Is it reasonable to build a computer case that uses heavy amount of leds?

[Functional]

TL;DR section provided!

Hey everyone :) First of all, to introduce myself shortly; I'm just a 19 year old worker from Finland. Yesterday, while doing the usual routine of searching for parts to upgrade my computer from germany... I accidentally bumped while looking for water cooling into computer cases.

Suddenly, I felt desire of a new case as well. Old one is missing the side that can be opened and well, is just practically holding my components. Even my internal HDD's aren't inside of it. Not to mention, weights are holding them down because of broken connectors and input slots.

So, I went on to searching them. Found literally nothing interesting. Everything seemed to be just dull in many ways. So I decided; Why won't I build my own?


The first thought I instantly stumbled upon was a computer case that has lights that go along the music. I'm a producer as well, after all. Now, this is nothing new probably. But my idea was not actually to have them work along the digital input - my idea was to have them automated completely with specific songs.

To give an example - imagine this (http://vimeo.com/29093748) running around inside a square plexi glass. (Well, the parts with only lights).

I certainly have the imagination, but I feel that my way of expressing it would be to build a custom system that allows me to do it...

TL;DR

Is it practical in any possibly way to have, say, 1000 leds in an external case that goes on top of my custom built computer case, each one being separately controllable?

It feels like that less controllers I have, the less wire mess there will be. But how expensive are the controllers? And also, what are small 5mm leds called that can change their colour and be dimmed?

The first thing that worries me the most is the wire mess and how practical it would be. I'd imagine that it would be reasonable to have around around 100-200 per controller and that I can identify which controller uses which group (in case any gets broken).

Most of the leds would be attached to a center pylon as well in a manner that there won't really be much space for anything else. The pylon is a round pylon, so is it really even possible?

And I'd love to have leds/cold cathodes running along the plexi glass wall on the inside. But I have no clue what could be smooth enough to attach them on it. It would be also be also a round plexi glass.


And the last issue - programming a song towards them. I do have imagination, but having around 1k leds would probably be a pain to automate with Arduins per song. Is it really realistic? I mean, is it possible to hire a programmer that could create a custom software interface to control them?

I'd say all it would have to do is to have just every led rounded in their groups in the software and a timeline. From point zero to the defined end, whereas I could choose the colour of any lights that have customizable colour and brightness.



So, any thoughts? Sorry for the long post. Also, I don't need the answer of how to do it exactly (unless it is simple!). I'd rather get direction and then start researching myself exactly what to do.

[x88x]

First off, welcome to TBCS!

As for your idea; not exactly to my tastes (hell, even the MBB LEDs in my system bug me ;) ), but should be possible. It's all just a matter of how much time and effort you're willing to put into building it. And, as you say, the complicated part would likely be controlling them all in a dynamic manner. What it would break down to is:
physical design/build: relatively simple but very time consuming
control: complicated and possibly time consuming, depending on your backgroun

I don't know of any off-the-shelf controllers that would do what you want, but it would be possible with a bit of work. Do you have any experience with electronics and microcontrollers?

Check out some of these for ideas:
http://hackaday.com/2011/09/18/attiny-hacks-2313-driving-a-4x4x4-led-cube/
http://hackaday.com/2011/01/11/scrolling-marquee-made-from-ge-christmas-lights/
http://hackaday.com/2011/09/29/what-has-114-leds-and-is-always-running/
http://hackaday.com/2011/09/28/making-a-simple-addressable-array-from-led-strips/
http://hackaday.com/category/led-hacks/

[xr4man]

i was thinking that a controller for Christmas lights like this http://www.lightorama.com/
might do what you want. just use leds instead of Christmas lights.

[Functional]

Hey, thanks for the replys!

I don't have any real background as is. Spent too much time around computers, but never doing really anything more complicated than fixing CPU with broken pins (yes, it sounds complicated, but is far more easier than you'd imagine)


But when it comes to programming.. That's a bit another story. As a musical producer, automation is no new term to me. It's pretty much the essential thing that makes electronical music sound so dynamic apart from the melodies and the harmonies inside the notes used.


What I'd imagine is, the more controllers I have, the easier it'll be. I could have easily them on defined groups, less time to search and will take less time to make them all accurate (just not sure how realistic).

Unfortunately, there's another truth to that. More controllers: More expensive. Though less wires and easier repair if anything goes wrong. Let's say, I mess up the whole circuit by whatever means.


The xmas lights controller would be a great choice if it wasn't for the price. Arduino is more on budget. They have boards with 54 digital in/outputs designed for leds and their software is also with more freedom. Not to mention, 60 euros price tag on one isn't that bad. Considering that specific leds I want already will cost me around 1.2k at the very least, if I get an discount for ordering that many of them.


Meh, I'll have to buy one arduino, a bunch of leds and see how it works. I want anyway something for the case itself where the computer is as well. Lights that change along the temps using the mobo sensors and lights that also could change colours depending on what digital input from spotify will have on it's EQ.

A smaller scale project in the end will tell the truth about realism of so much bigger one. In this case anyway ;)


But knowing that it's possible is already a good start. Knowing that it will get me bankrupt isn't though.

[Airbozo]

Many years ago I worked on a project that involved thousands of LED's used in a color organ. The LED's were not controlled by anything except a custom circuit board that captured a stereo sound source, then fed that source to a circuit that filtered out certain frequencies and triggered the LED's according to volume of those frequencies. If I remember right we pulled about 100 or more different frequencies (which were tunable). I will see if I still have any schematics since it was a very long time ago. Only 3 of the devices were ever made as they were huge (6' x 4') and expensive. It took us weeks to drill and finish all of the holes for the LED's. Another couple of weeks to route and finish the channels for all the wires and another couple of weeks to install and wire the LED's once the panel was finished. It would be so much easier with a CNC machine...

It was basically a sound organ in a custom hardwood panel, only instead of 3 channels like most kits, there were over a hundred channels. We called it art although some would not. LOL!

For some ideas on color organs, check here:
http://wolfstone.halloweenhost.com/ColorOrgans/clobuy_CommercialColorOrgan.html

http://www.ledhoops.net/tag/color-organ/

or this:
http://www.youtube.com/watch?v=6mXM-oGggrM
Which can be built by going here:
http://www.instructables.com/id/Led-Cube-8x8x8/


...or search color organs on google.

[Functional]

Color organ is not what I'm looking for, I'm afraid :(

However, I think that Arduinos software can do similar stuff as a color organ. Problem is, it can only control 54 units. Would be a boring organ with just 54 leds. Well, not necessary, depends where you place it.


I'm more into automating, programming the lightshow myself per song basis. I know, it takes probably a long time at the start to get used to it. But might be worth the show eventually.


I'd love to give you an example of what I mean, but no such thing exists in youtube. And to put it in text.. well, don't got a week worth of time to explain what I have in my mind :P


PS. That Led Cube isn't really a color organ, but it's goddamn brilliant oO

[Airbozo]

There are christmas light controllers that run thousands of lights.

http://www.youtube.com/watch?v=CquGeUCGXFE&feature=related

I don't know much about the Arduino's but there must be a way to gang them together...

[Kayin]

That color organ is pretty sweet. Never thought of something like that...

[Functional]

There are christmas light controllers that run thousands of lights.

http://www.youtube.com/watch?v=CquGeUCGXFE&feature=related

I don't know much about the Arduino's but there must be a way to gang them together...


There's just the difference between channels. Please note, at this point, I know honestly very little about electronics. This all comes off my head from how I think it works, based on what I see in that video.



You can control 10000 of lights with just one channel. But that means, whenever you control that channel, you control them all.

But when you have 10000 of lights and you want to control every one of them without influencing another....

That requires 10000 channels then.

So, the xmas controllers I saw that had 14 channels costed around 300 euros or bucks, not sure. The Arduino that can control 54 channels is 60 euros.

There's however a simple explanition to that; Arduino is lot more complicated and it can't control a whole bunch of xmas lights, because it doesn't give enough power output to support them all. Meanwhile, it can give enough power support for say, 100 of leds. But with 54 channels, you can't control every one of them.



I know, controlling each one goes bit over the top, but that's really the plan. More control, more versatility. Necessary? Don't know, at least yet.

[Airbozo]

There's just the difference between channels. Please note, at this point, I know honestly very little about electronics. This all comes off my head from how I think it works, based on what I see in that video.



You can control 10000 of lights with just one channel. But that means, whenever you control that channel, you control them all.

But when you have 10000 of lights and you want to control every one of them without influencing another....

That requires 10000 channels then.

So, the xmas controllers I saw that had 14 channels costed around 300 euros or bucks, not sure. The Arduino that can control 54 channels is 60 euros.

There's however a simple explanition to that; Arduino is lot more complicated and it can't control a whole bunch of xmas lights, because it doesn't give enough power output to support them all. Meanwhile, it can give enough power support for say, 100 of leds. But with 54 channels, you can't control every one of them.



I know, controlling each one goes bit over the top, but that's really the plan. More control, more versatility. Necessary? Don't know, at least yet.

Necessary Schmecessary, do what you think is right.

The christmas light controllers are expensive because they have other bits you don't need for you application. Hell I bet they are similar to the Arduino's in the way they are programmed or just the basic logic plus the power circuitry, etc...

Let me check on another device I used to use in fire alarms. I know it sounds off base, but they are just simple boards that take a signal and act on it depending on how you program the device. They are capable of 256 channels on one board (smaller than a pack of cigarettes...) and will power anything from LED's to advanced power circuitry. Give me a bit to talk with the manufacturer to see if I am being daft about their abilities (and cost).

[Functional]

Haha, that sounds whack. As far as I know, most fire alarms in Finland (yes, I live here, but I don't mind ordering from other continents because I work in largest fin logistics company!), they seem to send some sort of radioactive particles from point a to point b, and if anything comes to block a particle (say, hazard smoke with all kinds of dangerous things to inhale), it will set off the alarm.

As long as it's not radioactive, I don't mind really using them either. I mean, it's not like my whole idea makes that much sense either. I just hope that when I turn a my first program on it, it won't explode. And if it does, at least it could explode reasonably well to make me forget how much time and money I wasted down the drain!

[Airbozo]

Haha, that sounds whack. As far as I know, most fire alarms in Finland (yes, I live here, but I don't mind ordering from other continents because I work in largest fin logistics company!), they seem to send some sort of radioactive particles from point a to point b, and if anything comes to block a particle (say, hazard smoke with all kinds of dangerous things to inhale), it will set off the alarm.

As long as it's not radioactive, I don't mind really using them either. I mean, it's not like my whole idea makes that much sense either. I just hope that when I turn a my first program on it, it won't explode. And if it does, at least it could explode reasonably well to make me forget how much time and money I wasted down the drain!

These devices I am talking about are not the sensor end of things. They are used in very large installations (think high rises and manufacturing plants) to do just about anything you want. They can be programmed to monitor a detector and provide continuous feedback to the main panel, or they can be configured to send a signal to the flashing lights and horns. Very versatile little devices and that is why I mentioned them. On their own I have used them to turn on lights, turn off a stereo and send an alarm signal back to my computer when someone walks into my office (the same unit did everything).

I am looking up the name now and will check on the cost. That may be the killer.

[Functional]

Sounds good.... and bad.

I got it now what you mean. And yes, they cost a lot. Usually, I believe, manufacturing plants et cetera uses a large quantity of those for various tasks and I think they'll get a straight discount for that alone.

Hmm. How realistic is it then to build my own circuit? Let's say, I have the soldering skills for that, but not the knowledge.

I could head over to another forum that would give me the knowledge to do so.

The only problem left would be programming it (in the best scenario). Since it will have to be programmed in that case pretty much from top to bottom.


Having that quantity thing assumed - this is just simply something I assume nowadays a lot because the logistics company for example that I work for, charges insane amounts for small deliveries. A 6kg package would mean a 70 euro delivery cost. Oh and that was 250km distance and not cross-country.

But when you get to the larger quantities and make a contract, that's another story. It'll be cheap and pretty much the reason that it's the biggest finnish-based company in here.

[Functional]

http://www.parallax.com/tabid/706/Default.aspx


This seems to be one solution. Only problem I have is how big that circuit is. But I guess having 128 channels means it's either that way or the highway.


And I'm speaking only for that circuit, not the program itself!

[x88x]

Was hoping someone else would remember what I was trying to think of...without me mentioning it...come on people, why aren't you psychic anymore? ;)

What you want is a bunch of multiplexers. Basically, they're small, simple (read: cheap), ICs that accept a more complex input signal and use that to control multiple channels.
More info here:
http://www.allaboutcircuits.com/vol_4/chpt_9/7.html
http://en.wikipedia.org/wiki/Multiplexed_display
http://www.youtube.com/watch?v=lZyc6ulpkyM
https://www.google.com/#hl=en&sugexp=pfwl&tok=fEDDc4EO4265m5tIynq0Xw&cp=10&gs_id=27u&xhr=t&q=led+multiplexing&pq=multiplexer+ic&pf=p&sclient=psy-ab&pbx=1&oq=led+multip&aq=0&aqi=g4&aql=&gs_sm=&gs_upl=&bav=on.2,or.r_gc.r_pw.r_cp.r_qf.,cf.osb&fp=2b92913e328ed8dd&biw=1918&bih=938

EDIT:
Hmmmm, those LED cube videos are making me really want to make one....but big....with a stupidly huge number of RGB LEDs....hmmm....

[Functional]

I see now. Thanks a bunch! I'll start on planning then. First gotta get ahold of one and a bunch of leds and see how it directly works.

I'd assume that you could get at least 4 leds to work through that method with just one channel? One Arduino could be enough already in that case for over 50 leds, but problem is, you can only connect it to 16 led groups due to physical limiations. But it does have 16 channels!

[TLHarrell]

Check out the LadyAda Blog: http://www.adafruit.com/blog/ lots of good info on there about using RGB LED strips. They also did a project for a new year's ball with a mess of LEDs in it. It was set up to operate using Processing, which would do all the pattern and control work. Definitely worth going back through the blogs and seeing how some of this stuff is done.