we have trees spelling out "Merry X-Mas" when they blink or houses that blink "Welcome."
8 channel Christmas light controller
http://www.luminus.cx/projects/chaser/ 12 Dec 02 | "...a 8 channel christmas light controller, currently serial driven using a Microchip 16F84. Using a WinAmp visualization plugin, each channel can be either turned on or off depending on the beat of the music. Its not limited to being controlled by WinAmp, as any program that can talk to the serial port can control the lights." [bristolz, Oct 04 2004]
Chaser controller
http://lights.fcpages.com/seq10a.htm Designs and schematics for a variety of controllers. [bristolz, Oct 04 2004]
Commercial Chaser Strands
http://shop.store.y...y/chaserlights.html [bristolz, Oct 04 2004]
for 8th of 7
http://www.foreverbright.com/ led christmas lights use only 2 watts per string of lights [mihali, Oct 04 2004]
Light Emitting Polymers
http://news.bbc.co....sci/tech/390220.stm [my face your, Oct 04 2004]
[link]
|
|
A plague on all your, flashy, blinky, twinkly, sparkly overdecorated tastelss houses. Bah, Humbug ! It's a waste of good electrical power that could otherwise be used to power the iron lungs of poor orphans with terrible tragic diseases. Fishbone (unilluminated). |
|
|
I reckon these would not be the "200 light string for $1.97" variety. |
|
|
The real key to this will be the production of an inexpensive integrated-controller LED which can be addressed easily. Perhaps a four-terminal package with a daisy-chained shift-register arrangement. |
|
|
There's lots of strands that can do chaser light effects as well as fades, etc. |
|
|
Yeah, all this stuff above, that's what I meant. |
|
|
You had half a mind to mention all that stuff? |
|
|
Oddly enough, I've thought about building Christmas lighting along these lines. My thinking was more limited, though. I was thinking of making a version of those "net" lights in which the bulbs are indivdually controlled. |
|
|
First of all, there'd be no reason to run AC120 out to the lights. Low-voltage stuff is much easier to control, and regulatory approvals would be easier as well. |
|
|
Second, while the DS2045 would be an option, I think using a shift-register arrangement would be much more practical. I don't know what exact restrictions exist on GaAs semiconductors, but I would expect that one may be able to do a simple shift-register stage monolithic with an LED. Alternatively, incorporating a silicon-based shift-register stage would be no harder than incorporating a "blinking" chip [which are already quite common]. |
|
|
I would suggest using a four-pin LED package. Vss, Data-in, Data-out, and a combination power/clock/strobe pin: a 1us dropout on the PCS pin would cause data to be shifted through the LED's, while a 20us dropout would cause the shift registers to be copied to the LEDs themselves. Alternatively, a 1us dropout on the power pin would trigger a shift, and the LED's would be disabled until power was stable for at least 10us. |
|
|
Otherwise, if one wanted nicely-addressible "net" lights, I'd suggest that one should wire the lights as a matrix and run everything off line-isolated low voltage so as to allow use of thinner wire than is possible with stuff that's exposed to AC120 (use of lower voltage would require higher currents and thus potentially require a heavier conductor, but in low-current AC120 applications insulation thickness dominates over conductor thickness). Depending upon the application, one could either have 256 lights wired as a 16x16 matrix (which would require somewhat bulky cable, though perhaps not too bad if thin wire could be used) or else have many groups of 16 lights, each of which is a 4x4 matrix operated by a separate controller. |
|
|
Why DO people put crazy lighted stuff on their houses in December? I've been wondering about this and all I could come up with is that somebody, somewhere, at some time decided it might be fun to make their real house look like a gingerbread house covered with gumdrops and other colorful candy. |
|
|
I think I'm outgrowing Christmas. |
|
|
Gee, now we know who the engineering egg heads are around here... When I was writing this I was thinking of all the bluetooth annotations I would get. Also, why couldn't you just use a fiber switch to address all the lights and light them? Can't you handle more with fiber then you can with serial? The only thing is you would have to have it hooked directly into a computer instead of into a plug like the original idea, but still... |
|
|
[Steve Deg], [Supercat]; I'm terribly, terribly sorry about this, but I'm afraid I'm going to have to hunt you down and kill you. |
|
|
Nothing personal, and have Happy (Nasty flashy light-free) Xmas. |
|
|
I've seen strings of individually addressable lights. |
|
|
Given that, we have to solve the positioning problem, and here's how. Use a cheap-ish webcam. Attach it to the controller. The controller lights each light in turn and records where in the webcam it shows up. Then you move the webcam to a new location and repeat. Do this two or three times, and it can triangulate pretty well, and now you've got your 3-D model, no magic locators needed. |
|
|
OK, now back to addressability. I wonder if it wouldn't be cheaper, rather than building logic into each light, to use a ribbon cable. 32-strand ribbon cable could theoretically control 256 lights. Maybe you couldn't pump enough juice through an individual wire to actually light the bulb, so instead you have a single transistor at each bulb, but that's got to be cheaper than a big ol' IC. Discussed at great length in other forums, though. |
|
|
LED's are available which incorporate a flashing circuit inside the package. A simple shifter as I described shouldn't be significantly more difficult, especially if one did away with the secondary latching circuitry and simply designed the device to run off 1.5 volts during the "shift" cycle and 3 volts between shifts (at 1.5 volts, the LED's wouldn't have enough power to light, so the shifting would be "invisible"). |
|
|
BTW, if one is trying to really maximally simplify the circuit in each LED, there are some tricky methods for building dynamic (capacitor-based) logic which would require goofy driving waveforms but would reduce the circuitry to practically nothing [e.g. on the order of 2-3 transistors per LED] One such approach would be to use two transistors on each LED to form a memory "bit"; an resistor-capacitor-coupled input from the previous stage would bias one of the two transistors. The effect would be that every time the memory bit was powered on it would "default" to the last value held by the previous bit and then hold that value as long as power remained. Total circuitry per LED would be on the order of two transistors, one small capacitor, and six resistors. I don't know whether GaAs can be used to make transistors, but if so, a circuit like that might be constructed in a monolithic chip with the LED. |
|
|
< quietly adds [egnor]'s name to Death List> |
|
|
But you're all looking at the problem as if it's something to do with cabling or watt-ever. It isn't. It's to do with the viewer's perspective. |
|
|
If you have lights arranged in 3 dimensions flicking on to form a pretty pattern from one viewpoint, the odds are that from any other viewpoint, the lights will simply make a random pattern. The odds of this happening reduce dramatically if you go down to 2 dimensions with the viewer outside that plane, and *then* it simply becomes a question of cabling. |
|
|
Yes, yes, [8th], you don't like Christmas lights, we get it. |
|
|
[Supercat], I see what you're saying. Unfortunately, what's potentially most efficient for mass manufacture isn't necessarily what's most efficient for hobbyist construction. Both problems are interesting to think about, but even for mass market, you have to have a *seriously* mass market before designing your own custom GaAs LED fabrication system starts to make sense. |
|
|
[PeterSilly], I don't think it's all that grim. Imagine a Christmas tree where the lights make a "wave" pattern (as in a stadium) along the Z (trunk) axis. I think that would look pretty neat. |
|
|
In general, since lights tend to follow the outline of some regular and even visible object, people don't just see a point cloud; they see the outline of a shape. Since they're seeing the lights in 3-D, not 2-D, I think you could play tricks with spatial addressing that would work well. |
|
|
<GROAN succumbs to temptation to join discussion> |
|
|
The effect of spatial addressing would depend a lot on the size of the array, the POV of the observer, and the acievable maximum parallax. I don't think you could get a sufficiently deep array if the POV of the observer was about 50 metres away without having a very big structure wich would be vulnerable to wind and weather. A 2D array is the best idea from a practical aspect. |
|
|
Don't we already have enough of friggin "Merry Christmas" blinking signs in the windows, or those gawd-awful lighted twig reindeer on the lawns? Do not these constitute a 2- or 3-dimensional represenatation of das blinkin' lights used to further the merriment of Christmas-goers or epileptic seizures of others? |
|
|
Really, are houses, shrubs, trees, and lawns really supposed to look like a circus just because of the calendar and "tradition"? |
|
|
Hey, I resent that, [XSarenkaX]! My house doesn't have *that* many animated juggling clowns this year. Hmmph! |
|
|
<aside> Guys .... look, that's where [PharoahM] lives, the one with the juggling clowns in the garden ..... right ...... pointy end goes in first ...... down a bit ...... now, stay clear of the back of the tube, these recoilless jobs have a bit of a backwash .... |
|
|
Ready, everyone ? </aside> |
|
|
Firing upon me, eh? Hmmph! Now I'm outraged! You're Fredo to me, [8th]! *puts up anti-missile trampoline screens* |
|
|
Every year, dad and I hang up a giant 35 foot cross he welded together and then strung lights to, on his TV tower, which sticks up about 70 feet in the air. Since the house sits on the highest point for about a 2 mile radius, you can see it from 3 miles away.
Really impresses the local Amish.
I love coming home for Christmas. Dad has a way of doing things over the top. |
|
|
You post a symbol of Jesus' death/resurrection on a holiday that supposedly celebrates his birth. Don't EVEN get me started! |
|
|
Alternatively - Light Emitting Polymers. Linko. |
|
|
//You post a symbol of Jesus' death/resurrection on a holiday that supposedly celebrates his birth.// Though at first glance it might seem illogical to do this, it's entirely consistent with a Christian theological framework. Christ himself said that the entire purpose of his incarnation was to be crucified for the sins of humankind (see Luke 13 32 & 33 for one of many instances of this). It is at least within the pale of acceptability to acknowledge the purpose of Christ's coming when celebrating his birth. |
|
| |