Projector mod conversion - UHP lamp to laser... RGB? - help!

[farbe2]

Thats very nice!

I am quite interested in the optics used. Seems like you converted one Osram array to white and used another one to have blue? I also see several lenses and a "diffuser" wheel? Do you have problems with speckle as you use blue directly? Do you use a despeckler?
How is the Phosphor holding up? Which one did you use? Why not a allready manufactured phosphor wheel (should be more efficient)?
How do you do the black frame detection? Whats your Laserdriver?
SOO many questions, please dont let me die dumb. Thanks for sharing your project!!

I think Water-cooling is the way to go for a upgrade.
It could (if designed correctly) be much less noisy than the cooling for the UHP lamp. Quickconnects for the (small) waterlines are easy and cheap. This gives the possibility to have the radiator a few cm/m away from the PJ, reducing the noise at the watching position even further.

As this would make the PJ much colder inside, one could think about reducing the existing fans speed (or number). However one would need to program a little board that emulates a fan with the right speed. As most PJ are checking the fans for speed/function. Design of this board would be easy for me.

For wavelength: 410/405nm is impossible for the Osram array. Its likely around 445nm. How hot does your laser run? These modules drift quite a bit with temperature. Increase it if possible, this would give you a higher wavelength blue thus less violet. Just dont run it too hot, check the data sheet. Otherwise 455nm is the way to go.

I dont have any PJ related measurement equipment, only a spectrometer and Lux meter but i hope i can still develop something that i could use for my VW270.

 

[Harcoreta]

Thats very nice!

I am quite interested in the optics used. Seems like you converted one Osram array to white and used another one to have blue? I also see several lenses and a "diffuser" wheel? Do you have problems with speckle as you use blue directly? Do you use a despeckler?
How is the Phosphor holding up? Which one did you use? Why not a allready manufactured phosphor wheel (should be more efficient)?
How do you do the black frame detection? Whats your Laserdriver?
SOO many questions, please dont let me die dumb. Thanks for sharing your project!!

I think Water-cooling is the way to go for a upgrade.
It could (if designed correctly) be much less noisy than the cooling for the UHP lamp. Quickconnects for the (small) waterlines are easy and cheap. This gives the possibility to have the radiator a few cm/m away from the PJ, reducing the noise at the watching position even further.

As this would make the PJ much colder inside, one could think about reducing the existing fans speed (or number). However one would need to program a little board that emulates a fan with the right speed. As most PJ are checking the fans for speed/function. Design of this board would be easy for me.

For wavelength: 410/405nm is impossible for the Osram array. Its likely around 445nm. How hot does your laser run? These modules drift quite a bit with temperature. Increase it if possible, this would give you a higher wavelength blue thus less violet. Just dont run it too hot, check the data sheet. Otherwise 455nm is the way to go.

I dont have any PJ related measurement equipment, only a spectrometer and Lux meter but i hope i can still develop something that i could use for my VW270.

Thanks a lot!

I'm using optics from two "cuashios" M14O projectors, and harvested lens, mirrors and motors instead the LDs xD. Also using a camera 37mm close up lens +8, here you have a very advanced photo detail... xD


I'm using now a nichia NUBM36 instead the Osram for the yellow color, It's a bit more efficient. For white color you only have to mix the yellow with blue. You can see a dichroic filter where the yellow color is reflected and the blue goes through it.

This diffuser wheel is just made for prevent the sperckle and do the blue color more homogenous , it's working very well. I made it with a 58mm UV filter camera, and applied a glass etching cream on one surface.

I've purchased the phosphor powder from this alibaba seller

https: //www.alibaba.com/product-detail/LED-phosphor-LED-phosphor-powder-LED_60617755571.html

Purchased 10gr of:
TKY-4155-C
TKY-4255-C
TKY-4255-CC
TKY-4354-C

(around $150 :O)

After some testing I'm using a mix of 3 parts of TKY-4155-C + 1 part of TKY-4255-C, and have the perfect color temperature for my projector.
About conversion light efficiency I'm sure it have room for improve it, I'm not using an optical silicone, It's just a clear high temperature resistant silicone, I was unable to find it. It's the same problem if you want to buy a manufactured phosphor wheel, very difficult to find one.

About laser dynamic lighting please wait for tomorrow , I don't have the code where I am now, you will see that it is very simple. The power supply are two DPM8608 working in constant current mode.

A custom wartercooling loop is the best way for move all the heat outside, and as you say it is very silent. I installed not a custom, it's a corsair AIO. An AIO is worse for assembly, disassembly and maintenance, but I was looking for a solution without not much complication xD

For the Osram array, I'm running it powering only 2 rows and with a current around 1A, it is rated to 2.3Amps max. I suppose the problem is only its wavelenght, as you say probably near 445nm. I hope to receive the nichia NUBM38 next week for do more testing

 

[HDMaster]

I am really impressed

 

[farbe2]

Hello Guys,
I measured my PJ to get more informations about its light content.

As you can see, each RGB color has a quite broad spectrum. I didnt expect that! My PJ is rated at BT2020 operation, but i dont have measurement equipment to verify that.

Blue starts at around 435nm and ends at around 485nm. Its peak is at 445nm with a second peak at 475nm.
Green starts at 515nm, peaks at 550nm and ends at 575nm.
Red is even broader, it starts at 595nm and extends into the infrared band.

So it seams like your Blue looks to violet because its to close to the edge of the colorspace. Its to "pure" and to low. It would be interesting to measure the spectrum of the UHP bulb to see if the filters are only bandpass or if they alter the spectral distribution of the input light. This could help to better tailor the phosphor converter to better suite the internal filters.

 

[Harcoreta]

Thanks for sharing this info
Yesterday was playing a bit with my xrite colormunki photo, I've purchased it last week and have not spent much time using it, and finally found the spectrometer graph in the HCFR program



yes! the laser is peaking at around 445nm. Seeing the CIE triangle gamut I assumed that the blue laser was further away.
In conclusion... I want my new nichia array now! xD

This is the code for powering dynamically the lasers.
You need....

A video player that works with ffdshow wideo decoder, like "MPC-HC" (64x)

https:// github.com/clsid2/mpc-hc/releases

"Avisynth" plugin

http:// avisynth.nl/index.php/Main_Page

"ffms2" wrapper library for Avisynth

https:// github.com/FFMS/ffms2/releases

Extract the "ffms2.dll" to a folder (using C:\brillo), activate "Avisynth" in "ffdshow wideo decoder" and paste this code.



It is going to calculate the average luminance of each frame and log it to a file called "output.txt". The recorded values are going to be between 16 (Black frame) and 235 (white screen)

And finally create a new powershell script and paste this code (change voltage and current values for your application)....
I'm sure the code can be improved, and some better programming language could also be used

$port1.Close()
$port2.Close()
$port1 = new-Object System.IO.Ports.SerialPort COM8,9600,None,8,one <#COM PORT for Yellow Power Supply#>
$port2 = new-Object System.IO.Ports.SerialPort COM10,9600,None,8,one <#COM PORT for Blue Power Supply#>
$port1.Close()
$port2.Close()
$port1.Open()
$port1.WriteLine(":01w11=2500,") <#Setting current in miliAmps for Yellow PS#>
Start-Sleep -milliSeconds 200
$port1.WriteLine(":01w10=5020,") <#Setting Max Volts for Yellow PS, "5020" indicates the voltage setting is: 50.20V#>
Start-Sleep -milliSeconds 200
$port1.WriteLine(":01w12=1,") <#Activate output in Yellow PS#>
Start-Sleep -milliSeconds 500
$port1.Close()
$port2.Open()
$port2.WriteLine(":02w11=0990,") <#Setting current in miliAmps for Blue PS#>
Start-Sleep -milliSeconds 200
$port2.WriteLine(":02w10=5300,") <#Setting Max Volts for Blue PS, "5300" indicates the voltage setting is: 53.00V#>
Start-Sleep -milliSeconds 200
$port2.WriteLine(":02w12=1,") <#Activate output in Blue PS#>
Start-Sleep -milliSeconds 500
$port2.Close()
Start-Sleep -milliSeconds 500
$port1.Open()
$port2.Open()


do {

$data = get-content c:\brillo\output.txt -Tail 14 | Select -First 1
$dato = [math]::Round($data,3)
$dato

<#these commands open the log file "output.txt" and take the value in the row 14 starting from the end, try another number for best sync with your projector#>


$lastModifiedDate = (Get-Item "c:\brillo\output.txt").LastWriteTime
$dateA= $lastModifiedDate
$dateB= (Get-Date).addseconds(-3)


if ($dateA -lt $dateB) {
$port1.WriteLine(":01w11=2700,") <#Set max current in miliAmps for Yellow PS#>
$port2.WriteLine(":02w11=1060,") <#Set max current in miliAmps for Blue PS#>
}

<#if there are no active playing on the last 3 seconds, the laser are going to work at Max power#>


elseif ($dato -gt 15.9 -and $dato -lt 16.05) {
$port1.WriteLine(":01w11=0600,") <#Set minimum current in milliAmps in Yellow PS for a Black frame, this must be less than its laser emitting threshold#>
$port2.WriteLine(":02w11=0250,") <#Same for Blue PS#>
}

elseif ($dato -gt 16.05 -and $dato -lt 16.2) {
$port1.WriteLine(":01w11=1900,") <#Current in milliAmps in Yellow PS to start to light up the laser#>
$port2.WriteLine(":02w11=0800,") <#Same for Blue PS#>
}
elseif ($dato -gt 16.2 -and $dato -lt 16.4) { <#make your illumination power curve #>
$port1.WriteLine(":01w11=2100,")
$port2.WriteLine(":02w11=0865,")
}
elseif ($dato -gt 16.4 -and $dato -lt 16.8) { <#make your illumination power curve #>
$port1.WriteLine(":01w11=2300,")
$port2.WriteLine(":02w11=0930,")
}
elseif ($dato -gt 16.8 -and $dato -lt 17.4) { <#make your illumination power curve #>
$port1.WriteLine(":01w11=2500,")
$port2.WriteLine(":02w11=0995,")
}
elseif ($dato -gt 17.4 -and $dato -lt 260) {
$port1.WriteLine(":01w11=2700,") <#Set max current in miliAmps for Yellow PS, normal operation#>
$port2.WriteLine(":02w11=1060,") <#Same for Blue PS#>
}

} until (start-Sleep -milliSeconds 120) <#configure a delay to start the loop again#>

 

[farbe2]

Thats what i was thinking. That does not fit my needs (mostly youtube sometimes Netflix).
I would go for a video grabber and a raspberry pi. The ambilight people do it this way.
Maybe one could also hack the firmware of the PJ to output some signal on a unused pin. The mainboards for the PJ are almost always identical per product line, so if the product line offers a PJ with a iris changes are good that the mainboard already has the capability to calculate average brightness and output it.

 

[Harcoreta]

I was considering your approximation using a raspberry pi. It will be more flexible and will work with any kind of content.
I've been looking some ambilight codes but not went much deeper with it. I think it is your best way to go
Taking signals from the projector is possible but a seems bit complex to implement.

 

[farbe2]

Taking signals from the projector is possible but a seems bit complex to implement.

You are right! That would be really complex! I think with a PJ with a Iris, there should be no problem using a Arduino to measure the control signal and output it to a Laser driver.

I already have a NUBM31 and a ColorDrive6, this should be enough to test stuff. I just need to get my hands on some phosphorus.
However I am quite hesitant to break my expensive PJ that I use almost every day.
So my plan of action would be to buy a (cheap) broken PJ to experiment with. If everything works out, I will rip my normal PJ apart.

 

[Harcoreta]

I have bad news with the experiment, I have a new issue and it's very difficult to solve easily. The lenses are collecting dirt with the laser operating at high power, I can measure about 10% of light loss in about 10 hours of operation. I can see small spots of dirt in the lenses where the laser passes through. The dust particles are burnt due the high power laser beam and then it stiks to the lenses. The only solution is make a totally sealed laser+lenses package and I don't have the tooling to make it.
Another option is buy a laser engine for other projector that can fit into my projector housing, as HDMaster said a few posts ago, I'll take a look...

 

[HDMaster]

Hi Harcoreta,

Thank you for your update.
Seems that it is highly important to seal the whole lightengine.

The laserengine comes from the EPSON EB-L630U.
Part# 1860230 Reference# B0030, LSM ASSY,HA25,PH,AS

The EB-L630U has 6200 Lumens.

EB-L400U : 4500 Lumen
EB-L510U: 5000 Lumen
EB-L520U: 5200 Lumen
EB-L730U: 7000 Lumen

The price varies between 500,- and 1200,- Euros.
It is independent from the power.

Two weeks ago i installed the EB-L630 in a company and i was really flashed by the light output.
You can dimm it down to 50%.



I am just waiting for the 6000 Lumen version.

May the force be with us !

 

[kecked]

An idea but it’s just that. How about a fan so the dirt can’t settle. I think with high power lasers a charge can build up at the spot and attract dirt. A fan would make it harder to collect this dirt. Humm or easier because I just increased mixing…..

second idea…again a guess. Put one of those anti static things on the lens next to the spot. Or tin foil with a charge to attract the dirt stronger?

seailing the optics is really the best idea.

 

[Harcoreta]

Thank you guys!

Hi Harcoreta,

Thank you for your update.
Seems that it is highly important to seal the whole lightengine.

The laserengine comes from the EPSON EB-L630U.
Part# 1860230 Reference# B0030, LSM ASSY,HA25,PH,AS

The EB-L630U has 6200 Lumens.

EB-L400U : 4500 Lumen
EB-L510U: 5000 Lumen
EB-L520U: 5200 Lumen
EB-L730U: 7000 Lumen

The price varies between 500,- and 1200,- Euros.
It is independent from the power.

Two weeks ago i installed the EB-L630 in a company and i was really flashed by the light output.
You can dimm it down to 50%.

View attachment 73612

I am just waiting for the 6000 Lumen version.

May the force be with us !

I have some ideas to seal the lightpath, but It has to been made in metal due the high temperatures. Only have a plástic 3dprinter not suitable for that, and a hand made lightpath using aluminium could be a nightmare :/
I was looking for those Epson projectors, seems very interesting, they are using a nichia NUBM31/34/36 arrays.

Are there any chance to get one of those laser engines?
I saw some Epson projectors in eBay but cheapest are about 1500€
Please share your advances when you have the laser engine

An idea but it’s just that. How about a fan so the dirt can’t settle. I think with high power lasers a charge can build up at the spot and attract dirt. A fan would make it harder to collect this dirt. Humm or easier because I just increased mixing…..

second idea…again a guess. Put one of those anti static things on the lens next to the spot. Or tin foil with a charge to attract the dirt stronger?

seailing the optics is really the best idea.

Already have a fan facing to one dichroic mirror, and its collecting more dust xD.
Yes, the only way to go is sealing the entire laser lightpath.