Eric is bang on the money for this one.
Put a single white beam into a prism and out the other side you get the full colour spectrum. So turn that around and put different frequencies of light into a prism and it acts as a dichro and you get a single white beam out, simple. Telecommunications companies use this method i believe with RGB light and some company has patented and copy righted etc...
Were not dealing with a laser company here were dealing with a big multi billion pound company, just bare that in mind.
So using the above principle you could add different frequencies of red light into a prism or other unnamed optic and get a single red beam out. The prism would act as a dichro so each beam would be perfectly over lapped even at aperture, its the holy grail of diode combining and would produce a more powerful and better beamed red than an opsl 5w red.
Think of it like this, if you could get 10 red diodes @ 170mw each with a 1nm difference then you could combine them using the above method into a 1mm by 1.5mm beam @ 0.5mR ( or whatever spec a 2mm fl lens gives on a single mode red – cant remember off the top of my head).
So thats 1.7w in a 1mm beam
Now stack 6 sets together so you have 3 wide and 2 high that would give a beam of 3mm by 3mm @ 0.5mR at a power of 10W in a tiny beam. Then cube another 6 in and you have 20w of red in a 3mm by 3mm beam at 0.5mR, beat that coherent!
It all sounds so easy and perfect ... yeah sadly its a right cunt to actually get to work.
Each diode would need to be temperature controlled to probably 0.1 degree
You can only set it up using a spectrometer, from testing every diode gives a slightly different frequency. At the same current and same temp diodes can be as far as 3nm different. you have to adjust the temp of each diode until the required frequency is reached, the spectrometer would need to be accurate to around 0.2nm to make it work properly.
The most nm difference i managed to get out of a diode with a temp range of 0-40 degrees was 8nm, so that means you could only combine 8 diodes into the prism.
The further the distance between the diodes and the prism the easier it is to get it to work. Im sure someone could do some sums and work out the distance and angle each diode would need to be. But the closer to the prism the smaller the angle, the further away the bigger the angle.
If you have 8 diodes on an aluminium base plate, at one end a diode is cooled to 0 degrees then 150mm away the furthest diode is heated to 40 degrees, with some more diodes in the middle at different temps. Managing the temp of the base plate is tricky. Not looked into this in the slightest.
When you first turn the unit on each diode is at the same temp, room temp, so out the prism would be lots of little dots. You would need a way to turn the diodes on once each has reached is pre set temp. So you couldn’t turn the a projector on and start using it straight away.
Ok now for the fun part some real world testing
3 diodes, 2x 110mw 683nm diodes tec’ed and a 660nm diode un tec’ed – all single mode. One 638 tec’ed to 0 degrees to other to about 20 degrees, no idea on what wavelength they were giving out as i don’t have a spectrometer.
All 3 diodes had 2mm fl lens on them
As you can see beams all tight together
The combining optic isn’t important at the moment, but i was able to get all 3 beams perfectly over lapped into a 1mm beam
The really fun part is dicking around with the temp of one diode and watching it slowly drift across the wall, kept me amused for hours.
Interesting shit, i just need a really really accurate spectrometer to complete the project now