Pure speculation, but I would guess that the existing de-speckle hardware ended up being smaller (or cheaper) than a fiber optic solution.
Yes. Exactly.From what you described I can have the initial beams and their collimators before the fibers smaller in diameter than the DLP/LCoS chip and use the collimator after the fiber to get the correct diameter?
Yes, you can re-collimate it, and it will be very close to the original beam (not perfect, but close enough that you won't notice).whether I can re-collimate it and expect the same laser-like collimation
You always want to use an achromatic lens when dealing with a full-color beam. (Achromatic lenses are specifically designed to minimize chromatic aberration.)Won't a single lens for a combined laser beam refract each wavelength differently and properly collimate only one of them?
No idea. A good starting point would be to find out how long the fiber was that VDX used in the pictures he posted above though... The number of bends in the fiber may also play a role.how would you determine how long of a fiber you need to get a proper "homogenization".
Adam
... for the fiber length - I'm using this fiber-coupled IR-diodes with fiber lengths of around 50mm to 1m without notizeable differences in behaviour ... and too - didn't see any difference, if the fiber is straight, or rewound several times ...
Viktor
*** EDIT ***
My idea is, that this "homogenization" through the fiber will kill any coherency, so the beam behaviour should be the same, as if it were emitted by a LED (if you get the same energy density into the fiber).
The good optical quality of the recolimated beam should be due the small output diameter of the fiber, what's much better than any LED ...
Last edited by VDX; 07-28-2018 at 11:19.
You suggest I put the fiber directly in front of the laser dye, but I can't do that as I have 3 laser diodes for each color channel, not one. I could use a separate fiber for each diode but then I would need to combine those.
Should I go that route or perhaps I can just use the regular dichro combining but have the collimators for each diode focusing at a single point after the dichrois where I could put the fiber? Would that produce a different result than having the fiber exactly in front of the laser dye without any collimating lens between them?
Also do I understand correctly that fiber will affect divergence because the diameter of the fiber will become the new illumination source diameter? I don't know the dimensions of the laser emitters, if they are significantly smaller than say a 0.05mm fiber the difference in the emitter diameter might become noticeable here where a 1920 pixel row has to remain in focus from 50cm to 4m. Sure it will be smaller than a LED but I'm trying to preserve the optical properties of lasers here.
The ability of the projector to be in focus over a large distance comes from it being illuminated by a small source. If you insert a diffuser, that source size and location in the optical system will be different from what was originally designed for the projection optics. You probably will loose the focus range and a whole amount of brightness on the screen. Also, there are two types of laser speckle in a projection system; objective and subjective. You may be able to kill objective speckle with an optotune or diffuser wheel, but if the diameter of the source is very small (as seen when looking at the projection lens from the screen), you'll still get subjective speckle. Removing subjective speckle from very small projectors is extremely difficult.
Well, I'm not sure what you mean by source and small here. When you look at the projection lens you see the reflection of the DMD or LCoS chip reflected from a prism, and less commonly the DMD itself. The DMD in a 1080p pico projector is 0.47 inches diagonal, 0.66 inches for home cinema and cinema projectors. The size isn't that different between DLP pico and larger ones for the same resolution but I don't know if it is different enough for this. The laser beam illumination source has to be magnified or collimated to the size of the DMD to illuminate it fully yet not require very large lenses for its reflected beam. However, the de-speckler or fiber may be put in the illumination system before it is collimated/magnified to the size of the DMD to save cost and space, if that'll work.
With LBS scanned laser projectors it's a different story and the beam diameter is small enough to be scanned 45 dergees on both axis with a 2mm diameter mirror. I don't know if that's what you're referring to by a very small projector or same issue will apply to DLP/LCoS system I described above. If it will, please explain why. Thanks.
Last edited by joec; 07-29-2018 at 06:11.
... you can "combine" the three diodes by focussing them to the entry end of a single fiber -- here's a "combiner" with 6 blue diodes, focussed on a spot of 0.3mm ... with shorter focal distance, and your three diodes closer together, it shouldn't be a problem to reduce the combined focus diameter to 0.1mm and inject into a 0.2mm (or thicker) fiber:
Viktor
*** EDIT ***
... or you set 3 fibers to every diode and combine the output of the fibers by focussing into another fiber - here's a sketch showing the "combination" of many fibers:
Last edited by VDX; 07-29-2018 at 06:26.
That will eliminate the need for dichros so sure, I could do that.
Can you please respond to this question?
"Also do I understand correctly that fiber will affect divergence because the diameter of the fiber will become the new illumination source diameter? I don't know the dimensions of the laser emitters, if they are significantly smaller than say a 0.05mm fiber the difference in the emitter diameter might become noticeable here where a 1920 pixel row has to remain in focus from 50cm to 4m. Sure it will be smaller than a LED but I'm trying to preserve the optical properties of lasers here. "
And how small can I expect to focus the beam with a collimator on a diode housing VS having a fiber directly in front of the laser cavity and focusing the end of the fiber instead?
How small of a fiber do you think I can and should use with the single mode Osram and Mitsu diodes?
With <100mW focused diode beams should I use plastic or glass fibers?
Thank you again.
... most of my fibers have glass cores - but with powers below 5Watts polymer fibers could be better.
The emitter of the SM-diodes has something like 1x100 microns, but most energy is centered around the centre in 30-50 microns - so a 50 micron fiber, placed in 5 microns distance will catch around 80-90 percents.
With diodes in housings with emitter window refocussing with a small lens would be the better option - expect a focus diameter of roughly 50microns with a short FL lens (at or below 10mm FL).
With my IR diodes out of a 105microns fiber and a 10mm lens and a geometrical arrangement with same distances from fiber to lens to focus (around 20mm) the "refocussed" spot diameter is roughly at 0.1mm too.
The divergence of the beam out of the fiber depends on the NA (Numerical Apperture) - don't know exactly, if the NA with thicker cores should be better ...
Viktor
Shouldn't I go with thinnest fiber (0.05mm) to have the emitter "source" (end of the fiber) be as close to a laser-like point source as possible?
Will that simply affect the distance the collimating lens will be from the end of the fiber? Or will it also affect the divergence I can get from the final beam after collimating lens?The divergence of the beam out of the fiber depends on the NA (Numerical Apperture) - don't know exactly, if the NA with thicker cores should be better ...