Sextuple red build

[Doc]

I think the hysteresis problem could be improved by using dowel roll pins between bolted together components i.e. throuch the foot of the mirror mount into the base plate.

We had an issue years ago with exhaust manifold studs shearing off flush to the cylinder head on the old 5 cyl UR Audi Qattros. This was found to be caused by the manifold walking across the face of the cylinder head over many hot cold cycles; the remedy was to dowel the manifold at the centre cylinder forcing the expansion to always walk back to the centre during cool down.

Same problem with drive plates breaking on ZF auto transmissions - rectified by fitting dowels to keep everything aligned during temperature cycling.

[logsquared]

I think the hysteresis problem could be improved by using dowel roll pins between bolted together components i.e. throuch the foot of the mirror mount into the base plate.

We had an issue years ago with exhaust manifold studs shearing off flush to the cylinder head on the old 5 cyl UR Audi Qattros. This was found to be caused by the manifold walking across the face of the cylinder head over many hot cold cycles; the remedy was to dowel the manifold at the centre cylinder forcing the expansion to always walk back to the centre during cool down.

Same problem with drive plates breaking on ZF auto transmissions - rectified by fitting dowels to keep everything aligned during temperature cycling.

That was part of my theory with the combiner pictured above. And it does help to keep everything centered through temp cycling. I had thought about trying roll pins but didn't think they had the precision for a combiner. So, I used regular dowels.

[logsquared]

[QUOTE=Zoof;184024]

.. I have found through testing that even 5deg can really throw off alignment, even with all brass mounts. ...QUOTE]

So you actually already tested the all brass set screws in a brass mount for alignment issues, and the result was not better than brass mounts with steel set screws? If so it is unexpected, and I'm very interested in hearing more about what else you tested.......

My test set-up was very simple. Basically, I put a 50 watt power resistor on a small aluminum breadboard. The resistor was driven by an on/off temp controller. Next I would mount a minimum of 2 laser diodes to a brass base on the breadboard. All the diodes shared the same mount. The beams were made to reflect of a mirror ~30 feet away and reflect back to a wall where I could record there position. Next I clamped down the mounts to the brass base with brackets to minimize torque that would be present from using a screw to mount them.(found earlier that using a single screw for holding down a mount is not best practice) Next I would align the spots to overlap. I would mark the position on the wall. Then I would raise the temp of the whole assembly marking the position and temp on the wall. Then I would wait for the temp to settle back and see where the dots land. The temp in my shop was around 45deg F so cooling didn't take too long. I tried this with different mounts as mentioned above. The beams moved a lot more than I would have ever imagined they would. Its makes sense though. I have done rave type parties in the summer where my rgb was perfect at the start but badly off at the end when the temp was 95deg F.

Maybe I am going overboard here? I am just sick of opening my projectors up to align things Maybe I am making things harder than they need be. If someone knows a better way the mitigate temp drift please let me know.

[danielbriggs]

THINK ABOUT IT

i.) Price of PBS cubes >> mirrors.
ii.) Without use of a half wave plate before each PBS, mismatched beam profile -> hard to knife edge properly.
iii.) You cap yourself to 6 diodes.

[badger1666]

you know my idea WILL work because if you stack 3 beams then 3 beams again via knife edging you will have 2 sets of 3 knife edge beams going into a cube but in the + configuration....3 beams wide and 3 beams high UNLESS you use a Waveplate to get both 3 beam stacks into either 3 high OR 3 wide.All 6 beams using 3 PBS cubes have the same losses as 6 beams going through a single PBS cube except you don't have the losses of a waveplate.

THINK ABOUT IT

guess you have never built one before
3 beams x 3 beams good luck getting that on a set on scanner mirrors
unless you have some of daves super duper lens
for a lower cost build the way andy has his setup is best

[badger1666]

If you said six beams like + ,what do you expect people to think, you should have said triangle if the first place then i would not have questioned it
Also i stated andys way was good not a waist of time
What lens are you using ?
Did you also plan on using telescope lens ?

[logsquared]

I think I understand your plan. I don't see any reason why it won't work. As long as you are doing six why not just add 2 more diodes and do 2(pairs) over 2(pairs)? I don't see how the triangle stacking gains anything. If you use those tiny PBS cubes from dvd burners you could potentially make the whole thing fairly small. Also, not sure if less knife mirrors will offset the loss in fill factor from the combined "+" beams. Only one way to find out for sure..... start building

[danielbriggs]

Derek,
Get building!

[andy_con]

WTF...

have you actually put any thought into this at all?
what lens / beam sizes are you suspecting?

is there a reason no one has done it that way before kingkong?

[norty303]

Ok, so I've been thinking about how I would personally go about this, and my initial thoughts would be the following:

2 sets of 3 diodes, knife edged to give 2 sets of ...
Run them both into a PBS (one set through a half waveplate)
Use a set of correction optics/prisms to bring the wide beam a bit tighter.

Because they are overlaid through the cube, I'd have thought it might be a little more resistant to drift effects than just tightly grouped beams. Does that seem a reasonable theory? I appreciate there might be a few extra losses due to the amount of optics involved than a simple stack arrangement. I guess you might even get away with 8 diodes like this too.