Planning an aspheric scratch build

djozz and some german forum and saabluster (deft X creator) did tests on it and it beat the old XP-G2, and the new XP-G2 is even worse.
that’s the LED I will use for my next light, sorry Cree :stuck_out_tongue:

I’m pretty sure by “very precise” vinz means he bought real optical grade precollimator lenses rather than some cheap chinese ones. When you have something that close to the LED, any imperfection will make the light unusable.

all right… start at the NA of the BEAM.

NAbeam = sin(theta/2) = sin (60/2) = sin(30) = .5
NAbeam = .5

now, THAT is a LOT nicer to work with. sweet jesus yes. thats “do able”

next?
we can single lens collimate that source beam? with any lens, so long as the NA of the lens is at least .5

so… lets soc-cah-toa up THAT… this will take me a few minutes… (theres a lot of drudge work to this, lol)

assuming for ease, 100mm diameter lens…theta 30 above the x axis… locks in the OPPosite as the radius of the lens, which is of course 50mm… and the ADJacent is now the true focal length…

sin 30 = .5
hypotenuse (useless, just a step stone) is… coming in at 100mm
cos 30 = .866, i am getting… adjacent (FOCAL LENGTH) = 86.6mm

meaning, a lens to completely capture the escaping 30 degree half angle, in one lens? is… 100mm diameter, 86.6 focallength… which is of course F-number = .866

for a single lens zoomie? using waiven collar and the 60 degree full angle??? any lens of F number .866 “kicks butt”.

=

equation #2? full beam divergence (post collimation), expressed in radians?

THETAprime = (DIAMsource / Fcol )

THETAprime = (1mm / 86.6mm)

THETAprime = .011547 (rounded off, useless digits out the butt, i guess six is enough) “radians”

(i wanna convert radians to regular degrees, just for visualization purposes)

.011547 Radians = angle in radians * 180 / Pi
(.011547 Rads * 180) / pi
= 0.661594 degrees “divergence”
(eh? what? whatever… keep moving…)

===

this collimated beam diameter… right after this lens, can be estimated by….

BEAMdiam = 2 * NAbeam * Fcol
BEAMdiam = 2 * (.5) * (86.6)
BEAMdiam = 86.6 ??? what are mu units? mm, i think… so 86.6mm ? (worry later, keep moving…)

=

okay, barring any gross errors (smeone follow along the playbook, and make sure i didnt boo-boo this, ha ha) thyen THIS is the completion, of what is purported to be, a complete mathematical model of a “proper” one lens zoomie, if the waiven lets out a 60 degree full angle, 30 degree half angle.

(with the example lens, of course, 100mm diameter wa a nice round number to aid in calcs)

=

let me enjoy a SMOKE, and i can take another WHACK at the million dollar question, LMAO…

aint but 2 stinkin equations left, we will just have to check THIS work, naturally… and see “what” the answer to the next 2 equations “means”… but who cares? lets just GET IT, we can worry about “exactly what” it is later… i never GOT this far before…

lets DO THIS…!

now i am going to enouy another smoke, and wonder what in the &^%$ i am going to “pick out of a hat” for the “ratio” of focal lengths, that somehow “sets up” the Numerical Aperture of the “precollimated” system.

i think you guys just got “your answer” unless i missed something or made a gross mistake, but… you guys wanted to know the description of the “optimum lens” for your waiven collar fed single lens zoomie, right?

glad “youre happy” i dont have MY answer… (yet)… lets see what we can come up with…

(and i am open to suggestions about that ratio of lenghs, to set up the numerical aperture of the system, lol)

=

still? we’re in a better place than we WERE… arent we?

now… stopping right now even, wh should or might have a 100 hundred percent complete optical engineering workup, or exactly how a one lens zoomie works. complete math model… barring i made a stupid mistake, or, barring the engineering white paper contains some “updated error” we dont kow about… but still.

this should “work” as the complete math model, acceptable to an engineer, to describe ANY emitter, and any one lens zoomie set up. we might need “a crazy lens” made up out of 2 lenses touching to do it, but… still. at least we know, we are not “guessing” and saying things about “throw, duuuuude”.

i am beginning to be something resemling “contented”.

of course? its still an “epic fail” if i cant do the last 2 equations, and get MY million dollar question answered then understood.

(and i am still open to “ratio” aperture set-up ideas…lol)

right before i take a whack at this final assault? i would like to point out…

is it just me? or did i finally “prove mathematically” that in a one lens zoomie… “diameter does not equal throw” and that it is now “mathematically incorrect” to say things like…

“only diameter gives throw… focal length is meaningless”

its simply not true… its the numerical aperture that means everything… and even a SMALL diameter of lens? is just as acceptable at pushing “moon bats” thru the lens, so long as the ratio of focal length to lens diameter is acceptable.

which is to say, barring mistakes…

F-number is, tentatively proven…to be the final authorityof the suitablity of a lens? not “diameter only”.

just saying…

okay. 2 down… 1 to go.

now that we are done feeding the multitudes? we might s well try to take a stab at walking on water while we are at it, shall we? *^%$ it… ha ha

(i either hate or love, working above my pay grade like this, lol… its a rush…)

well… since bulldogging thru, worked this good so far? might as well go with THAT for this session… no sense wasting being in the zone…

QUOTE
2 Aspheric lens pairs
A second aspheric lens can be added in front of the collimating lens
to focus the light over a desired area. The illuminated
spot diameter at the focal plane (FP) is given by:
ENDQUOTE

well? we can do THAT now, if everything else was correct… screw it…

spot diameter = DIAMsource * (Ffocus / Fcol)

spot diameter = ? well, we know if nothing else, the diameter of the source, thats 1mm, right?

spot diameter = 1mm * (Ffocus / Fcol)

=

we got half of this, anyways… the spot diameter is 1mmTIMES THE RATIO, and a real nice number. more importatly? the millimeters inthe ratio will cancel out wharever they are… the answer is in millimeters, left over last unit from the 1mm source diameter, after canceling out?

million… dollar… question….

what freaking ratio? we are 1.5 equations away, from walking on water… i only know,i got the tiger by the tail and dont wanna lose it…

okay… stop… think??? we can SET the spot diameter after the lens pair? to be whatever we WANT it to BE…just, what do we ant it to BE?

uhmmmmmm. in the LAST one lens zoomie math model? using “best lens scenario”, above which the extra lens power is “useless” as the entire set of “rays” escaping the waiven collar is captured mathemastically? what more is there to “gain” than THAT?

okay… lets “go with that” assumption… we only GOT THAT optimum setup? simply because we used a waiven collar, which fired up the emitter hyper, and contained an escape angle of 60 full, 30 above the x axis… right?

so… maybe, and i mean just maybe? this portion is reserved for “getting optimum” assuming we DIDNT use the waiven collar? that we COULDNT find a suitable lens with the proper stupid numerical aperture… which is simply the F-number hiding under an assumed name…

so… we… what? we want to SET THE RATIO so as… to accomplish this, assuming we NEEDED To.???

i’m thinking out loud…. i am talking myself into this line of thought… we are only stuck here, simply because we are ALREAY capable of finding lenses that are “optimum capture angle” already with the one lens model.

what if we WERENT ABLE TO be “here”? which is any time without THIS special emitter, without THIS special case of using a waiven coller… we WOULD want to “be here” regardless, right???

and… setting THAT RATIO, will PUT us where we want to “be”.

=

okay, i talked myself into it…. i am forgetting the whole single lens zoomie workup? that was just an academic exercise to get me HERE, working with THIS portion…

go back, see “what” was optimum about that setup…we RE-make it HERE, with this stupid ratio?

shrugs i like the logic train of it….

houston, we have a problem… we never calculated a “spot diameter” back there… all we calculated was a “beam diameter”. is that the same thing? engineer dont use terminology lightly… whatever the differenc ebetween a beam and a spot “is”? it might mean something… of course, it could also just be a convention based on whether you use one lens or two lenses.

decisions… decisions… decisions….

also? any clue i take from the previous session? is “switching backwards”… if i like to think of calculating this spot diameter, as the same as the beam calculation previously? the one lens collimated beam diameter, was the 2NA*Fcol equation… here? its “more like” the previous divergence formula… and my next equation, also containing this holy ratio we set? is switching the collimator/focus ratio upside down… its a reciprocal… and, thats the divergence equation in THIS section, looking more like the other equation back there…

what to do?

okay…. pondering the pnderables? lets say maybe the spot diameter and the beam diameter arent the same thing… i dont know… BUT, i am just bout to calculate the divergence, and while the equation is way different than last time? a term is a term… divergence is… divergence.

since i am trying to assume that we couldnt set divergence “optimum” and now would be our chance to do it? we couldnt get the cool lens we had for optimum last time? then i want my divergence set to “optimum” we got lucky and HIT optimum in the single lens, maybe i can use that as an “answer” then solve for the ratio i want? that will produce it?

its a thought… well, thats one variable out of the way… i need ONE of those ratios to be set for me…

hmmm. maybe we COULD use the “single lens” optimum, use that as the bigger forward focusing lens? i mean, lets not waste it… thats another variable down… where we at in the second equation? if i can SOLVE for the ratio last need, i will have the ratio to plug into the previous equation?

scrtew it, lets try that…

the optimum theta divergence, only because we HAD an attainable lens? WAS an answer of… ehat was it?

optimum radians? WAS 1 / 86.6…

so…

(1mm / 86.6mm) = 2 * NAbeam * (ratio)

mm’s cancel out…

.011547344 = 2 * NAbeam * ratio
.011547344 = 2 * .5 * ratio
.011547344 = 1 * ratio

BUt, following this line of thought? ratio = reusing the cool lens, as the “forward focus lens”?? equals 86.6 as the bottom of the ratio…

.011547344 = 1*(Fcol/86.6)

? eh

now, need the “new” collimating lens, to be… smaller than the focusing bigger lens… which is 86.6…. if i solve this, i might get the focal length of the new collimating lens, hat used with this lens, would make the optimum lens setup, assuming i didnt already have it available with one lens previously… hmmm, watch this give me a nonsensical answer, LMAO

1 times anything is anything, i am left woth…

.011547344 = Fcol/86.6

.011547344 times 86.6 = Fcol?? try it…

i get “1mm” as the proper focal length of the collimating lens….

wait a minute! that actually might make sense!

since we already HAVE the “optimum one lens”, and i was using that radians answer to derive the rest of the ratio? it is stating that the “ratio” is already set up? maybe?

okay… thats either nonsense, or it means maybe what i think it might mean… lets try this… with… a “lesser lens” and solve for the same radians answer, the optimum we already know from the one lens example…

maybe i can “hit it” based on any focal length lens i CHOOSE to use for the focuser? okay, lets se what it would be for… a 100mm FL focusing foreard lens… nice round number…

.011547344 = 1*(Fcol/86.6) scratch that….

.011547344 = 1*(Fcol/100)

1 times anything is anything

.011547344 = (Fcol/100)
.011547344 = fcol/100

multiply both siddes by 100, duh…

1.1547344 = Fcol

heck, a ratio is a ratio, right? lets fire this puppy in with a… 600mm forward focusing lens… we’re just setting a ratio, right? maybe THAT will get the focusing collimator up into something reasonable to fiond in a col. lens??? worth a shot. i can create a ratio out of thin air like this… and i can still run any ratio back up thru the “spot diameter” equation see what it spits out…

.011547344 = 1*(Fcol/100) scratch that… bigger number, to get a realistic collimator mm fogure…

.011547344 = 1*(Fcol/600) now we’re either totally lost? or were cooking with gas…

.011547344 = (Fcol/600) cos 1 times anything is anything…. multiply both by 600?

NOW i get Fcol = 6.9284064 = 6.3mm collimating lens

THATS more realistic…. to “find”… in asmall collimator…

=

i am NOT calling this solved? i’m saying this is a possibility….

just, let me feed this “600 ratio” back into the above equation? see what he spot diameter is… maybe it will be similar to the “optimum” one lens beam diameter was? i am either close or done? or completely lost…

spot diameter = DIAMsource * ( 6.9284064 / 600) source diameter always was 1mm…

spot diameter = 1mm * (.01157344)

spot diameter = .01157344

does that even make sense??? check it… look for it…

HEY!!! i got rthe SAME RADIANS!!! witha totally different equation!!!

==

i’m am NOT calling this solved? i am just calling this… best whack i ever took at it…

someone suffer thru this logic trainwreck see what i might have had abrain fart? or else… i am a golden god? (and i just cant be… i aint that smaert… there has to be a monumental logic *&%$ up… this is way above mypay grade…)

still… i got “something” working… sort of? pinch this off, before i hit a bad key and lose all this work…

if i even think thid might be a solution? backfeeeding mt ratio, in the spot diameter estimation? what the &^%$ is my answer supposed to be units wide? radians? see i was thinking millimeters…

nut… still… the WAY the equations are flipping around, and reciprocals of reciprocals back and forth? maybe the answer IS IN radians… i dont know…

and? i am about to come out of my “zone” and “lose it”… i need a bath… and sleep… butonly after i decide if i epic failed or walked on water…

i mean, these new equations? are “sort of” looking like the opposite equavelants back trhere….

anyone knoe if calling something a beam or a spot, indicated whether the anser is experessed in radians or millimaters?

all i know is, i am getting “realistic LOOKING” small collimatror lens figures out of this…

my brain is about to explode….

good night and god bless. i need a bath… and hey? atleast you guys got your answer… i just pray i got MINE.

(i wanna be able to define the collimator lens build, gawd dammit)

and? just in case i did walk on water? THESE collimator/focusing lens focal lengths? would be the puppies to look at ray traces… if i am even CLOSE to anything remotely resembling “right”? something might “pop out at you” on the ray traces.

keep in mind? you would want to use “these” focal length figures… and move them back and forth see if the rays started lining up cool, i guess? (i always assumed if things were working, you could just “see it” heppening in rays… i really dont know nothing though)

bed… bed… bed…. i’ll find out tomorrow how close i got, maybe…

how in the *&^% do we even “check my work” ?

Sedstar and Enderman,

Guys sometimes all maths and calculations means nothing without experiment. Lets build something out of available and known lenses scraps for cheap to get approximately close to something that we could call multi aspheric flashlight.

And just to mention old de domed XP-G2 S4 2B(R.I.P) was and remains the best throwing emitter and the most suitable emitter for fet driver setup.

Osram Oslon Flat could serve for smaller setup and it is fet driver friendly but it is XP-E2 size (so stick it on 70 + mm lenses and wow! Tiny laser :slight_smile: even with precoolimator ) but we have XP-E2 which after de doming should be same or better than that even at lower currents.

Djozz about Osram oslon flat black:
“Much encouraged by these results, I did a Brinyte B158 mod with this (very) led and a direct drive AK-47A+FET driver (ledboard glued to the pill with an as thin as I could make it layer of Arctic Alumina Adhesive to get the required electrical insulation between led-minus and pill) in the hope that it would crush the throw of the dedomed XP-G2 S4 2B led, but unfortunately it did not, I got ‘only’ 285kcd from the Oslon BLack Flat, with 322kcd for the S4 2B”

luminarium?

technically speaking… equations are equations.

assuming the following…

1) the given stated equations i worked with were correct?
2) if no one can detect any “algebra errors” in my flight plan?
3) if i didnt make some huge logic error in setting up my final solution?

the work should be “checkable”.

i am a “lot more confortable” with the mathematical modeling of the single lens zoomie…
the “way” i went about my (possible) solution of the pre-collimation calculations? flipa coin, i figure i could be totally right, or, totally off base…

i cant check my own work? i’m “in love with it” i need peer review…

but…my gut says i am “probably right” on the single lens model…. and that i cant rise above “50 percent” on the collimation equations…

but, “either or both” if they would be right? are now repeatable, and if right, i could use to find stuff. and this is ll assuming i wasnt “given errors” woth anything wrong in the edmunds scientific paper.

but… those “precollimaor” fogures? look reasonable…
which dont mean nothing, but still…

question?

what was that “height” that vinz claimed was the holy height for the precollimator? and what lens was the focusing lens? i mean rough focal length?

if i got close to THAT,a given working solution? that would be a slightly warm fuzzy feeling…

Here: Dereelight DBS with de-domed XP-G2 and pre-collimator lens | Candle Power Flashlight Forum

Everything he wanted to reveal is in that thread(4 years ago). Main lenses are Dereelight DBS which are same as nightmaster(V1 V2andV3) and I really don’t know nothing about focal lenght.

He said around 0,7mm from emitter and it is… The higher you’ll go the larger projection and larger lux drop therefore you must hit the sweet spot.

Vinz is the best modder in a world.

Edit: Right one are main Dereelight 50mm lenses Vinz used in mentioned build:

…and i have to entertain the NOTION that vinz would lead us all astray, on exactly “how”… this is his bread and butter, and his reputation as top dog on the line…

those pre collimation equations? they “compleyely” rely, one way or the other? on the RATIO, not on specific vales…. that much is obvious and evident form looking at them.

technically? a RATIO, has any possible (infinite) number of “solutions”… one you assume ONE, the other should be solveable?

thats the way i’m seeing it, though… giventhe forward focusing lens? you can calc the FL of the precollimator… and given the precollimator given? you can then calculate the new forward lens.

unless i am missing something. but if i am, i am not seeing it. (duh) why would i see it right now…lol

all right… my ego completely aside? I need this from you…

1) COMB thru my “work”. if i make any error in algebra? or in my trig “soh cah toa” descriptions? someone needs to POUNCE on me for it. if i dont stand to peer review? i dont have a CHANCE.

2) if no one can find an obvious “error? then theres a decent chance my math description of the single lens zoomie, best build might be an accurate model.

3) same thing in my “crazy man rantings” on the second set… if theres an algebra error? POUNCE on me. tiniest thing matters.

4) then, if anyone finds a “logic error”, i could have easily made a false logic assumption. if you think you see something? maybe you see something.

5) one major assumption i am making? is that all these equations? form a giant “tautology”, which of course i might be in huge error of that… the first set of equations, might not have anything to do with the second set, and i am making the giant intuitive/deductive step that they are.

6) on the initial one lens model? if the ray trace even remotely resembles anything in real life? then, we should be able to see SOMEthign in the ray traces… one lens should at least INDICATE something… and namely that “a bigger diameter lens wouldnt do more” if it reads like i think it does.

7) the second solution? flip a coin… at this point i cant know.

I’ve researched it, the XP-E2 curve starts dropping down much sooner than the flat black, at about 3A.
Everyone knows the old XP-G2 had the best intensity, but both djozz and saabluster tested a single flat black and it had higher intensity than the G2. Djozz probably got worse with the flat black due to not enough current due to the Vf. This is why I go with constant current controlled test results and not “in-flashlight” test results.

This is what saabluster had to say:
“Sample size of one in my tests.
Relative Beam intensity
2A
BF 1040
XPG2 942”

Nobody is talking about “lower currents” here, or inconsistent FET. The LED will be driven at the peak of it’s curve through constant current drivers. Since the black flat is not sold in sorted bins, I plan to purchase about 10 or more and bin them myself to find the optimal one. If Saabluster and DJozz got better intensity at the same current with a sample size of ONE led, I’m interested to see what a binned one can do.

Another plus is that the flat black doesn’t need to be dedomed, which will make testing all of them easier.

PS- when I do build this light, I will use coolaboratory liquid ultra, which is a liquid metal paste and performs much better than regular thermal pastes. The only thing that would beat it is soldering the MCPBC to the heatsink, which I would rather not do, ever :stuck_out_tongue:

mannnnnn. i hope i didnt screw the pooch royal, lol…

coolaboratory liquid ultra :question: sounds interesting and now I got to order it… :slight_smile:
I did soldering of mcpcb to brass pill but I did not notice any significant improvement over silver paste neither I noticed significant difference between copper, aluminum and brass pills (well up to 5A setups I probably can’t see any difference… At higher currents that variables could matter)

But Osram is still to tiny imho i even consider de domed XP-E2 to tiny for my regular setups (last time I used them was in Jacobs A60 reflector light since it was more than fine for Jacob stock driver that draw 2,2A of current and for some small aspherical builds, still have some of them).
It would need some crazy precoolimator to make it usable on 70+ mm size aspherics.

Waiting for delivery of lenses from surplus shed. I think I have a couple of two lens combination that should work nicely. One uses the 100mm 135mm fl that I ordered from fasttech and the other a 73mm 148mm fl from ss.

Good luck with lenses :+1: