Specs:
- Controller can power up to 360 watts of LED lighting - between 3 and 100+ power LED's
- Analog dial or slider for Red, Green and Blue - create an infinite number of light colors!
- easy to make a room light or a spot light by switching lenses / covers
- very fine analog control for precise color matching (or white matching)
- Includes current-limiting protection circuit
- Simple 7-component thru-hole circuit - No microcontrollers or programming!
- Controller cost: $25
- 50-watt, 15-LED lamp cost: $50
- Efficient: 50 watts of LED light equals 100 to 150 watts of incandescent light.
- Efficient: 80-90% controller efficiency
- Controller can be used with a variety of power sources
- Recycled materials & electronics: 20-40%
I've included a lot of detail in this instructable that applies to any power-LED project, and i've got several other power-LED instructables too, check those out for other notes & ideas.
why not digital?? this controller was specially designed to be as simple as possible by using analog-only. it turns out that it works really well. yes, a digital PWM controller is "better", but it requires "chips" and is thus more complicated to build. i've included a circuit for a digital PWM controller here.
This article is brought to you by MonkeyLectric and the Monkey Light bike light.
Remove these ads by
Signing UpStep 1: How does color mixing work?
lets start with your eye, which is an incredibly complex and sophisticated image sensor. how does your eye see an image? your eye has millions of individual sensor cells, a bit like the pixels in a digital camera sensor. your eye has 2 major types of cells: rods and cones. there are about 100 million rods, they are very sensitive to light but only see black and white images - they are your night vision. there are also about 5 million cones, which come in three types. each of the three types of cones sees different colors. you can think of the three types of cone sensors as 'red cones', 'green cones' and 'blue cones', although this is a bit of a simplification of how it really works.
so, how do we see colors? every color you can see is a result of how your brain interprets the signals it gets from the three types of cones. if you look at something and only your 'red cones' see it, then your brain sees red. if you look at something and both your 'red cones' and 'green cones' see it, then your brain interprets yellow. every color you can see is the result of how the brain interprets the different strength of signal from the three types of cones.
how does that help us? it means that using only three different colored lights (red, green and blue), we can easily fool the brain into thinking it is seeing any possible color. lets say we have a real yellow light: this light activates both the red and the green cone sensors in a specific proportion. but we can also activate the red sensor with a red light and the green sensor with a green light. if we activate the two sensors in the same proportion that the yellow light would have, then our brain can't tell the difference from a real yellow light and we see that red plus green equals yellow. we've fooled our brain!
it turns out that this is a very popular trick. every color television and computer screen ever made uses this exact principle! also most printers, cameras, and other imaging devices are based on this principle. if you have an older 'tube' screen or a plasma screen - look closely at it while it is displaying images and you can see all the small red, green and blue pixels (a magnifying glass helps).
for more information:
http://en.wikipedia.org/wiki/Color
http://en.wikipedia.org/wiki/Cone_cell
http://en.wikipedia.org/wiki/Eye





















































Visit Our Store »
Go Pro Today »




I tried to assemble a circuit like that with 4 SEOUL LEDs of each color (R42182, G42182, B42182), however the 10k slider I added does not seem to work like expected.
With my 19V power source I can switch only between like 13.2V and 12.9V incoming to the LEDs. It joes does not dimm anything...
No matter what resistors I use with R1/R2 (tried several combinations, even plain trial&error), even without any resistors the circuit is just lighting. Without of course at maximum brightness.
Anyone got an idea where I have failed here? (Also have some 12V computer power supplies laying around if that would be all it takes...)
So I came up with a VERY simple way to run and control LED's.
Assuming the power supply is 12V - or on a charging car battery - say 14.2V.... or even 15V for the silver / calcium car batteries...
OK just for simplicity's sake we will have a 12V power supply.
Each LED has a PEAK forward voltage.... which is only relevant for the amount of current in the supply. I have run single 3.5V 20ma LED's on 240V mains supply with only teeny capacitors to limit the current - they were getting 200V... and almost zilch milliamps... and they ran quite dim.
So OK assuming that the 3W LED's peak forward voltage is 3.9V.
Each LED has an internal resistance - and with enough of them in series - they LOWER the available voltage to each LED - thus acting as the current limiter at the same time.
So if we add 4 LED's in a string and run them on the 12V, we get a total peak voltage capacity of the LED's at 15.6V, thus giving each LED 3V - which is about 20% less voltage than the "accurately regulated voltage and current supply".
Less voltage = less current - so the LED's will also be a bit dimmer, but you will also have more of them...
QED.
It's really just a case of suck it and see.
For 2.2V LED's on a 12V supply, I'd be using say 8 of them in series to start off with and then perhaps dropping to 7 LED's in the string.
There is a magic point of about 15% less voltage than the peak forward voltage that gives enough resistance to stop thermal run away from occurring - and the light is still quite bright, and you have an extra LED or two in the series...
AND you don't have to mess around at all with any BS current limiting circuitry.
Doing the calculations, getting the parts, making the circuits, and they also tend to waste power etc.....
This way - dead reliable, and slightly dimmed LED's last almost forever.
http://www.learningaboutelectronics.com/Articles/Why-attach-a-heat-sink-to-a-voltage-regulator
But great job on putting this together. Great stuff.
i have also sinilar led costs .3 USD here ( yes 1/3 rd of a dollar us ) in pakistan.. they say its one watt.. i tried it running at 750mA it was muchg much brifghter but was also very hot .. so i returned back to 350mA usng pot .. so i think if i attach a large number f these to a pc heatsink and attach a 12 volts (1.5 Watt) dc brushless fan it can be very bright as well as cool and give more life ? what r ur words? have u over driven it ? can u share more info ..
Thanks again for all the helpful instructables.
ex.
60 degrees and under bright blue
Sunny Day Yellow
Get home from work/school Cyan
2.9V @ 700mA
3.35V @ 3A
I want to control it up to the max current. Seems you define R3 for a single current. Do I calculate that with the max (3A) current, giving a 0.2-Ohm resistor (closest value)? Thanks!
Datavideo CKL-200 Dual-Color Chromakey Light System
Lets say i have a power source V and a luxeon Led, and Resistor(R1) in between that makes the led shine in full power.
and now i insert in between a pot meter in series, This would change the resistance of the circuit lowering the current, meaning making the led give off less light, never allowing to go above the R1 value which is the max.
now do this three times for RGB and you got the same project with simpler circuitry don't you? why wouldn't that be good?
why do you choose to use voltage regulator and transistors?
This is my only question before i start making this instructable so could someone please explain this in detail to me? Thanks
Say the current flowing through each LED is 700mA (0.7A) which is quite common for high powered LED's. Now, if the voltage across the LED is 3.2V (full brightness), the the power being dissipated is: P=IV = 0.7*3.2 = 2.24W. If you connected a potentiometer in series with this LED then it would also have to be capable of dissipating this much power. However, most pots have an absolute maximum power rating of 0.5W which is well below what is necessary, which is why it starts smouldering.
Using the voltage regulator, the current on the input branch (the one with all of the resistors in it) is kept very low and therefore the power dissipated by each resistor is very low. But, the current in the output side can be SET (using R3) to anywhere up to 3 or 5A depending upon what voltage regulator you use.
http://www.superbrightleds.com/cgi-bin/store/index.cgi?action=DispPage&Page2Disp=/specs/LDRF-RGB4.htm
but I am not quite happy since the analog driver that makes this project so beautiful is not included, which is neigh impossible with a remote I imagine. Anyway, do you have any advice for me? Thanks in advance!
I'd like to construct the controllable driver with voltage regulator for use with (2) 10 watt LEDs; V= 23 volts, current = 1000ma. Cut off voltage =18 volts, max voltage= 23 volts. I'll use the 2.5K pot for R4. I need to figure values for R1&R2, but I am math/calculation deficient ,Any help will be appreciated.
www.sparkfun.com/commerce/product_info.php
Used single 10W led for each colour. Also I used potentiometers with switch, so I can turn off each chanel completely without the VREG burning. The hardest part in all this was getting the right R1 and R2 and insulating the VREG.
excellent instructable
Three Things:
1. The link http://www.lumiledsfuture.com/ is outdated, you should remove it.
2. The other link http://www.futurecb.com/ works fine, but
3. In the FutureCB link, there is a parenthesis at the end of the link, making it so you can't directly access the site from the link. You should edit the link so the parenthesis isn't messing up the link.