For those who stumble upon this note, some history is mandatory in order to better understand my goals on this LED projector light.
So I suggest you put the LED light v first. 1.
I posted 0 instructable a while ago.
When you finish the second part of my adventure, please come back and have a look.
For this version of the projector lamp, I know I have to find a way to at least double the amount of light on the projector screen. As high-
The Power LED here is very expensive (
I live in Sibiu, Romania)
I want to try to make my own led array.
I chose the NS6W183T LEDs from Nichia.
For those who don't know, Nichia is a Japanese company that produces high quality led, the first company to invent high power blue and white led in 1990s.
With these LEDs here, the light output is tempting.
This structure is not finished yet.
In addition, the time frame of the project is not specified.
I'm posting the progress of it because in order to get it done, I need opinions from other users on something.
If you think you have something really valuable to add to this project, please be my guest and share your comments in the comments.
I will try to answer all the questions.
My choice of the NS6W183T LEDs is mainly based on light output and size.
These are SMD components, each of which has an amazing 245 lumens depending on their size (
When powered with If = 700 mA).
I had to settle for an array of 9 LEDs, mainly due to price issues.
My budget is very limited.
Still, I paid around € 55 or $66 on these LEDs.
Expensive for Romania and cheap for others.
The main goal is to fill the entire projector condenser lens with light, and that's what the array does.
But there is a problem.
The beam of the SMD LEDs I get is distributed at 120 degrees.
I have no choice but to hold on and hope I can find a way to focus this light in some way.
Selected cooler (
See cooling steps)
Ron is 87. that's 24. 5 Euro or 25. 5 USD.
I got this because regardless of the outcome of this lamp project, I have its future use.
The rest is the part that has been cleaned up, which is really not worth mentioning.
People in the DIY community know the importance of bin diving.
So far, the final cost of the material is about 100 euros.
There is also a failed project, and I will hear my wife say, "You know, you could have bought that original light, considering that a lot of money has been spent ".
This is what I call a bitter irony.
To avoid producing too scattered areas of light, the array is as compact as possible.
This is because all the projector light is emitted from one point.
With arrays, you get a few, and the challenge is to focus them in one place.
I will deal with this later and there is still a long way to go until then.
9 NS6W183T LEDs are arranged in series on a custom PCB board of 3x3 cm.
The Led is welded in the center of this board, within a square of 2x2 cm.
I could have made it smaller, but I can't weld SMD components with reflow soldering, all of which have to be done manually with soldering iron.
Each Led is checked for polarity, glued to the board using the northern Silver thermal composite glue, and welded quickly and carefully.
You don't want to cook too many LEDs, they are very fragile when it comes to temperature.
My goal is to transfer the LED
The heat generated goes to the PCB and then to the cooler behind the board.
Four Screws fix the PCB on the cooler so as to avoid PCB warping and use hot paste between the PCB and the cooler.
After welding all the led on the circuit board, fast power check (a few seconds)
It is proved that the array is running normally and the light output is huge.
If you make the mistake of checking the led (like I did)
I promise you will see nine points of light in a very long time.
Hopefully I can make good use of this very uniform light.
Like the 9 LEDs I'm using need 700 mA at 31.
5 v is the total amount of 22.
The power consumed is 05 Watts.
I should get at least 2025 lumens under these parameters.
If I can keep the led temperature below 70 degrees (Celsius)
Then I can try to power them at 800 mA and then I will get 2340 lumens.
I think the light loss inside the projector will reduce this output by half.
My goal is to use this projector in the original ecological environment.
Mode, when the light output is only 1000 lumens.
Hopefully I can get closer to this now than I did last time.
I will use the same power supply as the lamp v. 1.
0, is a constant current power supply of 700 mA.
It outputs up to about 30 V so I will be close to its maximum rated power output.
Checked with a multimeter, the current does stabilize at 700 mA, but the voltage is only 29. 2V.
Good enough for the initial test.
Another concern is the power consumed with a rated power of 3 per LED.
Dissipation power 2 w.
This figure is as high as 28.
I need to remove 8 watts of heat.
Given the limited space inside the projector (5x5x9cm)
I have to find a cooler suitable for this space.
A very extensive Google image search makes it clear that the best option for me is the Nexus NHP-
2200 chipset cooler.
This one uses heat pipe technology to cool the computer chipset.
It is made entirely of copper, and in front of the projector's cooling fan, it fits into the available space, and the cooling piece is exactly where I want it.
This cooler is really beautiful and I try to handle it with gloves to avoid fingerprints everywhere.
The cooler base is already a bit oxidized, but that doesn't affect its performance, just the overall look.
The tricky part is to let the led cool because the pcb I use should be made of metal, but I don't have this technology to use.
I am counting on the projector fan, there are two in front and one large on the back of the bulb.
Hopefully these will help control the temperature.
The initial test proved that the LED array gets hot very quickly, and the PCB is too hot to touch in a minute without cooling the array.
The temperature probe on my multimeter reaches 70 degrees (Celsius)
After a few seconds, the projector fan did not cool.
I don't have a way to check the array with the cooling of the projector yet, so I'll keep using it.
Hope I don't blow up the led too early.
The Nexus cooler works fine and warms in a minute, but the heat transfer rate between the array PCB and the cooler base is slower than expected.
The LED array is already too hot when this cooler gets warmer.
Man, I wish I had a metal pcb . . . . . . Three projector fans should help, at least I hope so.
Believe it or not, this is the hardest and hardest work for me --intensive part.
Because the whole LED conversion should be right
Destructive, I don't have the original bulb at all, I have to copy the original frame and mount as much as possible.
There are two mounting screws inside the projector and several plastic pins need to be considered.
None of them are in line with others.
So I have to measure everything I can measure in that small space inside the projector and try to make a model
First, remove my lamp shell from the cardboard and check the measurement results.
Later I had several cardboard models and all the distances were correct.
The PCB board I used for this case is a good material because it is strong and can be easily welded at a 90 degree angle, and I will fix everything that I can put in the frame firmly in place.
The hardest part is to cut the PCB board into the various shapes needed to assemble the housing.
I used a little puzzle and I can only say it was a painful experience.
Since this array is square, I can't use a circular reflector like I used on v. 1. 0 lamp.
So I will make a square with a thin aluminum plate.
It will also control the light overflow as it will interfere with the projection experience.
The second benefit of this aluminum reflector is that it can also double as a radiator as its base touches the LEDs, thus taking away some of the heat generated.
So it's more important to control the light overflow than to guide the beam because, as I mentioned earlier, the LEDs have a beam of 120 degrees, and I don't have a real way to concentrate too much.
The lens is useless here as I have 9 separate light sources distributed on 2x2 cm.
There is no Fresnel lens here either.
I'm thinking of using a small cd. ROM lens (
From the head of light)
For each LED, but the focal length of these tiny lenses is very short.
This project is stuck here.
I try to use this light in the projector.
Light output is not good.
My lights are everywhere.
The light output is blind, but there is no way I can control it.
The projected image is weaker than the one I took with the LED projector lamp v1. 0.
I blame the beam angle.
If I could find a way to effectively focus the square LED array, I believe the end result would be a success.
But there has been no effect so far.
Maybe this forum will help me find the answer and that's why I posted this unfinished project. (
And get rid of those who ask about their status)
Okay, I'm just kidding.
Update: cancel this LED bulb version.
Due to time constraints and the lack of encouraging results, I will postpone the project until I can come up with a better solution.
Instructures will be the first place to read it when I do so, so please keep an eye on it.
At the same time v. 2.
Version 0 is my new emergency light. :-)
Perfect White and high brightness, just like the operating room inside.