Big Red Banger using 12 to 16 Opnext HL63133DG Diodes

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...
Attachment 36898 Attachment 36899
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...
Attachment 36900
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.

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

Quote:

---

Originally Posted by -bart-

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!

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 :p.
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.

Quote:

---

Originally Posted by MIIKKKLLLL

Hi taggalucci,

---

Hi :)

Quote:

---

Originally Posted by MIIKKKLLLL

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 :confused:

Quote:

---

Originally Posted by MIIKKKLLLL

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.

Quote:

---

Originally Posted by MIIKKKLLLL

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.

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

Attachment 36903

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 :)

hey Mon frére.. :)

Quote:

---

Originally Posted by taggalucci

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... :cool:

..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.. :p But, if-not, 'dos centavitos', fwiw... :cool:

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

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 ;)

Quote:

---

Originally Posted by taggalucci

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

---

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).

Quote:

---

Originally Posted by taggalucci

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.

Quote:

---

Originally Posted by taggalucci

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).

Quote:

---

Originally Posted by taggalucci

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 :p.

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).

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

yep what andy says , or you will alighn everything and after an hour or so and the tecs have warmed up the baseplate and mirror mounts,
it will all be wrong again

Something I learned with time and projects :

Less mirrors you have ( = less mounts you have) lower will be the power losses, and also the misalignment as there is less flexing points.

Exemple ; (sorry guys if it's one of your project... I don't remember where I found the picture).

Here, 8 (!) mirrors should not be there :
Attachment 36910

Quote:

---

Originally Posted by MIIKKKLLLL

Something I learned with time and projects :

Less mirrors you have ( = less mounts you have) lower will be the power losses, and also the misalignment as there is less flexing points.

Exemple ; (sorry guys if it's one of your project... I don't remember where I found the picture).

Here, 8 (!) mirrors should not be there :
Attachment 36910

---

hi. that's a very early pic of my projector build.

the "extra" mirrors are there quite intentionally. the beams are double bounced before hitting the dichros to ensure the vertical alignment is spot on. the alignment is actually quite stable.

Attachment 36911

http://youtu.be/nLhs5k40ulk

Ok, so the mounts are there for beam height adjustment purpose. Sorry for my last post, I didn't want to criticize your work but just help Taggalucci with some advices ...

BTW, nice abstract show :D. Good frames optimization, absolutely no flickering. And nice alignment too !

oh no offense taken ;)

it's absolutely a valid point that each additional mirror reduces total overall power and increases the likely hood for alignment problems.

Quote:

---

Originally Posted by MIIKKKLLLL

Ok, so the mounts are there for beam height adjustment purpose. Sorry for my last post, I didn't want to criticize your work but just help Taggalucci with some advices ...

BTW, nice abstract show :D. Good frames optimization, absolutely no flickering. And nice alignment too !

---

if your worried about the amount of mirrors used then muti diode builds are not for you.

Quote:

---

Originally Posted by MIIKKKLLLL

Something I learned with time and projects :

Less mirrors you have ( = less mounts you have) lower will be the power losses, and also the misalignment as there is less flexing points.

Exemple ; (sorry guys if it's one of your project... I don't remember where I found the picture).

Here, 8 (!) mirrors should not be there :
Attachment 36910

---

Hi Andy,

Quote:

---

Originally Posted by andy_con

been there done it...

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

---

Yes, I've seen your builds and regularly take inspiration from your site. I'm now finally doing something about that with this build!

Quote:

---

Originally Posted by 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

arctos, kvant and rgb all tec the entire base plate

---

I plan to TEC the entire module base plate. I just plan to do it with 16 TECs placed underneath the diode blocks, BUT importantly between the module baseplate and the heat sink. If you look at my latest drawing, you'll see that the diode mounts rest directly on the module base plate. That's why I question stability concerns being raised when there are 16 dispersed TECs rather than 2 bigger ones e.g. Like a building built on 2 big piles rather than an even and spread out set of smaller piles.

If someone can explain scientifically why this would be less stable, please chime in :)

On the mirror count, I've deliberately reduced mirrors by having all 8 in "series" rather than doing 2 sets of 4 in parallel and then stacking. So I get the whole more mirrors = more alignment issues. I'm not sure where that came from!

I appreciate all your comments and views. I'll be working through the design tonight with my industrial designer friend and will discuss your points with him for his input.

So your saying the tecs are going to be under the main base plate that the diodes and mirror mounts are on ?

And if that is so with 16 tecs better make sure that baseplate and heatsink is dead flat
As i have found in the past just using two big ones gave be problems getting good contact across both
Let alone 16 small ones

Quote:

---

Originally Posted by taggalucci

Hi Andy,



Yes, I've seen your builds and regularly take inspiration from your site. I'm now finally doing something about that with this build!



I plan to TEC the entire module base plate. I just plan to do it with 16 TECs placed underneath the diode blocks, BUT importantly between the module baseplate and the heat sink. If you look at my latest drawing, you'll see that the diode mounts rest directly on the module base plate. That's why I question stability concerns being raised when there are 16 dispersed TECs rather than 2 bigger ones e.g. Like a building built on 2 big piles rather than an even and spread out set of smaller piles.

If someone can explain scientifically why this would be less stable, please chime in :)

On the mirror count, I've deliberately reduced mirrors by having all 8 in "series" rather than doing 2 sets of 4 in parallel and then stacking. So I get the whole more mirrors = more alignment issues. I'm not sure where that came from!

I appreciate all your comments and views. I'll be working through the design tonight with my industrial designer friend and will discuss your points with him for his input.

---

Quote:

---

Originally Posted by badger1666

So your saying the tecs are going to be under the main base plate that the diodes and mirror mounts are on ?

---

That's exactly my plan right now (subject to further scientific scrutiny)!

Quote:

---

Originally Posted by badger1666

And if that is so with 16 tecs better make sure that baseplate and heatsink is dead flat
As i have found in the past just using two big ones gave be problems getting good contact across both
Let alone 16 small ones

---

Ultimately, this will be about the quality of machining and thermal interfacing. I want to achieve a good result in both.

Ok then let me put it this way
Get yourself some digital calipers
And measure the thickness of all you tecs ;)
And letme know the results
Its going to be a mirical machining job
If you can get good thermal contact across
All 16 tecs

I have always used bigger TECs primarily to minimize the wiring. I have also sourced the TECs from one company (lazy) and at the same time and so these are quite uniform in thickness, specified and lapped flat.
I agree that before setting this base plate up, taggalucci should carefully measure them. I suspect that if they measure out then his bigger problem will be the plate. Unless your mill's spindle is dead on square then any surfacing tool ( and you must surface raw stock) will leave a very shallow scalloping at the center of the cut. Final flattening with a surface grinder is easy and better if available.

A interesting alternative would be to use the 0.5mm thick thermally conductive pads popular in the electronics/computer area. 3M makes these as well as others and they are inexpensive and available through over-clocker supply houses. This will also remove some concern in engineering the distribution of hold down screws to avoid bending of a flat plate as it is secured. A fair amount of pressure is required if using grease or indium and the all hard and flat conventional TEC mounting.

Nice one Tony!

I feel a chubby coming on!

Is that too much information?

Grafoil. Comes from a local company, used for making high temp gaskets. Comes in nice flat sheets, you can cut it with scissors. Excellent thermal and electrical conductivity. Doesnt flake to short out the TECs either. Cost effective.

Similar materials made by the same company are used for CPU coolers etc in laptops and servers.

Steve

Quote:

---

Originally Posted by badger1666

Ok then let me put it this way
Get yourself some digital calipers
And measure the thickness of all you tecs ;)
And letme know the results
Its going to be a mirical machining job
If you can get good thermal contact across
All 16 tecs

---

Quote:

---

Originally Posted by planters

I have always used bigger TECs primarily to minimize the wiring. I have also sourced the TECs from one company (lazy) and at the same time and so these are quite uniform in thickness, specified and lapped flat.
I agree that before setting this base plate up, taggalucci should carefully measure them. I suspect that if they measure out then his bigger problem will be the plate. Unless your mill's spindle is dead on square then any surfacing tool ( and you must surface raw stock) will leave a very shallow scalloping at the center of the cut. Final flattening with a surface grinder is easy and better if available.

A interesting alternative would be to use the 0.5mm thick thermally conductive pads popular in the electronics/computer area. 3M makes these as well as others and they are inexpensive and available through over-clocker supply houses. This will also remove some concern in engineering the distribution of hold down screws to avoid bending of a flat plate as it is secured. A fair amount of pressure is required if using grease or indium and the all hard and flat conventional TEC mounting.

---

4 x 3.30 mm
8 x 3.36 mm
1 x 3.38 mm
3 x 3.40 mm

So the largest difference is 0.1 mm

I don't plan to machine for the difference in thicknesses, I plan to use thermally conductive foam under pressure to deal with this. I have a sample from my industrial designer friend that provided promising results.

He recently worked on a project that required complex thermal management and they ended up using a similar approach - creating "islands of cold" in the heat sink located adjacent to the sources of heat.

Quote:

---

Originally Posted by dnar

Nice one Tony!

I feel a chubby coming on!

Is that too much information?

---

Yes, way too much info Wayne!

Quote:

---

Originally Posted by mixedgas

Grafoil. Comes from a local company, used for making high temp gaskets. Comes in nice flat sheets, you can cut it with scissors. Excellent thermal and electrical conductivity. Doesnt flake to short out the TECs either. Cost effective.

Similar materials made by the same company are used for CPU coolers etc in laptops and servers.

Steve

---

Grafoil sounds interesting, but I'm not sure which of their products you're suggesting would be suitable Steve.

I'm just finalising the baseplate design and will post some more when I'm done.

I wonder if you could use 4 of the 71 red diodes to replace all this. Considering the number of diodes the beam size can't be much different and special optics to tame the didoes would be smaller and cheaper than all this. Now if you did 12 of the 71 diodes that would be something to see! 8w or so of 638nm would be incredible.

Quote:

---

Originally Posted by kecked

I wonder if you could use 4 of the 71 red diodes to replace all this. Considering the number of diodes the beam size can't be much different and special optics to tame the didoes would be smaller and cheaper than all this. Now if you did 12 of the 71 diodes that would be something to see! 8w or so of 638nm would be incredible.

---

This is all discussed in http://www.photonlexicon.com/forums/...lashlight-Fail

This thread is about this build with the parts list shown as I pretty much have all of the parts (and have had for many months)!

So until I finish building what I already have, I'm not (personally) going to try using the 300mW or 500mW wide stripe multimode diodes (though I do have one 71 diode and some Kiyoukan optics).

The beam properties of these Opnext diodes are amazing and I'm not convinced you could achieve the same power density and divergence with multimode diodes. But this has already been discussed before.

I agree with kecked. The problem with combining a large number of diodes is that all the problems of alignment and reliability are proportionally greater. And the larger spread of all the components is associated with more instability problems. If the absolute maximum power (forget brightness and beam size) is the goal then you need the maximum number of the most powerful diodes you can get. But, if you have a target with your current project of 2-2.5W then 4 G71's run at 1A each will generate in excess of 3W. Two PBSed vertically stacked stripes expanded laterally produces a very nice beam. You'll need a 2-3 power telescope to get a DPPS to match the divergence and you're cooling plan fits in nicely with this much more compact set up. I have worked with as many as 20 diodes in a module and the alignment is pretty tough. And remember if cooling; when one pesky diode gets a little out of alignment or focus (and that's another issue) then you have to warm it up, remove the enclosure, adjust and then replace the cover.

Guys, this is all great stuff honestly. But it's not helping this build!

Quote:

---

Originally Posted by taggalucci

Guys, this is all great stuff honestly. But it's not helping this build!

---

what help do you need?

Quote:

---

Originally Posted by andy_con

what help do you need?

---

Nothing right now unless someone has any advice with regards to "how to" with the parts I'm using or the design of the baseplate (which is really the only variable in this build).

My comment about help above was in regards to posts that described a completely different set of parts i.e. it's not helping me (my sanity or my pocket) basically suggesting I should ditch this set of very expensive parts and buy some more! I appreciate this was probably not the intent :)

I also appreciate that the posts may be of use to someone looking at this thread and thinking whether to buy similar parts to mine or go the alternative. However this is a build thread not a general "what's the best/cheapest/least-diode way to get 2+ watts of red" thread! Given many have lamented the lack of build threads, please lets try and stay on topic to help a builder out?

The comments regarding my choice of TECs and how these will be integrated into the plate were great and have challenged me to further consider the factors raised. They were discussed at length with a highly experienced industrial designer and I found his feedback on this enlightening.

I respect that Andy, Planters and others have already built multi diode set-ups, so I very much respect and appreciate your input. Please just don't tell me to chuck my parts in the bin and start over especially when these diodes can and have been combined successfully in the past! I.e. constructive contributions from the masters most welcome :)

Fair enough. And no, DO NOT CHUCK PARTS. As your build will be unique it will be very worthwhile and although this is not a divergence competition, the performance will be interesting to follow. My suggestions were not intended to divert you. I suppose I miss interpreted your OP to be a review of the potential components of the design, I didn't release you had already acquired many of the components and were a lot further along.

OK. Regarding the base plate, your CAD image shows the diode mounts and the flex mounts recessed into milled depressions in the plate. It looks cool, but I believe that this may actually make the task of adjusting the mounts a little more difficult as well as significantly reducing the bending modulus of the plate. As you tighten the plate down against the TECs it will flex. The less it flexes the better both optically and as it effects TEC contact, thermally.

You haven't described your intentions for the cooling. Do you plan to cool aggressively below ambient? If so, do you have any experiments showing the effect on these diodes and how will you house the assembly to prevent condensation? Have you some idea of the TEC power requirements and will there be some consideration for the heat removal from the main projector base plate? I can make some suggestions.

Hey Tony, I'm in the same boat as you - diodes that were purchased a while ago, and a project which has been a bit slow to complete...

Here is one of my modules to give you some ideas ;)

Getting ~1.1w and a beam that fits on 5mm aperture mirrors nicely @ ~1mrad, much nicer far field spot than one of the wide stripe reds.. Big TEC sitting under the base plate driven with a PI( proportional integral ) driver ( lots of headroom - the TEC hardly gets driven to maintain the set temp )

Attachment 37101

Quote:

---

Originally Posted by planters

I didn't release you had already acquired many of the components and were a lot further along.

---

Yes, I didn't make that clear. Here's what I have...

Attachment 37239

Quote:

---

Originally Posted by planters

Regarding the base plate, your CAD image shows the diode mounts and the flex mounts recessed into milled depressions in the plate. It looks cool, but I believe that this may actually make the task of adjusting the mounts a little more difficult as well as significantly reducing the bending modulus of the plate. As you tighten the plate down against the TECs it will flex. The less it flexes the better both optically and as it effects TEC contact, thermally.

---

Point taken, also it simplifies the milling significantly. Don't quite understand your point about the bending modulus. I saw these not as recesses, but rather as protrusions, so I'm not sure how these protrusions would compromise the rigidity of the remaining plate. Anyway regardless of that, this is how it's now looking...

Attachment 37240 Attachment 37241

(Big thanks to Kitatit for knocking up the Opto Sigma mini prism table in SolidWorks for me so quickly) :)

With the design shown above, I have a big 12mm baseplate that's sinking the heat from the TECs. Sandwiched between the TECs and the baseplate are thermal conductive pads to even out any variance in the height of the 16 TECs. Above the TECs, is the sub-baseplate which has the two "steps" to facilitate the 3mm stacking of the back 8 lasers with the front 8 lasers. At the shallower end, the TECs have 5mm of sub-baseplate between them and the diode mounts (and therefore 8mm at the deeper end). This allows the overall height of the beams to match my big 2.1W CNI green (@ ~45mm).

Quote:

---

Originally Posted by planters

Do you plan to cool aggressively below ambient?

---

I'm just looking to maintain a comfortable operating temperature, especially as these are beautiful 637nm diodes :)

Quote:

---

Originally Posted by planters

Have you some idea of the TEC power requirements and will there be some consideration for the heat removal from the main projector base plate? I can make some suggestions.

---

The spec of each 15mm x 15mm TEC is as follows:

Max. thermal transfer power: Qc max = 3.9W at Th = 27°C, I = 3.3A, ∆T = 0°C
Max. temperature differential: ∆T max>=66°C at Th = 27°C, I = 3.3A, Qc = 0W
Power Supply: I = 3.3A
Power Supply: V = 2.0V DC

I'm looking for a suitable TEC driver(s) to run all 16. Solarfire suggested when asked about 12, that these could be run in 2 parallel x 6 series from his driver. Still waiting to hear back from him on whether his would run 16. He also suggested using 10 x 10k thermistors in series to get an average across the array.

Quote:

---

Originally Posted by dave

Hey Tony, I'm in the same boat as you - diodes that were purchased a while ago, and a project which has been a bit slow to complete...

---

I'm struggling to find time, so I'm with you! Pulled a late night ('til 2AM) to get the CAD done so that it can go off to my designer mate for his magic touch before the file is sent to China.

Quote:

---

Originally Posted by dave

Here is one of my modules to give you some ideas ;)

Getting ~1.1w and a beam that fits on 5mm aperture mirrors nicely @ ~1mrad, much nicer far field spot than one of the wide stripe reds.. Big TEC sitting under the base plate driven with a PI( proportional integral ) driver ( lots of headroom - the TEC hardly gets driven to maintain the set temp )

---

Hopefully I'll get the specs I'm expecting and require minimal driving of the TECs.


Here's some other concept images...

Attachment 37247 Attachment 37250 Attachment 37248 Attachment 37249

M5 Nylon bolts to prevent thermal transfer between the hot and cold sides.

I'll clearly need to think about an aperture!

Quote:

---

M5 Nylon bolts to prevent thermal transfer between the hot and cold sides.

---

I recommend using m4 with an isolation washer.

You'll lose in little thermal isolation, but gain a lot in mechanical stability.

What happened to this thread

Quote:

---

Originally Posted by andyf97

What happened to this thread

---

Hi Andy,

This project has stalled due to work commitments :( Here's a view of where things are at...

Attachment 39037

I know the feeling, I am also loaded with too many things to do.

That design looks very professional..

Quote:

---

Originally Posted by taggalucci

Hi Andy,

This project has stalled due to work commitments :( Here's a view of where things are at...

Attachment 39037

---

Funny that this topic is resurfaced, because I'm re-building my 16x diodes this week :).

I'm currently measuring 2.65W at the output, after PBS and wave plate. Quite happy with the results.

Attachment 39039

Attachment 39038


I have an advice as you still don't drill the wholes to fix the flex mounts : Fix them as close as possible to the beam exit, like that you can decrease the beam height deviation.

I also see that the main aluminium plate is really thick (should be really stable). As the generated heat on these diodes are really low (<0.7W/diode), *maybe* you can go with passive cooling :confused:

Looks neat, what diodes are you using there.

Quote:

---

Originally Posted by MIIKKKLLLL

Funny that this topic is resurfaced, because I'm re-building my 16x diodes this week :).

I'm currently measuring 2.65W at the output, after PBS and wave plate. Quite happy with the results.

Attachment 39039

Attachment 39038


I have an advice as you still don't drill the wholes to fix the flex mounts : Fix them as close as possible to the beam exit, like that you can decrease the beam height deviation.

I also see that the main aluminium plate is really thick (should be really stable). As the generated heat on these diodes are really low (<0.7W/diode), *maybe* you can go with passive cooling :confused:

---

I use the diodes we offer for sale, single mode Opnext 638nm, P0=170mW. Diode's beam properties are really nice.

The module can output more than 3W, but the diodes runs here at a conservative current to extend lifetime :p.

Hi, I have a little problem with my mitsu... I cannot set it to "one" beam- on picture you can see my the best settings and spot after 7m.
Much easy to adjust is on anamorphic prism but beam is large ;/
I have fun with this from last four day, how to set up?

Attachment 39377

Attachment 39378

Attachment 39379

You have it set up all wrong , i suggest you read this thread ( all of it ) http://www.photonlexicon.com/forums/...ght=Holy+grail
You will be needing additional parts , pbs cube and wavplate ,another set of cylinder lens
Those reds need to be stacked in pairs ,and then pbs/waveplate them together
And all distances for reds are critical also i would not try building a quad without using 2mm fl lens
And daves adjustable mounts
I built a quad p73 and the holy grail thread showed me how to get it right ;)

Quote:

---

Originally Posted by bolo

Hi, I have a little problem with my mitsu... I cannot set it to "one" beam- on picture you can see my the best settings and spot after 7m.
Much easy to adjust is on anamorphic prism but beam is large ;/
I have fun with this from last four day, how to set up?

---

How did my thread about Opnext diodes turn into a "help me" thread on the Mitsubishi diodes???

Shot to admin- you have my agree for delate my post.

curious as to what lens you use , not using a telescope ?

Quote:

---

Originally Posted by 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 :p.
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.

---

I was hoping if we could spend some time on the Mitsu diodes? I really like them, a lot.

But seriously, Bolo raises a good question that is diode independent and that is near field and far field alignment. Forget the telescopes and prisms for now and visualize what is happening when two diodes are co-aligned.
The knife edges are not used to bring the beams toward convergence because if they do then the beams are farther apart than they could be near field and will inevitably diverge away from each other as they progress away from the projector. The angle adjustment for the knife is only to adjust the beams to be parallel to each other while the POSITION of the knife edge determines where the two parallel beams lie relative to each other. First, adjust the diodes in their housings to be parallel and if say 2cm apart in the projector then they should be 2cm apart @ 5M. Then, intercept the beams with the knives. Fix one knife and slide the other carefully on the base plate while watching the far-field spots. It is OK here (for now) to overlap these spots. You will quickly see that there is a limit to how close you can place the movable knife in toward the first beam before you begin to shadow the first beam. This limit is at its lowest when the second diode is as close to the edge of its knife as it can be without light failing to strike the mirror.

A big help here is to project the whole array on a large mirror that is 1/2 the distance of your typical, anticipated projection distance when the projector is complete. The reflection from this mirror can be adjusted to strike a screen close to your projector so that as you make these adjustments you can see both the "far field" spots and what you are adjusting in the same field of view.


If additional optics are used they will be added and work best if this basic alignment is done first

Quote:

---

Originally Posted by badger1666

curious as to what lens you use , not using a telescope ?

---

I use lenses with 3.2mm focal length ; longer FL and beams can't pass through scanners mirrors (for 4.5-5mm beam), shorter and you obtain a too high divergence & lense centering problems.

Where can i find 3.2mm fl lens ?
What divergence did you get ?
I see you also sell diodes please pm me price ;)
Thanks

Quote:

---

Originally Posted by MIIKKKLLLL

I use lenses with 3.2mm focal length ; longer FL and beams can't pass through scanners mirrors (for 4.5-5mm beam), shorter and you obtain a too high divergence & lense centering problems.

---