Best approach to multi-module heatsink design? (concerned about alignment)

[rhd]

Does anyone have any wisdom or advice on best practices for designing a single heatsink that is meant to hold multiple modules? I would like to be able to make the outputs roughly parallel. My thinking currently runs along two lines:

A) Integrate a active mechanism (set screws?) for fine adjustment of the modules within the heatsink.

B) Allow some wiggle room, and then patiently align each module using 2-part arctic silver that will harden and fix each module once aligned.

Neither approach is perfect. Option A is sounding better at present. Does anyone have any wisdom? I don't want to get too complicated, and this is still for a portable build, so I cant get terribly fancy. This still has to generally conform to normal heatsink dimensions (other than of course that it's going to be a large heatsink for a big host).

 

[Bionic-Badger]

I've given some thought to this as well, but I don't have the metalworking tools to carry it out. My conclusion is that you need to have at least 2-axis adjustments at each reflective/combining surface. There's really no way around it if you want something to aligned at relatively long distances.

If the cavity is relatively open you could attach part of the surface to a precision adjustment mount, set it right, and then epoxy it down. Then detach the precision mount and move onto the next one. My plan would have been to use UV-curing epoxy because it is extremely hard and sets extremely fast when you want it to. It's also used with glass work.

For the set-screw method, you'd have to set up a flexure-type mount inside the host to provide that spring action against the set-screws. You could also use magnets or a spring. The flexure method is a lot more rigid though.

 

[ARG]

I thought I would have this problem when doing a multi diode laser, but if it is milled correctly they should be parallel with low enough tolerances for the purposes of beam combining. I never had a problem with beams not being parallel for handheld multi diode lasers using a single setscrew.

However, I can tell you from trying to make a multi diode heat sink without a mill that option B is very finicky and will take a few attempts. Option A would be the way to go in my opinion.

 

[CDBEAM777]

RHD...I have had some success with making a " Combiner "...HH with multiple LD's. Fundamentally...it is a question of how does one achieve alignment and beam to beam proximity. The explanations below is posted for ANYONE curious about combining multiple LD's.

Let's address beam to beam proximity first....for this has impact on alignment. The only way to place one beam directly next to another beam is to use the knife edging practice. Please reference the " Tridentis " design for pics. As you can see....and as others have also done.... One LD is placed with its output axis parallel to the HH unit. Let's call this the Primary LD. This LD " Sets the Stage"....so to speak. If using two (2) or three (3) LD's in combination...I call the remaining LD modules... Secondary. These Secondary modules " join" the Primary LD output.

As a reference....remember the famous " Hello, hello, hello " signature opening greeting from the Three Stooges !!! Well....that is kinda what we are doing with combining multiple LD's ...Hello=Primary, hello=Secondary, hello=secondary !

With the Tridentis...the Primary beam does not bounce off any mirrors....but travels in a direct path....in this case...into the corrective Anamorphic optics. With knife edging...the Secondary beam is re-directed...at a 90 degree angle....and as you can see...the two secondary LD modules are placed at a 90 degree angle to the Primary LD module. The beam from each Secondary module hits the very edge of the mirror....takes a 90 degree turn...and then " joins" the Primary beam.

In this way....the output from all three beams are stacked next to each other....very closely....and appear as a single beam. This is really the only practical way to combine multiple modules. That said...let's move on to alignment.

For alignment...we have many considerations !!! I have found an End mill works fine to establish initial parallelism and beam output height...and the bounce knife edge mirrors do the rest. For example....with the " Tridentis"...an initial bore was made to accept the 12mm module OD for the Primary LD. This bore was parallel to the HH long axis. Then a additional block of Aluminum was milled....with two (2) bores...parallel....and at a height so that the module beam output was equal to that of the Primary LD beam. This Aluminum block was positioned such that the output from the two Secondary beams were at a perpendicular 90 degree to the Primary beam.

The real take away is that the knife edge mirrors and their mounts take care of the rest !!!! They align the pitch and yaw via the flexmount to which they are attached. The beams are aligned via the adjustabillity of the mount....NOT by physically adjusting the module !!!!

ALL three LD modules were secured within the bore hole by two(2) 6-32 set screws. No provisions to adjust the modules within the bore were done....the only consideration was to secure the modules within the bore hole and achieve the correct radial positioning.

The module beam height MUST be relatively close....and this is taken care of with attention to detail in the End Mill operation....but...the Flexmounts do the work !!! And...they are components one can buy off the shelf !!! Inexpensive too !!!

ALSO....the position of the Flexmount on the optical platform is important. THIS positioning consideration is why you see ALL the units in the " Combiner " class to have a semi-circular extension on the front section. This extension acts as a platform for the optics and allows the flexmount units to be positioned via the mount foot. The foot is bolted to the platform plane....and there is room for adjustment. This adjustment is needed to exactly position the mirror edge so that that a secondary beam is reflected...and the Primary beam passes right by. The adjustment screws on the mount itself then address the pitch and yaw alignment.

In this way...ALL beams are stacked....and their output is aligned to be parallel.....not perfectly parallel.....but practically speaking...perfect.( Unless you are on the moon...hahaha)

Note: Added Tridentis drawing

Lite'em up-CDBEAM=======>

 

[rhd]

This is all great wisdom above, and no doubt a good resource for others looking to do beam combining. Thanks for all the input!

However, I'm not looking to do beam combination

"I would like to be able to make the outputs roughly parallel."

I know it's less ambitious than beam combination, but I'm truly only trying to keep beams parallel - not trying to combine them. So it my case, it's really about module adjustment, because nothing else will be in the beam paths.

 

[IsaacT]

Subscribing. Sometime soon I want to do a Magenta build.

 

[CDBEAM777]

This is all great wisdom above, and no doubt a good resource for others looking to do beam combining. Thanks for all the input!

However, I'm not looking to do beam combination

"I would like to be able to make the outputs roughly parallel."

I know it's less ambitious than beam combination, but I'm truly only trying to keep beams parallel - not trying to combine them. So it my case, it's really about module adjustment, because nothing else will be in the beam paths.

MMMMMmmmm???? I have some thoughts on that...See the attached pic...

It MAY be possible to have a taper bore hole....like a LONG cone...12mm ID on the back end...13mm ID on the front end...and a series of 3 to 4 threaded 4-40 holes radially spaced around the 13mm end. These 4-40 holes would have allen head set screws in them. The set screws MIGHT be able to set the location of the module output end ...this being able to bring the beams in a parallel condition !!!??? Dunno...Have not tried it ?????

IF you were going for say three (3)modules...Access to all the set screws would be problematic ????? Might be doable...IF the location of the set screws were staggered/offset just slightly... along the module long axis ...but still down toward the 13mm end ???

Just a thought !!!!! Maybe spring load one side....and have two (2)set screws
engage the remaining sides....so you would only need to have access to two (2) set screws per module ????

In the words of Doctor Frank-n-stein..." It Could Work " !!!!

Also....see added pic for some " Combiner " humor....hahaha Just being a Wise Ass !!!

Lite'em up-CDBEAM=======>

 

[Bionic-Badger]

How much space do you have to work with? I had some ideas for for a stage-laser type module setup, which might be able to work. You'd use 12mm flexible shaft couplers to provide a module holder with a high-tension "spring" for steering purposes using set screws. It'd also provide moderate heatsinking. They may be too large for your purposes, but possibly if you stacked them up slightly you could fit them into whatever portable setup you're attempting. You could also make your own like that in two axises.

 

[CDBEAM777]

I forgot about heat sinking !!! MMmmmmm ???????? Sheeeeeezzzzz....It's ALWAYS something !!! MMMMmmmm... ????

Well...maybe just bore three(3) 12mm ID holes in an Aluminum block....Place your modules in the block, secure with set screws. Send the beams into three Flexmount units with three Prism mirrors attached to the Flexmount.

The beams hit the mirrors...and take a 90 degree bounce. Then... use the Flexmounts with attached mirrors to do your alignment/beam steering and place your beams in a parallel alignment !!!

Simple !! Small ! Inexpensive ! Easy to align....and the modules would be well heatsinked in the Aluminum Block !!!

Parallel beams...TO GO !!! :drool: :beer::beer::beer:

We have our cake....and eat it too !!!

Lite'em up CDBEAM=======>

 

[rhd]

I don't have nearly enough space for any of the above sadly. In fact, I'm essentially limited to the space available in a conventional heatsink. So any solution I adopt need's to work within the volume boundaries of the heatsink itself. I can get creative, because it's going to be a CNC job, but I can't get as brilliantly creative as some of the suggestions in this thread (although they're great suggestions!)

 

[CDBEAM777]

I don't have nearly enough space for any of the above sadly. In fact, I'm essentially limited to the space available in a conventional heatsink. So any solution I adopt need's to work within the volume boundaries of the heatsink itself. I can get creative, because it's going to be a CNC job, but I can't get as brilliantly creative as some of the suggestions in this thread (although they're great suggestions!)

:thinking::thinking::thinking::thinking:

Well.....No...Ya can't get 10 pounds a shit inta a 5 pound bag !!!! There are always trade-offs !!! Engineering is always a series of compromises !!!

Exactly what final dimensions are you looking for ?????? What are the dimensions of a " conventional heatsink " ???

I guess the take away is I need to know the end game, the final target dimensions of a completed unit. These constraints absolutely drive the possible optical solutions !

Finally....what you want MAY not be achievable.....within the final end dimensions you require ??? You may have to expand what is acceptable !!
....." Go Bigger....or Go Home" hahahahahaha...sorry !! Dunno ?? But....it is a fun challenge !!!

Lite'em up CDBEAM=======>

 

[ARG]

Well.....No...Ya can't get 10 pounds a shit inta a 5 pound bag !!!! There are always trade-offs !!! Engineering is always a series of compromises !!!

But you can get 7.5 pounds of shit into a 5 pound bag

 

[Jmillerdoc]

I am having Flamingpyro build two hosts for this purpose as we speak. My solution to getting the beams aligned was to use the tiny amount of play I have with some lens barrels I bought a while back. I don't know where I got these but they are different from the standard three elements like DTR sells. They are not threaded all the way down the barrel and are smooth on the output end of the barrel. They are somewhat looser than my regular barrels and have enough play to allow some directional control over the beam. I plan to use some epoxy on the threads, get them all focused to infinity, and hold them all with the beams parallel while the epoxy sets. It will not allow any focus once the epoxy is set but I don't need that. I haven't measured the amount of angle I can swing the beam from one extreme to the other yet but if I had to estimate I would say 2-3 degrees. There is probably enough play in a regular lens barrel but I haven't really paid that close attention to my regular lenses.

Maybe this will offer another possible solution. Don't really know what your endpoint is. I haven't spoken with Jeff in several days but I bet he's close to finishing my hosts. I will be glad to post my results if you're interested.

JM

 

[CDBEAM777]

But you can get 7.5 pounds of shit into a 5 pound bag

Yes......But.....you SQUEEEEEEEZZZZZZ out.......er......ah.....liquid.......The compression process could also be applied to a HH.....grind it up !!! Compress it !!!.....and it'll fit into a smaller bag !!!! hahahahaha !!!!:drool:

Lite'em Up CDBEAM=======>

 

[rhd]

For what it's worth, mythbusters recently fit 10 pounds of crap into a 5lb bag

Jmiller:
I could be wrong, but I don't think adjusting the lens has impact on the direction of the beam. Have you tested?

 

[Bionic-Badger]

What if you create a single heatsink mass with three holders, press the diodes/lenses into those holders, and then separate them by slots? Then use set screws to pull or push the modules towards or away from each other. It'll only give you very little adjustment, so you'll have to get the beams relatively parallel to begin with, but for longer-distances that little bit of set-screw adjustment might be enough to maintain the beam parallelism. It'd essentially be like those flex-couplers, but with three of them combined into a smaller space (and smaller/fewer flex points).

Other than that, you'll need to use some reflective surfaces and adjustable mirrors to direct the beams out some holes. It'll be a real pain to do.