Seeking links to good knife edging techniques material.

[CDBEAM777]

Probably Contact, the signal puzzle explained by the esoteric engineer and CEO, I've forgotten what his name was in the movie, but that is my all time favorite. I'm trying to puzzle this diagram out, but I cannot understand it yet.

You need a Primer !!! Hahaha

BTW...S.R Haden was the character in the movie " Contact "...from which I quoted....FTWGARA !!....and another puzzle....what the hell does
" FTWGARA " stand for ????? :wave:

Now...I will do a plan view render ...and perhaps a side view render...mainly to make sure....my head has not " vortexed " down a worm hole !!!

Stand by !

Beam out

 

[CDBEAM777]

OMG...My head hurts !!! See u now the attached render....End view and Plan view. The side view will be another day !!!!!

I remain confident this set-up will work !

What a nightmare to draw....likely even worse for you to figure out !!! hahahahaha:eg::eg:

Seeing the design drawn increases confidence that it might work !
I still have not spotted a design flaw....not yet.

But...then...this is just V1....along about V12....things become apparent !!

OH...overall width has increased from 55mm to 61mm...oh well...the real proof will come when I try to build it !!!:yabbem:

Beam out

 

[CDBEAM777]

Much appreciate! That looks like it is so narrow, I can offset it to one side to have the beam go out the center of my 2.7 inch wide plate inside my host. That way, I could expand the beam further.

Well ...yes...we folded it in half...and added height....so....made it almost a square !!

See the new drawing with some additional labels !

OK...... First...we have two banks...the Left and the Right. Each bank has two (2) LD units...one unit up and one unit down. The beams for each up and down unit start from the LD at the N side...and travels thru the G2 collimation lens...then traveling toward the S side...passing thru the Cylindrical optics for correction.

Once corrected, the beams then continue on to hit the knife edge mirrors. The knife edge mirrors stack/position the two (2) beams...side by side ...and at the same time...turn the two (2) beams 90 deg. At this point...both beams are head downward toward the base of the T-Bracket. This sequence is the exact same for each bank.

Now...let's take each individual bank.

The Left bank stacked beams are now headed down toward the bounce mirror #1. This mirror then steers both beams at a 90 deg turn...headed to the right or E side. The beams pass thru a small wave plate and then thru a hole in the T-Bracket...and toward the PBS Cube.

The Right bank stacked beams are now headed down toward the bounce mirror #2. This mirror then steers both beams at a 90 deg turn...headed forward...toward the S side... and toward the PBS cube.

SO...now we have both sets of beams...headed toward the PBS cube.

The Right bank set... entering the PBS cube from the back....and passing thru...and now headed toward the S side.

The Left bank set...entering the PBS from the side...taking a 90 deg bounce...and combining with the beams from the Right bank....and now also headed to the S side.

All four (4) beams are now headed toward the S side...leaving the PBS cube....All combined !!

Aren't optics wonderful !!! Yikes...I need several beers !!!
...and after you have read the above...blah, blah....so will you !!!

Note: While watching Michigan State get stomped...cleaned up the drawing...back to 55mm
wide...see V4

Beam out

 

[CDBEAM777]

Well...we are now up to version V7 ! FTWGARA

I have made no modifications to the beam path's.

Just moved the Bounce mirror mounts and PBS mount ... so that they are now attached to the T-Bracket base. Also relocated the Wave plate to below the right bank.

After moving things around...in the virtual CAD world...I am fairly certain this design will work.

Beam out

 

[jors]

CDBEAM I always enjoy your posts, thanks for that:wave:
IMHO I think easier, cheaper, and smaller:

Knife edge 2 and stack over another 2 + PBS + another set
+prism correction + b.e. (8 diodes)

yes, we have gaps increased due magnificaction but beams overlap soon thus reducing-eliminating gaps due beam divergence.

This is how several multimode octo prof modules are done

 

[CDBEAM777]

CDBEAM I always enjoy your posts, thanks for that:wave:
IMHO I think easier, cheaper, and smaller:

Knife edge 2 and stack over another 2 + PBS + another set
+prism correction + b.e. (8 diodes)

yes, we have gaps increased due magnificaction but beams overlap soon thus reducing-eliminating gaps due beam divergence.

This is how several multimode octo prof modules are done

I would really like to see a pic of this....sounds very interesting !!

I will attempt a Quad....sometime in the future....But...Eight LD's....

WAY over my budget !! But...given some space....sure...it can and has been done !!!

Now...I do not think that Anamorphic prism correction will be sufficient....and If this is so...we are stuck with Cylindrical Lens correction....which has more optical conditions !! I have a set of A-prism's....and will experiment.

I have experimented...and used Anamorphic Prism's...While they correct a beam....when the Aspect ratio is bad....really bad....as with many MM diodes....then....the Anamorphic's will still work....but must be adjusted at such a severe angle of incidence...drastic power loss is a result....like 15% or more loss !!...which is not acceptable for me !

Beam Out:evil:

 

[CDBEAM777]

I could be happy with 4 diodes, but 8, wow! The Fenix TK75 flashlight is perfect and can have plenty of battery capacity the way they series parallel additional battery packs in the extension tubes.

Well....let's give this compact quad design a final view....last drawing

See attached.

Beam out:thinking::thinking::thinking:

 

[RedCowboy]

I could be happy with 4 diodes, but 8, wow! The Fenix TK75 flashlight is perfect and can have plenty of battery capacity the way they series parallel additional battery packs in the extension tubes.

4, 9, 16
Those will come out better if you square up your MM beams before KE
i.e. Its a 2 level so 2x2, If 9 then 3 level 3x3, 16 means a 4 levels at 4 x 4

You could do 8 on 2 levels of 4 each corrected and KE but you will have a spot that is a rectangle, depending on how square the cylindrical lenses get the beam and how the beams diverge, but if you aim for perfectly square to start and see how you do, then build the monster 16 diode unit later.

Just my 2 cents, I am enjoying watching this build, good luck and good work to both of you. :gj:

 

[jors]

I would really like to see a pic of this....sounds very interesting !!

I will attempt a Quad....sometime in the future....But...Eight LD's....

WAY over my budget !! But...given some space....sure...it can and has been done !!!

Now...I do not think that Anamorphic prism correction will be sufficient....and If this is so...we are stuck with Cylindrical Lens correction....which has more optical conditions !! I have a set of A-prism's....and will experiment.

I have experimented...and used Anamorphic Prism's...While they correct a beam....when the Aspect ratio is bad....really bad....as with many MM diodes....then....the Anamorphic's will still work....but must be adjusted at such a severe angle of incidence...drastic power loss is a result....like 15% or more loss !!...which is not acceptable for me !

Beam Out:evil:

Well, severe angle or not, depends on magnif. you want to achieve.... you probably have seen modules like this.... This is a pic of 12 diodes MM module on PL (10w green)
Photonlexicon

 

[jors]

really nice those lasertack machining, what power, divergence and beam diameter you want to achieve?

 

[jors]

WOW...Look at that !!! It can be done !!! Great !!!:bowdown::bowdown::bowdown:

Looks like they are producing a pattern like....

ll
ll

And the BE is doing it's thing !! Well...in thinking about my Linos experiments...it was off-axis propagation that would not work ....so...if the pitch or yaw were off...that was a problem !!!

Clearly...there are four (4) beams...entering the 4X BE...and one would assume that their aggregate / combined centre line is entering the BE dead centre !!!

Beautiful work !!! IIRC....Tuned Cavity ( who built this) is outa Europe.

BUT...with Cylindrical correction ...or Anamorphic prism's placed in the beam path...then knife edged ...then beam expansion....one would get both beam expansion....and... less divergence ....BUT.... a larger cost/size and complicated build !! That is for sure !!

Perhaps I will experiment with the Tridentis build......remember...that was a three (3) - 9mm diode .. knife edged combined....then all three beams put thru a pair of Anamorphic prism's for correction. I wonder how my Lios BE will handle that output....which was 6.7W combined optical power.

As Doctor Frank N Stein sez......" It...............Could................Work " !!

Beam Out

I'm pretty sure It's 3-elements lens collis on this particular setup..."2mrad no correction". typical side knife edge llll , no vertical stacking...simple design due no axis correction but worse FF beam profile.....3-elem+Ke+be. IMO
btw....2mrad to 0,08mrad...so 25x beam expander.....??

 

[jors]

Initially I was wanting 2.5 inches or close to 64mm, but 2 inches/50.8mm is enough to take the divergence way down.

You could end out of prisms (or cyls) say... 6-7mm beam @1mrad aprox (G2)...+8x B.e. ...resulting in 48mm @0,12mrad beam...to have an idea...
on a 10x.... 60mm beam @0,1mrad etc
(Please Woody correct if I'm wrong)

 

[CDBEAM777]

Well, severe angle or not, depends on magnif. you want to achieve.... you probably have seen modules like this.... This is a pic of 12 diodes MM module on PL (10w green)
Photonlexicon

Very Interesting...I do see a set of Anamorphics at the final stage !!
Well....the diverergence of say ...the Red's MM ( P73/Oclaro) is too great for Anamorphics...that is certain...Now...maybe the Green's aspect ratio is not as bad as the Red !!! Great !!!

I will experiment with a NUBM01T ( Lower power version of a NDG7475 )...and a set of Anamorphics I have in the lab !! I will see what the incidence angle must be to accomplish correction. If the Zoof angle...perfect !!

If this works...Wonderful !! Thanx Jors !
In what section was that on PL ??

Just gather up all those beams...and put'em thru one set of " Glasses "
No problem !!

At the end of the day... The goal is to pac four (4) beams into the host...in the most efficient and less complicated fashion...and still achieve the desired results !!!

With the current Green Quad design in my posts...is far from ...er....ah...un-complicated....I am fairly certain...it will work...but...if there is a more simple, less complicated solution...let's go that way !!!!!! All depends on how much correction one wants/needs to achieve...and to what do to get there !!


Beam out:beer:

 

[jors]

Very Interesting...I do see a set of Anamorphics at the final stage !!
Well....the diverergence of say ...the Red's MM ( P73/Oclaro) is too great for Anamorphics...that is certain...Now...maybe the Green's aspect ratio is not as bad as the Red !!! Great !!!

I will experiment with a NUBM01T ( Lower power version of a NDG7475 )...and a set of Anamorphics I have in the lab !! I will see what the incidence angle must be to accomplish correction. If the Zoof angle...perfect !!

If this works...Wonderful !! Thanx Jors !
In what section was that on PL ??

Just gather up all those beams...and put'em thru one set of " Glasses "
No problem !!

At the end of the day... The goal is to pac four (4) beams into the host...in the most efficient and less complicated fashion...and still achieve the desired results !!!

With the current Green Quad design in my posts...is far from ...er....ah...un-complicated....I am fairly certain...it will work...but...if there is a more simple, less complicated solution...let's go that way !!!!!! All depends on how much correction one wants/needs to achieve...and to what do to get there !!


Beam out:beer:

Here's the Post on PL:

Post #34 #66

lightspace ascent-x series - Page 2

Sure!! Despite I'm a fan of cylindricals (less looses), IMO prisms are easy to work with when going with 4 or more MM diodes (except reds, of course).
So, yes, Nichia 1W green can be easily corrected with prisms for 1-1,5mrad

For 4 diodes...multiple options here...it all depends on budged, size required, beam specs...
We all know lots of possible setups: cyls vs prism, llll vs ---- vs ll+PBS+ll vs l over l +PBS+another set+only ONE cyl pair (like we do on P73 reds) etc etc

We could take long long time analysing each option but IMO, for a handheld device, where size is a MUST (thus simplicity), high align stability is required...so robust...near to hits immune... and we are not rich...(me not!!) I would like to go with:

*****3-layer lens (8mm EFL):
Taking into account we are going to end with high beam expander (increasing gaps), it is good to start with a 'reasonable' expansion before knife edge...so less prism expansion is required after = thinner gaps after prism correction

-G2+knife edge....more prism expansion required...fatter gaps :undecided:
-3-elements+knife edge....less prism expansion required...thinner gaps:wave:

on the other side.... G2 less losses BUT more losses on prisms (high prism expansion required)
3-elements: more losses BUT less losses on prisms (less prism expansion required)
g2: more power but more junk
3-elem: -15% power lost, but cleaner spot


BUT, As I said...we are going to END with HIGH spherical B.E., so we need thinner gaps before expansion...for me, no doubt...3-elements here.

*****'Standard' side knife edge: llll
yes, yes, PBS more energy density...vertical stacking better FF beam profile...I know...BUT I said simple, affordable, small footprint, stable...robust!!!

so IMO 4-in-line diode host, same diode weight, no vertical stacks... and 4 mirror mounts for side-knife edging...one bolt+prism mirror type best option (really stable)...or angled flexmounts with flat bounce mirrors..etc

-So we now have llll and 1,8mrad

*****PRISM CORRECTION 2-3x magnification
With 2x expansion....we get 0,9mrad...not bad :beer:. Have not data here, but not high losses at 2x prisms setup...If I remember.....5-8% ?¿?


****Then HIT the BE....for example 10x to get 0,09mrad and a really fat beam..I expect 60mm (very very aprox...not tested)

 

[CDBEAM777]

Here's the Post on PL:

Post #34 #66

lightspace ascent-x series - Page 2

Sure!! Despite I'm a fan of cylindricals (less looses), IMO prisms are easy to work with when going with 4 or more MM diodes (except reds, of course).
So, yes, Nichia 1W green can be easily corrected with prisms for 1-1,5mrad

For 4 diodes...multiple options here...it all depends on budged, size required, beam specs...
We all know lots of possible setups: cyls vs prism, llll vs ---- vs ll+PBS+ll vs l over l +PBS+another set+only ONE cyl pair (like we do on P73 reds) etc etc

We could take long long time analysing each option but for me, for a handheld device, where size is a MUST (thus simplicity) and we are not rich...(me not!!) I would like to go with:

*****3-layer lens (8mm EFL):
Taking into account we are going to end with high beam expander (increasing gaps), it is good to start with a 'reasonable' expansion before knife edge...so less prism expansion required=thinner gaps after prism

-G2+knife edge----more prism expansion required----fatter gaps :undecided:
-3-elements+knife edge----less prism expansion required---thinner gaps:wave:

on the other side.... G2 less losses BUT more losses on prisms (high prism expansion required)
3-elements: more losses BUT less losses on prisms (less prism expansion required)

BUT, As I said...we are going to END with HIGH spherical B.E., so we need thinner gaps before expansion...for me, no doubt...3-elements here.

*****'Standard' side knife edge: llll
yes, yes, PBS more energy density...vertical stacking better FF beam profile...I know...BUT I said simple, affordable, small footprint...robust!!!

so 4-in-line diode host, same weight, and 4 mirror mounts for side-knife edging...one bolt+prism mirror type...or angled flexmounts with flat bounce mirrors..etc

-So we now have llll and 1,8mrad

*****PRISM CORRECTION 2-3x magnification
With 2x expansion....we get 0,9mrad...not bad :beer:. Have not data here, but not high losses at 2x prisms setup...If I remember.....5-8% ?¿?


****Then HIT the BE....for example 10x to get 0,09mrad and a really fat beam..I expect 60mm (very very aprox...not tested)

WOW...not that is a very complete answer....Thank you...and I will need to re-read many times...to make sure I understand !! But...

I just did some power loss testing. :cryyy::cryyy::cryyy:

The Optic Gods have not smiled on us this day !

Set-up

NUBM01T with a G2 Collimation lens driven by a Lazereer driver.

All set-up measured at approx.equal divergence geometry at the Far field.

Here is what I measured...with a Scientech 362 10W Meter;

-------------------------------------------------------------------

No corrective optics other than the G2 Collimation lens @ 1450 mW

Cylindrical Optics BBar coated (6X) @ 1400 mW

Anamorphic Prism BBar coated ( set at the Zoof angle) @ 1125 mW


This calculates to;

3.4% loss of transmission thru the Cylindrical Lens pair

22.4% loss of transmission thru the Anamorphic Prism pair

It would be great if this were verified by someone else !!! Scientific methodology is picky that way !! Perhaps my method is flawed ???

IF this is indeed true....then...for my buck....I would not want to loose
325 mW per diode....or 975 total mW loss in a quad set up.

Yes...there will also be losses with the Cylindrical lens use...projected at 50 mW loss per LD....or about 200 total mW loss in a quad set up.

Trust me....I would much prefer the use of Anamorphic prism's....much less complicated a set up.... but....not at that difference !! I want to squeeze out every photon out the front end !!

Beam out :eg:

 

[logsquared]

Here is a design I did a few years ago. Just another way to arrange things.

[/URL][/IMG]