Projector Bulb Conversion to LED

Published

Introduction: Projector Bulb Conversion to LED

About: Electrical Engineer | Woodworker | Robot Builder | Maker | Fixer

I received a Projector from a friend December of 2010. It is a Sharp D100U from 1997. It did not have the original bulb, but everything else worked fine. I did some research online to see if I could get a bulb for it and found out that it would cost me $300. I could buy a reasonable projector new for that much. All I really needed was a new light source, so I thought a high-powered LED would work great. 

There were two stages to this project. The first was figuring out how to bypass the bulb checking circuit, and the second was building the new LED light source. 

I found these two Instructables by claudiopolis to be very helpful with this project:

https://www.instructables.com/id/LED-Projector-Lamp-v10/

https://www.instructables.com/id/LED-Projector-Lamp-v20-in-progress/

UPDATE!

I found the maintenance manual for this projector

http://diagramas.diagramasde.com/otros/Manual%20MAINT.pdf



Step 1: Bypassing the Bulb Checking Circuit: Observations and Research

The first thing that I wanted to do before I started taking the projector apart was to see how it currently ran. I plugged it in and turned it on. The fans came on and I could hear the sound of the fans and a faint ticking sound. along with two green lights that indicated Power and Bulb. After about 3 seconds the bulb status light started to blink, and after about 5 minutes the projector shut off and both lights turned red. I shut off the main power switch the red lights turned off. I then repeated that process a second and third time to ensure consistency.

I did some research online to try and find any documentation on the projector. All I was able to find was the user manual. Reading it I found the section that addressed the indicators: Power, Bulb, Temperature. There are three possible states for each indicator: Solid Green, which indicates Normal operation; Blinking Green, which indicates Starting up; and Solid Red, which indicates a problem. 

I removed the top case and accessed the main logic board. My intent was to find the logic signal that was telling the the board that there was a problem with the bulb.

I used my multimeter with the ground connected to a grounding point on the projector and started measuring values of all the wires going to the main logic board. 

I was lucky to find that each set of wires was labeled, but it was only abbreviations. After I recorded all the wires in each state, I made some assumptions as to what each set of wires went to.

RC - Remote control  (Since i did not have a remote for the projector and did not want to buy one I removed this wire and sensor)

BL - High voltage and lamp check circuit ( This is the wire set that I finally figured out was the one I needed to focus on )

LF - Bottom fan cover sensor  ( The two wires on this connector need to be cut and tied together if you want to be able to run the projector with out the fan cover ) 

FF - Fan 

LL - Lamp cover sensor  ( The two wires on this connector need to be cut and tied together if you want to be able to run the projector with out the lamp cover )

Q - Temperature sensor

FN - Fan 

There was also a set of colored wires in one connector that I also recorded, but these ended up not being needed to modify the projector.

WHEN CUTTING AND TYING THE WIRES TOGETHER MAKE SURE TO CUT A COUPLE INCHES AWAY FROM THE CONNECTOR. THE WIRES SHOULD BASICALLY BRIDGE THE CONNECTOR. DO NOT SIMPLY CUT THE WIRES OFF OF THE CONNECTOR. ON THE OTHER HAND, THE ORANGE WIRE NEEDS TO BE CUT OFF OF THE CONNECTOR.




Step 2: Bypassing the Bulb Checking Circuit: More Research and Tests

With no luck on the logic board I turned my sights to the main power supply. I did some more research and found that some projectors use optocouplers which are used to measure the feed back from the bulb (see the photo for a full illustration of how one works). I gathered from my research that I had to bridge the optocoupler that controlled that bulb to trick into thinking that it was working fine. 

I started to dissemble the projector even more; all the way down to the power supply. I found the three ICs that I thought were the optocouplers and through trial, error, and almost bricking the whole projector I was right back where I started. 

The projector sat in pieces on a shelf for a couple of months while I was busy with school and other things. 

I came back to it in the summer and tried again. I reassembled it and tried to figure it out with more research and more trial and error. I was able to get the high voltage arc to stop by unplugging the BL cable from the main board, so I was starting go get some where.

Step 3: Bypassing the Bulb Checking Circuit: Found the Circuit

Once again, the projector sat on a shelf until December. It had been almost a year since I received the projector and I had barely gotten anywhere. I was starting to think that it would never work. Just when I was about to tear it apart and use it for parts, I decided to give it one last try.

I focused on the main logic board. I figured that if there was any signal coming from the power supply it had to connect to the main board to tell it to make the led blink and shut off the projector. 

It was then that I realized that maybe for that 3 seconds when the indicator was solid green that it was sending the right signal that was telling that the projector was working fine, and the moment that it started blinking the signal would flip. I had not tested that in my initial testing of the main board, so I gave it a try. 

I was looking for a wire that was 5 volts or 0 volts with the light solid green, and 0 volts or 5 volts when the bulb indicator started blinking.

I focused on the set of three wires that were coming from the power supply to the connector labeled BL. This cable was the one that I had unplugged to get the high voltage to stop, so I checked it for the values that I was looking for. 

Just as I had thought, the middle red wire started at 0 volts and jumped to 5 volts when the indicator started blinking. I cut the wires and spliced them together to bridge the connector. Then I put the projector back together to test it. I also found that the orange wire was the one sending the signal for the high voltage start up. I cut that wire so that it was no longer connected to the main board.

I turned on the main power switch and then pressed the power button. The fans came on and the green indicator lights came on and stayed on. I thought that time had stopped. I stared at that little green light for almost 30 seconds just to be sure that it was not blinking. 

I had finally found and fixed the circuit that had taken me almost a year to find. 

AGAIN: WHEN CUTTING AND TYING THE RED AND BROWN WIRES TOGETHER MAKE SURE TO CUT A COUPLE INCHES AWAY FROM THE CONNECTOR. THE WIRES SHOULD BASICALLY BRIDGE THE CONNECTOR. DO NOT SIMPLY CUT THE WIRES OFF OF THE CONNECTOR. ON THE OTHER HAND, THE ORANGE WIRE NEEDS TO BE CUT OFF OF THE CONNECTOR. 

Step 4: Building the New LED Light Source: LED V1.0

The next step was to find an LED that would work. The original bulb was a 265 watt metal halide, 2000 lumen bulb. I did some more research and found a 20 watt, 2000 lumen LED for $20 and the 18 volt power supply for $7 from DealExtream.com. 

If you have ever ordered from DealExtream then you already know that it takes about 3 to 4 weeks to get what ever you ordered. I waited all month for the LED and driver to arrive. 

I first did a quick test to see if the LED actually worked, and it did.

Next I had to build a housing for the new LED to mount to. I used a piece of PCB board as the main piece and cut it to fit inside where the original bulb housing would go. I attached the LED chip to the heat sink and used heat sink compound to ensure a good heat transfer. I connected the heat sink to the PCB using some aluminum brackets that I made. I soldered the LED to the power supply and used heat shrink tubing around the wires to keep the heat from the LED from melting it. 

I put the new LED housing into the projector, set it up, and tested it out. The picture was not very bright, and I had to have all the lights off just to see it. I adjusted the projector's brightness and other settings which helped a little, but not by much. 

I did a quick comparison of the two light sources just to see home much energy the new LED saves. The 265 watt bulb used 0.265 kWh in an hour while the 20 watt lamp only used 0.02 kWh, about 13 times more efficient. 

Step 5: Building the New LED Light Source: LED V1.1

I decided that a reflector and a lens could probably help focus the wide angle of the LED. I used Google Sketchup to do the math and made a template that I could then use to make the pieces out of aluminum sheet metal. I found a lens that focused the most light and fit within the reflector. 

I'm sure that this could have been made way better and more mathematically, but at the time I did not know the math or the physics to do so. The hardest part about using this type of LED is trying to focus 20 individual light sources, each one with its own focal point. 

This modification actually helped a lot. The picture was much brighter and you could actually watch a movie on it, but I still needed to have all the lights off and the darker scenes were still hard to see. 

Step 6: Building the New LED Light Source: LED V1.2

For the next modification I tried actual mirrors for the reflector rather than the dull aluminum. This made the picture even brighter and easier to see, but again it still was not bright enough to enjoy a movie. 

Step 7: Building the New LED Light Source: LED V2.0

I did not use the projector almost all summer and I was disappointed that I couldn't get the LED light focused more. I was browsing around on eBay in the fall and I came across a 100 watt, 8000 lumen LED for $60. It came with the LED driver, lens, and reflector. I thought about it for a while and did some research about the increase from a 20 watt LED to a 100 watt LED. There would be a substantial increase in heat that this new LED would produce, but since the original bulb was 265 watts i figured that the fans and the heat sink would be able to handle the heat. 

I bought the LED and it arrived later that month. Changing the LED was not too hard at all. the new LED had the same mounting spots as the last one. All I had to do was take the old one off and put the new one on. I soldered the new power supply, and glued the lens and reflector to the LED. 

I was sure that there would be a huge increase in the picture brightness going from a 2000 lumen LED to an 8000 lumen LED, and I was right. I was able to see the picture even with the lights on. when I turned the lights off the picture was just right. Bright enough to watch a movie. The only concern that I had was all of the heat that the new LED was producing. I ran the projector for 30 minutes and then let it cool down; then an hour and finally for three hours. From these tests, I found that the heat sink and the fans inside the projector were enough to keep the LED from over heating.

I did another comparison of the two light sources. The new 100 watt only used 0.1 kWh in an hour and would cost only 1.2 cents per hour. Again, the 265 watt bulb used 0.265 kWh in an hour costing 3.18 cents per hour, 2.65 times more efficient. 

Step 8: Finished So Far

As for now the projector is finished until I get another idea to improve it even more. Over the whole project I spent about $100 on LEDs and other materials and it took almost two years to complete. It is bright enough to enjoy a movie and I am satisfied with the control of the heat. I'm sure a few years down the road there will be better LEDs that will work better for this project, but for now it is good.

Share

    Recommendations

    • Woodworking Contest

      Woodworking Contest
    • Casting Contest

      Casting Contest
    • Oil Contest

      Oil Contest
    user

    We have a be nice policy.
    Please be positive and constructive.

    Tips

    Questions

    56 Comments

    I've been wondering about replacing the lamp in my DLP projector... Would it be feasible to use a CREE torch? Would something like this work? http://www.ebay.co.uk/itm/5-X-Cree-LED-XML-T6-Rechargeable-6000-Lumens-Bike-Bicycle-Headlight-Headlamp-UK-/251680085950?pt=UK_SportsLeisure_Camping_LightsLanternsTorches&hash=item3a994d57be

    4 replies

    There's no way (DIY-way) to combine the multiple LED's into one single-point-source... which is what's expected for projector light sources. The arc-lamp reflector creates a single-point light source, which is easy because it starts with a single point light source... an electrical arc. Using large, multiple-emitter LED's will result in horrible efficiency. Single-large emitters can be focused, and can result in higher efficiencies, to where a 10W single-chip LED may work far better than a 100W LED that used 100 chips. Look for large-single-die (single-chip) White LED's.

    there's no need for a single point source, it's not a shadow box. just get the light to go through the LCD evenly and into the lens.

    I re-read what I wrote, and remembered what I was thinking about... 10 year ago, I was trying to convert a pair of 10,000 lumen DPI Lightning projectors from Xenon 5,000 Watt lamp to a LEP plasma lamps... yet using the original Xenon lamp reflector. It's the parabolic reflector that requires a point light source... then the reflector turns that into an evenly distributed light field across the optics.

    How did the blacks work out, what was the contrast ratio like - were the blacks actually kind of bright?

    I'm looking to make one and I'm trying to figure out how the LCD screen will fare with a brighter light coming through.

    2 replies

    No DIY LED conversion will result in a brighter image than the original arc lamp... those lamps typically have 20,000 to 50,000 lumens output from the reflector. Projectors are VERY lossey... often less than 10% of the lamp brightness comes out the front of the projector.

    yeah, 10% sounds about right, you lose half of the light in the polarizer, and two thirds of what's left in the colour filter you'd get 16% with ideal parts ans a perfect light path. If you can replace the color wheel with coloured LEDs you could improve matters threefold.

    Best way to do a bypass would be to buy one of these cheap chinese bulbs (you know, the ones that burn out in a month and are dimm as sh*t but cost 20$ instead of 300$). Those things are made of A the bulb and B the bulb's circuitry. Bulb's circuitry is what sends the ok to your projector. If you can find a way of taking out the nasty bulb while keeping the circuit active, you have yourself a bypass at no risk to your projector. At 100W for the LEDs, you may even be able to use the projector as a powersource if you can figure out what to do with the excess power.

    1 reply

    The only problem I would with using the projector power is the 1000V start up voltage - I didn't want to mess with that too much and burn out any LEDs.

    I'm gonna need some help with this

    pin1 is at 12.6V at startup for aprox 3-5 seconds

    pin2 sits at 3.3V

    pin3 stays at 0v (this one presumably gets a signal from the ballast if the light fails to strike so the projector will set the error lights appropriately)

    pin4 stays at 12.6v for a couple seconds then goes to 0v and jumps back up to 12.6v after aprox 5 secs

    I need to figure out how to convince the projector that there is a good bulb

    20171009_195626[1].jpg20171009_195700[1].jpg20171009_195720[1].jpgScreenshot_20171009-195559[1].png
    3 replies

    unless I have the pins reversed

    I just realized that I think pin 1 is the yellow wire....................not the brown wire like I thought

    that would explain what I thought was pin 3 being at 0v (ground)

    still what should I try? pin one to ground?

    I would try bringing Pin 3 (Lamp Miss) to 5V and see if the projector stays on. You would need to pull up pin 3 to 5V on the control side, and not the power unit side.

    use a load resistor to trick the projector into thinking there is a halogen bulb.

    Anyone doing this with a Panasonic pt-ae1000u? http://www.projectorcentral.com/Panasonic-PT-AE1000U.htm

    Great writeup! I was inspired to try my hand at swapping a 4000 lumen LED into a Dell 2400MP projector. But I'm stuck at bypassing the bulb check. I have this ballast: https://www.osram.com/media/resource/HIRES/333553/... Do you think I need to short pins 4/5 (type 3 - Front Projection/OSRAM mode)?

    Hope to hear back! Thanks :-)

    4 replies

    all you need to do is locate the optocouplers and short each side of it until the light stays on. on my epson i had to short the top optocoupler. all i did was drop a bead of solder on the pins on the right side of the optocoupler (The Switching side) though yours may be in a different location

    opto.png

    Thanks. I gave that a shot but it didn't work... I did some more research with the manufacturer and apparently serial communication is *optional* but it looks like Dell went that route.

    Like psron said, below, there's not really a way around that.

    Unfortunately, this appears to be one of the projectors that uses serial data to communicate to the ballast... the boards actually talk to each other, exchanging data.

    There's no easy way to bypass that one...