Big Red Banger using 12 to 16 Opnext HL63133DG Diodes

[taggalucci]

Well this project has been a long time coming, I've had most of the parts for quite some time.

The plan is to combine 12 Opnext HL63133DG 170mW 637nm single-mode diodes leaving room to expand to 16.

I'll be using:

• Dave's diode mounts
• Dave's flex mounts
• 10mm x 10mm x 2mm first surface mirror from this GB
• Optima 305-0464-780 4mm FL lenses
• Dave's Optima lens holders
• 1/2 Lambda wave plate from Dave
• Newport MT-RS rotary mount for wave plate
• Suitable PBS cube
• Opto Sigma 155-1440 Mini-Prism Table
• MM1 style kinematic mount
• 16 units of 15mm x 15mm x 4.34mm TECs from Wimo

Still to decide:

• Telescopics

Concept so far...

The following 2 images show 4 knife-edged beams stacked on top of another 4 knife-edged beams...

I will have 2 of these assemblies and then combine with the wave plate, 90 degree turn mirror and PBS cube.

I expect this will fatten up vertically with a little defocussing...

I included the wireframe in the above image to illustrate how the 0 shape beam becomes a flat oval (at 10 metres) and overlaps the neighbouring beams. This is the whole slow axis / fast axis business at work - hopefully you find this as interesting as I

Next steps:

• Design combiner base plate
• Incorporate cooling (recesses for TECs)
• Confirm TEC driver(s)
• Confirm telescopics
• Other stuff not yet considered...

I welcome your thoughts and advice as this is my first build of this complexity. I need to make sure I have factored in everything, tolerances, adjustability, stability, thermal performance etc.

[-bart-]

Why such a small tec, do you plan to cool all diodes individually ?

[taggalucci]

Why such a small tec, do you plan to cool all diodes individually ?

Well, I was inspired by Solarfire's build and thought he might be onto something with individual cooling of each diode. So yes, though I may well end up driving them from 1 driver using (for example) 12 thermistors in series. I've already bought the TECs so I'd like to try a design that incorporates them. If anyone knows of a reason why I shouldn't please shout loud now!

[MIIKKKLLLL]

Hi taggalucci,

Your project sounds interesting,

I did the same 1 year and a half ago, a 16x module with 637nm single mode diodes.

I stacked the beams like you, for each polarization : 2 row by 4 columns

But for stability purpose, I installed 2 big TEC under the whole aluminium assembly (also easier to install).

Here are some pics, that might help you :

Picture of the inside, you'll notice that the wave plate and one diode were removed
Attachment 36902

Picture of the module. So I believe that you'll ask me : "the inertia of the module looks big". Yes, and this is what I was looking for .
Attachment 36901


All in all, this is the best configuration available for 16 diodes. 3*3 configuration (doing 18 diodes with H+V polarization) is too big to fit 5mm galvo...

You will see that there isn't telescope, as it isn't necessary.

[taggalucci]

Hi taggalucci,

Hi

But for stability purpose, I installed 2 big TEC under the whole aluminium assembly (also easier to install).

Interesting. I wonder whether having 2 big squares to rest on versus 16 little ones really makes any stability differences

All in all, this is the best configuration available for 16 diodes. 3*3 configuration (doing 18 diodes with H+V polarization) is too big to fit 5mm galvo...

I have some large mirror galvos for this projector, so that's not too much of an issue.

You will see that there isn't telescope, as it isn't necessary.

Cool, good to hear The only reason I was thinking of doing this was to match the profile of my 2.1W CNI green. I don't have to do this right away, I could leave space in case I need to add it afterwards.

Why did you choose to mount your mirror mounts at 45 degrees? I'm trying to understand any advantages / disadvantages of each approach.

[taggalucci]

Here's another pic with the latest version of my design...



I'm thinking that a combination of lateral movement in the diode mounts combined with longitudinal movement in the mirror mounts should provide me with the required adjustability. But as I've not done this before, I welcome anyone else experiences that would help me get this right

[dsli_jon]

hey Mon frére..

I welcome your thoughts..

Only $.02 I'd toss-in, would be to at least 'consider' going with right-angle red-coated prisms vs 'mir', for the simple-reason that the 'near-zero back-clipping' you get from the razor edge of the knifing-side, will give you that much more 'efficiency', and, make it easier to adjust for any 'blooming' you may experience with current-gains.. Granted, there's less of that to 'deal with' with these Reds, than the 445s, etc, but the point is, to whatever-degree you need 'micro-translation' to help glean every-last precious µW prisms are about as good as it gets... here's a 'point of reference' / some ideas for inspiration / food for thought, etc..

http://i187.photobucket.com/albums/x...1/DSC04994.jpg ...note the 'staggered' positioning of the prisms on the mounts...

..Not sure if Herr Alex http://www.photonlexicon.com/forums/...potenuse/page2 has reds any-left 'in stock', but you might give him a ping, if-interested...

Maybe you 'already know this' / thought of it / analyzed pros/cons, for your particular-build, and had decided to go the route you're plotting for a darn-good reason, in which-case, I am more than happy to go piss up a rope.. But, if-not, 'dos centavitos', fwiw...

cheers, and will look forward to yer pix..
j

[taggalucci]

Thanks Jon

I would have gone for the prisms but I've already spent on the mirrors some time ago. I need to use what I've got before forking out more cash I don't have! I do think I'd get a better result with the prisms - I'm just going to have to suck it up and move on

[MIIKKKLLLL]

Hi
Interesting. I wonder whether having 2 big squares to rest on versus 16 little ones really makes any stability differences

Because you reduce the probability of misalignment with that method, if you use a "one" piece aluminium plate for each "8 diodes" side. Other point, you reduce the number of cables, the overall cost of peltier and maybe the cost of peltier drivers.
If I had to reproduce my module again, I wouldn't go for 12mm (1/2 inch) small aixiz module to hold the diodes, as the thermal conductivity is lower than a direct fixation in the "aluminium plate". All in all, this is not so important as one diode generate less than 0.7W of heat (compared with blue who generates +- 7-8W, depending of current setting).

I have some large mirror galvos for this projector, so that's not too much of an issue.

Ok, without telescope and with 4mm FL lenses, you can achieve 6*4.5mm beam.

Cool, good to hear The only reason I was thinking of doing this was to match the profile of my 2.1W CNI green. I don't have to do this right away, I could leave space in case I need to add it afterwards.

If you use telescope, you need to know that the misalignment will be higher than without. (angle of misalignment is directly proportional to the zoom ratio).

Why did you choose to mount your mirror mounts at 45 degrees? I'm trying to understand any advantages / disadvantages of each approach.

Simply because I've made my own mounts and I spend enough time to do them so the edges aren't cut .

As Jon said, the best solution is still the use of red-coated reflect prisms. The alternative is a mount with the edge cut at 45° (same story with the "pan" adjustement).

[andy_con]

tec'ing each brass diode housing and dumping the heat into the module base plate which holds the mirror mounts is a bad idea.

you want the entire module to run cool not just the diodes, it makes for better stability

been there done it...

http://www.laser-man.co.uk/projects/...ode-array.html

arctos, kvant and rgb all tec the entire base plate