optics noob.. cheap beam correction/module selection

[robinlawrie]

Ive got a choice to make, for a cheap party laser i am modifying.

ive created some lovely lumia effects reflecting the output from the galvos onto a convex mirror with transparent epoxy and broken glass.. its very pretty, however the output isnt really sufficient for my needs. its supposedly a 500mw system, but ive always suspected this isnt really true. now im running it through a lossy lumia setup its really not cutting it.

so my plan was to replace the laser module with either a 3w or 6w jobby from china. (and an uprated psu if necessary, the current one is rated for 30 watts total. )

the 3w one seems suitable, and claims a 3mm round spot on the output.. this is similar to the current module.

however the 6w one ive found for an excellent price and im tempted... nothing like a bit of total overkill.

the seller claims (no datasheet ive seen) a 3.5 x 6mm spot at the output. this is a problem since the tiny galvos only have a 6mm wide mirror, so when hit at a 45 degree angle, a 6mm beam would be too large.

so i need to decide which module to get, and was thinking about how i could marginally correct the wide axis of the 6w beam so it doenst overshoot the galvo mirrors.

im not after a perfect beam at 100m or anything, but obviously not a blurry mess either.. i might not always want to do lumia.

options ive seen/dreamed up:

3mm "pinhole" in a metal plate... ive read this can cause diffraction issues, and, obviously, will waste a ton of power.

cylindrical lens - this seems "the way" to do it, but ive no idea how to calculate the correct lens, and they seem very expensive. an ar coating would be needed to avoid diode damage. ive also read you should put them before the collimating lens in the diode housing... this means buying 3 lenses. i was hoping to do something to the combined rgb beam.

mirror curved in one axis? this was my own idea, and ive not seen it discussed. the beam is already turned through 90 degrees with a mirror, and i imagined replacing that with a machined mount which would allow me to slightly "flex" a thin mirror with a screw from behind.... it would have to be damn thin not to shatter, but i imagine the curvature needed would be minimal.

so.. do i go for the 3w and call it a day, or the meaty 6w and get creative with corrections? consider its a "budget" project.

any suggestions/ advice most appreciated.

 

[RA_pierce]

Cylindrical lenses or anamorphic prisms can improve the beam but it will not be "cheap" and it will not necessarily reduce the beam diameter at the mirror. Beam correction is meant to make the perpendicular and parallel axes of a laser's output diverge at the same angle to produce a symmetrical beam. The point is not to reduce the beam diameter but the fast-axis divergence.
The math involved is easy, but you will pay $$$ for the optimum lens configuration. You can get away with a cheap kit from China but you will pay for it in optical losses and it will not solve your primary problem.

There is a trade-off between divergence and beam diameter. In simple terms, you can reduce the spot size at the aperture but that always comes at the cost of higher divergence angle. Likewise, lower divergence requires a wider beam to start with.

My advice: Use the most powerful green laser you can get your hands on. A 200 mW 532 nm DPSS will be about half as bright as 6000 mW of blue but will generate MUCH less heat, will be safer for your eyes (this is a display laser, after all), and will cost far less. Also the beam diameter will be around 1-2 mm instead of 6 and divergence will be on the order of 1-2 mRad.
Alternatively, you could get set up with a 500-700 mW 520 nm diode laser which, with a long focal-length collimator (like a G-8 lens), will give you a better beam and will be brighter than the 6000 mW blue laser.
If your current setup is outputting 500 mW of blue light, in terms of brightness, even 100 mW of green will blow it away. Easily.
You could also try using a larger mirror.

Here's a tool you can use to look at relative brightness of different laser wavelengths. I find that it's reasonably good for an estimation of what to expect.

Keep in mind that the power ratings from a lot of China's laser product resellers are not trustworthy. I've seen listings for lasers where the power rating advertised is actually the POWER CONSUMPTION rather than the output power.

 

[robinlawrie]

Thanks for the detailed reply.. I've been reading and trying to understand as much as possible. I realised I didn't mention anywhere it's an rgb module I'm upgrading. Needs to be rgb for the lovely lumia effects.

Appreciate that you need $$$ for a good result. But given its a cheap rgb module the 3 spots will have different shapes anyway.. Unless I go to the lengths of correcting all 3 prior to mixing (not happening!). I guess I'm just after a way to squeeze the wider axis till it fits on the galvo mirror, spot quality be damned. Obviously without destroying the beam in the 10 or so metres it has to project to the cieling/back wall.

Ha.. That calculator is very interesting, and confusing! So you can never have a perfectly matched set of rgb diodes that match in both beam and spot? That's counterintuitive!

The module I'm looking at claims 1.5w 638nm red, 1.5w 520nm green and 3w 445 nm blue, and module power consumption of 65w.

According to that calculator a 400mw green would be more appropriate.. But I can always balance the outputs after I guess.

 

[RA_pierce]

Thanks for the detailed reply...

I see - that makes more sense.
In that case, it will not be possible practical to apply beam correction to the combined RGB output and get a good result.
The reason is that each laser component (presumably red, green, and blue diodes) will have different divergence angles and different wavelengths of light refract differently in the same material, so you can't apply a single correction module to simultaneously fix all 3 beams.
Achromats could maybe do it, but there's no point in getting fancy with a budget project like this.

You may be able to get away with a sort of reversed "beam expander," though. The principle of a beam expander is to reduce divergence by first enlarging the beam with a concave lens, then collimating it again with a convex lens to take advantage of the inverse relationship between divergence and beam diameter.
You could do it backwards - converge the beam with a convex lens, then use a concave lens of appropriate focal length to collimate it at a slightly narrower diameter. This should allow you to fit the beam on the mirror. I suppose the increased beam divergence won't matter much if it's going into the lumia plate. This requires only two lenses (no exotic prisms or cylinder lenses) and so can be pretty inexpensive if you go with uncoated optics.

See here: https://www.edmundoptics.com/knowle...optics/can-a-beam-expander-be-used-in-reverse

 

[robinlawrie]

Well i saw the prices of the beam expanders.. They generally cost more than all the other components of my project together! i did see a 3x sanwu one that was cheap.. however in the end i restrained myself and went with the 3w module, which claims a 3mm circular spot.. so that should be fine. 6w was probably too much anyway.. id probably have set my epoxy/glass lumia reflectors on fire. (the beam is scanned over the surface and its also rotating, so im not worried really)

in any case thanks for pointing me at the various solutions, nice to learn!

 

[atomd]

As long as you have high quality TEM00 beam you can compress / expand it using telescope. The problematic factor is divergence * spot size