What current are you running those 71s at?
What current are you running those 71s at?
I am running them at 1025 mA. and the plate they sit on settles down at-1 to-2 C.
Eric;
Your spacial filter design is considerably more complicated (and adjustable) that I had first thought. I'm curious if you've experimented with other, simpler designs. In theory, once you are far enough from the diodes such that the artifacts begin to be seen around the edge of the beam, it shouldn't matter how you mask the artifacts.
I'm thinking that a simple fixed-size hole of whatever shape you want for your final beam (square or round) would work just as well as the 4 movable blades in your design. The only caveat would be that the plate containing this hole would need to be able to move in all three directions (up-down, left-right, and front-back relative to the focal point) in order to allow it to be "centered" on the beam. Moving forward or back would then carve out more (or less) of the beam.
Your thoughts?
Adam
It's only more complicated because I chose to make it that way. What I mean is that I wanted to be sure that if the method failed it would not be due to a lack of precision/stability. Nor would the adjustment flexibility prevent the maximum effect to be seen. I really would not be too excited to use a fixed size/shape(round) spot because depending on the quality of your alignment and the precision of each diode's focus the best hole would be difficult to size and nailing down the right size would be iterative. I really like the razor blade approach. Fabricate a block with an oversize mounting hole that allows lateral and fore/aft positioning on the optical deck. Then place a few small threaded holes around an oversize (way oversize) central hole. Slide each blade into position (you need 4) and then tighten a screw to fix the position.
To see how easy this is just slide the blade free hand near this focus and watch how well you can improve the beam with even this crude level of precision. The really big projector I am working on now will use this approach and I will apply it to the red AND the blue (not sure about the green yet).
I don't think you'd need to get the hole size perfect on the first try. I'm thinking you could start with a hole that is only slightly larger than the beam at the focal point and them move the hole towards the final lens to effectively shrink the size of the hole.
In other words: a relatively large hole (relative to the size of the beam at the focal point) is actually a very small hole compared to the final beam diameter just prior to the final optic. So moving the hole closer to that optic will mask more of the original beam (including any artifacts).
As for the differences between each diode's individual focus, that problem is the same regardless of which method you use to mask the beam, so I don't see why it would be an issue.
I don't dispute that it is easy to adjust the individual blades in your design. I'm just trying to simplify it even further. Assuming that the artifacts are more or less symmetrical (which they appear to be in your system, just as they are in the diode pair I've been working with), then moving the aperture forward and back should accomplish the same thing, assuming you've already positioned it correctly on the center of the beam (up-down and left-right).
Adam
a couple of questions mr planters , what do you use to control those giant tec's , also you have the red last why ? where you limited by filter choice ?
thanks
When God said “Let there be light” he surely must have meant perfectly coherent light.
Ι have more or less come to the same conclusions buffo. I'd just add a fourth, "tilt" function to the three directional movement you specify above, just to compensate for any sort of minor tolerance in the system, in the event that the beam is not exactly parallel to the baseplate. Or at least make the plate around the pinhole thin to minimise the chance of the beam not coming in perfectly square
What I have found very usefull is an XYZ adjustable mount that looks like this http://www.photonlexicon.com/forums/...5&d=1333573639 and allows for approx 1cm travel on all axis with a resolution of higher than 80tpi.
I also second your idea of the spatial filter hole not being very small. I have seen good results with a 445 diode and a hole of around 400micrometers by moving it towards the first or second lens (no precise way to measure it, but it does look to be around that much http://www.photonlexicon.com/forums/...6&d=1333573639 )
This is just a quick proof of concept I set up some time ago, where you can see that the beam is not exactly parallel to the baseplate, but it does convey the point http://www.photonlexicon.com/forums/...9&d=1342125279
Plus, i heartily second planters's point of using long fl lenses, that give you more "room" to work in and make the whole kepplerian setup more "forgiving"
I think that beam shaping by any means we can think of and matching all laser sources in a projector is the way to go (and not just chasing absolute powers)
"its called character briggs..."
If you watch his video (link above), he explains that after testing several different dichros he discovered that the low-pass filters had far less losses. Even when he compared the expensive Semrock dichros to the cheaper one's that Edmund's Optics sells, there was little difference in performance so long as he passed the long wavelenth (red) and reflected the shorter ones.
I see your point. But I'd like to make it as simple as possible, so adding another axis of freedom is something I'd like to avoid. Making the beam parallel to the baseplate should be easy enough anyway, since there will probably be a couple bounce mirrors in the path before the spacial filter. (That seems to be the only way I'll ever fit all this stuff inside a standard-sized case...)
Very interesting! What does something like that retail for?What I have found very usefull is an XYZ adjustable mount that looks like this http://www.photonlexicon.com/forums/...5&d=1333573639 and allows for approx 1cm travel on all axis with a resolution of higher than 80tpi.
Yeah... Given the size of the stack in Eric's setup above, I think even a 1 or 2 mm square hole would have worked just fine.I also second your idea of the spatial filter hole not being very small. I have seen good results with a 445 diode and a hole of around 400micrometers by moving it towards the first or second lens (no precise way to measure it, but it does look to be around that much http://www.photonlexicon.com/forums/...6&d=1333573639 )
I've been thinking about the optimal direction of movement for the pinhole quite a bit lately. It's obvious that moving it to either side of the focal point will mask off more of the beam, but I think moving it towards the output lens will allow the portion of the beam that diffracts around the edge of the hole to completely miss the output lens.
Originally I was thinking that moving it towards the input (or 1st) lens would be better, because the path for the light that diffracts around the edge would be longer. (Further away from the output lens) But thinking about it, that light will end up diffracting back towards the middle of the beam as it strikes the output lens... I dunno - I guess I need to set it up on a bench and just play with it.
Yeah, your setup in that picture above has a rather short focal length! I like Eric's layout better. I think you really want 6 inches or more between the lenses, with 9 or 10 probably being the sweet spot. Of course, I don't know if I'll have that much real estate to spare on my optical table... Sigh.This is just a quick proof of concept I set up some time ago, where you can see that the beam is not exactly parallel to the baseplate, but it does convey the point http://www.photonlexicon.com/forums/...9&d=1342125279
Plus, i heartily second planters's point of using long fl lenses, that give you more "room" to work in and make the whole kepplerian setup more "forgiving"
Agreed, provided you are already up beyond the 5-6 watt range. Eric's projectors are rated for nearly 10 watts, so he can afford to lose a ton of power via beam shaping and still have plenty of power left over.I think that beam shaping by any means we can think of and matching all laser sources in a projector is the way to go (and not just chasing absolute powers)
But if I were building a 3 watt projector (1 watt each of red, green, and blue), I don't know that I'd worry so much about it.
Adam
I had originally approached the sizing issue with exactly the same idea you are presenting and it will compensate to some extent, but the gain in matching the size is offset somewhat by a softening at the image plane. The issue with the compound focus is somewhat unique to the Mitsu diodes. The sensitivity to rotation is very high and when this rotation is foreshortened by the cylinders the residual effect is to produce a little rhomboid/parallelogram distortion. Furthermore, the distance of a given diode to the first cylinder is also very critical and the result of both of these is to require a compromise position for the spatial filter. This means what you might perceive as a slight de-optimization of the best focal plane might be a significant distance for a particular diode.
I'm not trying to be argumentative. And, obviously this is a little speculative, but when you look at the far field spot of a quad (or greater)at a significant fraction of maximum operating power ( I use a mat-anodized surface) the spot is a BLOB. I have focused, adjusted and readjusted, but a classic square or circle eludes me. The focus at the pin hole reproduces a much smaller, but otherwise identical image of the remote focus on the screen, It is therefore that shape that needs to be isolated . If the spot shape mattered more than power then a closely sized pinhole would get my vote. But, the flexibility to custom fit the blades to balance my preference for power vs spot size...priceless
You posted as I was typing (a very slow process). That's what is so nice about TECs, they are a pure resistance. Just try a few different driving voltages with an adjustable bench PS for a given TEC (resistance is a spec.) then measure your equilibrium temp near the diode. When you're satisfied then replace the varible source with a fixed source. A cheep switched mode with sufficient current capability is fine.a couple of questions mr planters , what do you use to control those giant tec's , also you have the red last why ? where you limited by filter choice ?
thanks
The red is FIRST. The filters I am recommending are long wavelength pass beam combiners. The first filter(there are two) passes 638 and reflects the 532. It is called a 552 nm cutoff filter. The second filter reflects the 445nm and is called a 466 cut off.
With long FL lenses the tolerance is not extremely tight and just allowing the block on which the filter sits to be manually slid L/R and back and forth before screwing down is sufficient. Seperating out these two DoF makes the filter a lot simpler. Also remember. all I did to filter the lateral component was to slide a small blade into position with my fingertip.The only caveat would be that the plate containing this hole would need to be able to move in all three directions (up-down, left-right, and front-back relative to the focal point) in order to allow it to be "centered" on the beam. Moving forward or back would then carve out more (or less) of the beam.
Last edited by planters; 10-20-2012 at 14:06.