True green diode laser rgb build

[kecked]

Well I got one of the 50mw true green laser diodes. Going to match it with a 100mw 450nm blue and a 110mw 638nm red. Just need to find a dichro to pass red and blue that reflects 520nm green. Seems most dichros transition round this wavelength. I want to goin order red through, reflect blue reelect green but I might be forced to dothe traditional rgb route if I can't find good dichros.

Build will use Dave's micro mounts and p3 drivers. I'm hoping for 200mw white light with perfect bem profiles and matched on off times. Ie about the same as an argon hene combo through a pcaom. Quality for this build rather than power. If it works I can knife edge 4 units and finally call it a day.

[Stoney3K]

May I dare ask how many kidneys you had to sell to get your hands on these diodes?

[Psi]

Probably both, he has 24 hours to play with his new projector before he goes into complete renal failure.

[kecked]

I can't say costs or source but it was not what you might have expected. I think you'll see these on the market in the next year at most but in the 530nm range. I wanted the 515nm range because it is so argon like. Hence 638/520/450 is about as argon as I will ever get from diodes. The 638 was a compromise between krypton like and brightness. As I said I'm after quality now rather than power. I have 2.5w in the living room now but the beams are not as good as I'd like. The modulation on dpss will never be good and I can't afford opsl. Thus this little box should and I'd unline should give excellent beams. The question is still will the rise and fall times match. I may still need to employ some kind of noise gate that senses the signal and doesn't turn on until a certain level is reached and that turns off as soon as a threhold is reached. BADPIPS drive might be the ticket for that function. That could eliminate the trails issue. I expect to find capacitance related problems in the driver that are the real cause of the trails. This means I might have to use telecom drivers to drive the lasers. These modulate in the ghz range so those issues should have been solved and since I need no more than 100ma I don't need a high power FET driver with intrinsicly high resistance. So that means that a little stray capacitance will not lead to large RC delay values. But then I'm dealing with loading on the inputs. It is a game I'd rather leave to the experts. Hence telecom drivers. Gee I am starting to sound like Steve!

Pics will come but don't wait I expect this to take me a few months to get around to finishing and I still have to find the right dichro stock. For those curious the diodes look just like any other 3.8mm diode. Nothing special to see. I have to give props to the guys on LPF for the machined copper diode holders. They are really nice.

[mixedgas]

When he sees the cost of the Semrock dichro, he'll have to sell the OTHER kidney.

You'll loose a lot of red/blue power, but Omega Optical Inc is your friend on this one. The president of the company shows up as Ebay Seller bjomejag. He usually has his overstock biomedical dichros on ebay. No one will accuse him of ever making a sharp cutoff dichro, but he's cheap and dumps his surplus on fleabay.


Steve

[kecked]

I wonder if using a full wavelength PBS cube might be a better way to go. I can combine the red and blue using a normal dichro and then use the cube to combine the green with the red and blue.

[swamidog]

i was about to suggest it. a broadband pbs cube may be a nice cheap way to do it. i've used them to combine wavelengths.

I wonder if using a full wavelength PBS cube might be a better way to go. I can combine the red and blue using a normal dichro and then use the cube to combine the green with the red and blue.

[kecked]

http://www.semrock.com/FilterDetails...id=LM01-552-25 $225.00 not cheap but hey this is a cutting edge build.

steve how does this one look. I can use it to combine red and green and then just use a stock one to combine in the blue. It's still half the loss of doing it the traditional way.

I think the cube idea might be the cheapest way to do it though. Anyone got those nice 5mm full wave cubes that were in the group buy last year? I had a few but gave them away to others.

[mixedgas]

The Semrock part is not bad.

What exactly is the center wavelength of the green diode on its test sheet?

Steve

[kecked]

Thats the issue I'm not sure. I am assuming somewhere between 510nm and 520nm. I figure if I am away from the edge it will not matter and I think the same can be said of the blue dichro. I think loss will be low through the second dichro. When you only have 50mw you need to conserve it. I sent Bob on ebay a message with my wavelenghts. Maybe he has what I need in stock. At $20.00 each that is not bad. some of that stock is 98% reflect 96% transmit. That is amazing stuff.