How to create super bright laser beams

[logsquared]

I considered two solutions if this proves more than trivial and I think it will. The first is to use a single scanner/amp controlled by the same modulation voltage signal that controls the red laser. The angle will probably be only a few mrad and after asking a few experts, I'm confident this should work. The second and I believe more elegant solution it to use two prisms, each made of a different glass to create what is basically a achromatic pair. The first, high dispersion prism corrects the converging beams and the second negates the deviation without undoing the convergence correction introduced by the first. I then confirmed that prisms with the right glasses are commonly available and the cost only effects how good the compensation can be. The setup would require the first prism to over correct the incoming beams depending on the level of dispersion of the second glass and then the second prism is inserted and is adjusted to eliminate the deviation of the pair. A little back and forth experimenting will be needed, but the result should be stable and essentially perfect.

I had thought of a few ways to mitigate the problem. i had considered the active route as well. Either just as you said or using a galvo for analog "blanking" by cutting into the beam with a flag. I had also thought of "digital" control. Basically turning the diodes on/off independently. Four diodes would give 2 bits of control.

I am trying to wrap my head around the extra prism of different glass

I don't entirely get it? It seems that it could be done with two extra prisms. One to re-converge the spread created by the injection current shift. Then a following prism to diffract the varying beam to the same angle on exit.

Sorry, that is probably hard to follow.

[planters]

The active control I envisioned was a very small angular correction introduced by the galvo that exactly opposed the small angular movement created by the diode's red shift with current. As the red shift is proportional to modulation voltage input from the computer then why not tap this voltage and use it to control the galvo amp at the same time? But, I think the series prisms are better. The deviation created by the prism or the gross redirecting of the beam is due to its refraction, but the collimation of the converging beams is due to the dispersion. The dispersion is the variation of deviation with wavelength. Although high refraction glasses tend to have greater dispersion this property is not tied to the refractive index. So, two glasses could in theory (and in some cases nearly do in practice) have the same refractive index, but have very different dispersions. If the prisms were set up like we use anamorphic prisms "\/ /\" they could cancel any deviation, but introduce a net dispersion that we need to collimate the beams. This is how an achromatic lens works. There is no glass that I am aware of that has a negative dispersion, but there are some with about 1/10 the dispersion (essentially zero). With that glass as the second prism you basically use the second prism as a deviation eliminator and the dispersion is determined completely by the first prism.

[logsquared]

The active control I envisioned was a very small angular correction introduced by the galvo that exactly opposed the small angular movement created by the diode's red shift with current. As the red shift is proportional to modulation voltage input from the computer then why not tap this voltage and use it to control the galvo amp at the same time?

I get it. Had the same idea.


Still working the prism idea out.... I drew it on my dry erase board. Looks like it should work. It should have a very small (shouldn't hurt) lateral displacement on the output face of the last prism. It looks as though rotation of the final prism would have the effect of changing the distance from the prism where the different frequencies all overlap. Am I getting it? Still not totally grasping the need for different glass yet.

[CDBEAM]

In comparison with the very smart minds working on this issue...I am a " Very Small Dog" indeed.....but I thought I would throw in a tidbit...

In response to the comment by LS==>" I had thought it may be possible in theory to use a long glass parallelogram to accomplish the combining. The difference would be the entering beams would be exactly parallel and "stacked" next to each other when entering the glass. ".....Please see the attached pic of the internals of a " LaserKing " projector...OK....if you are done laughing....you must admit they have tried a novel beam combining approach approach !! I have no details as to what the beam propagations are....except that a long glass parallelogram is employed to combine both red and blue diodes.

CDBEAM=======>

[planters]

CD I think this would work. The lack of a correction element for the dispersion would mean that here the beams might slightly unstack as the diodes were modulated, but this would be so small that it should not matter because the effect would not magnify as the light traveled down field. You tune with an angle adjustment.

The more I think about this, the more I like it. The very difficult to determine issue is how much could this shift the beams before the angle of incidence caused too much reflection from the surfaces. The item would be a rhomboid prism.

[ams09]

In comparison with the very smart minds working on this issue...I am a " Very Small Dog" indeed.....but I thought I would throw in a tidbit...

In response to the comment by LS==>" I had thought it may be possible in theory to use a long glass parallelogram to accomplish the combining. The difference would be the entering beams would be exactly parallel and "stacked" next to each other when entering the glass. ".....Please see the attached pic of the internals of a " LaserKing " projector...OK....if you are done laughing....you must admit they have tried a novel beam combining approach approach !! I have no details as to what the beam propagations are....except that a long glass parallelogram is employed to combine both red and blue diodes.

CDBEAM=======>

Hi all,
@ CD, on your pic is just two polarized mirror, one for blue and other for red, but each diode ( on the same wavelenght ) should have a different polarization, V+H.
The final track on the picture is not correct, the red and blue join the green (at 90° )in two dichroic RGB that are hidden.

Cheers.

[planters]

CD,

I like the idea of the parallelogram or rhomboid prism as I said above, but looking at the angles required to cause the parallel entry beams to converge at the exit face might require a pretty long passage through the glass. There might not be enough clearance within available prisms before the converging and also angled beam hits a wall. Now, the beam might still be OK if it zigzagged to the other face and at exit was properly stacked. Nevertheless, I think it might still need 10-20cm within the glass and this is still a pretty long prism. I'm basing this on a calculator that predicts the dispersion and deviation in a prism that you can access at:

http://www.optical-calculation.com/i...ultat_form.php

What about a hybrid? Two prisms of identical make (same glass) in the "\/ /\" orientation, but spaced some distance apart. The parallel beams from the knife edges enter the first prism as soon as it can be positioned to intercept them and it's this prism not an initial convergence by the knife edges that starts the convergence of the beams. Then, as dictated by the achievable convergence, the second prism is positioned and properly rotated to collimate the converging beams just as they are optimally stacked. In this case, the 10-20 cm is within the air space between the prisms and the two surfaces of the first prism double the convergence angle achievable as compared to the single entry face of the rhomboid.

Now, if there is a modulation red shift of say 2nm and the inter-diode temperature step is 4nm/diode, the maximal shift of the combined beam would be 1/2 the distance from one knife edge to the next, There would be no angular variation because the red shift would effect both prisms identically. This would broaden the appearance of the beam during fast modulation, but only by 50% no matter how many diodes you stack.

[logsquared]

What about a hybrid? Two prisms of identical make (same glass) in the "\/ /\" orientation, but spaced some distance apart.

Yes yes yes. This is it! When I said "long parallelogram" I was thinking really long (1 meter) to account for the lever arm we need. It hadn't occurred to me to "split the ends" off. Dude you are brilliant. And it works! I took my single mode rgb and put it through a 30 deg prism like we use for anamorphic correction. Then I put a second prism (same glass) about 1 meter away and turned it so the original co-aligned beam and the exit beams were parallel. The RGB lines exited the second prism spaced about 2mm apart and parallel. So, if we reverse and think of the RGB lines as the spaced red lines going in parallel they should exit as one co-aligned beam out the second prism. Frequency variations should move the combined beam laterally by a small amount negating the need for angle correction. Of course we will need to us 60deg prisms to get the dispersion we need. But should work.

Thanks Beam for putting the parallelogram idea back in motion.

[swamidog]

this is an old ion trick used by laserium and others. one prism to split the lines and another prism to make them parallel.

Yes yes yes. This is it! When I said "long parallelogram" I was thinking really long (1 meter) to account for the lever arm we need. It hadn't occurred to me to "split the ends" off. Dude you are brilliant. And it works! I took my single mode rgb and put it through a 30 deg prism like we use for anamorphic correction. Then I put a second prism (same glass) about 1 meter away and turned it so the original co-aligned beam and the exit beams were parallel. The RGB lines exited the second prism spaced about 2mm apart and parallel. So, if we reverse and think of the RGB lines as the spaced red lines going in parallel they should exit as one co-aligned beam out the second prism. Frequency variations should move the combined beam laterally by a small amount negating the need for angle correction. Of course we will need to us 60deg prisms to get the dispersion we need. But should work.

Thanks Beam for putting the parallelogram idea back in motion.

[kecked]

Can you fold the path to make it smaller than 1 meter say 10-50mm by bouncing sideways on two mirror like you do in a long path spectrophotometer?
Certainly bounce loss to be considered. 1 Meter is just not practical.

I wonder if judicious use of muscle wire might be able to compensate in the opposite direction for the diode shift during modulation by passing the modulation single through the muscle wire to push on the diode or collimator.