I have one of those that i plan to try out, i hope i got a good one.
I seems the general consensus is that cheap glass lenses very often are of pretty rubbish quality unfortunately.
This is what I have been working with. Well, glass substrates vs metals, and losses/gains through some different production methods. Trying to get something useful that’s still affordable using an adjustable mounting design. The biggest problem for people is in attaining the proper focus of the reflective aperture while mounting it.
I have been working on a much bigger platform design for a RA to sit, though.
Do I have any volunteers in here with caliper and any or all of 1405/1406/1407 models? I would like to get pill dimensions on all 3, also body interior dimensions on all 3. I just finally made my order, so a couple weeks before I even hold one. 
The higher the “F#” of the RA (I’ll say higher F#, meaning “smaller aperture hole size”), the more tint shift that will occur with the same white LED. If more “white” LED light is recycled back at the die, the ~450nm blue light within it (originating from the InGaN semi-conductor), gets converted to more green/yellow/red light by the YAG:Ce (Y3Al5O12:Ce) in the phosphor. The energy remains from the blue light in the form of its initial eV, but its eV is distributed to lower energy wavelengths by the photoluminescence process. The LED die will not produce or excite any more blue light when light is recycled, because blue LED light does not originate from the phosphor composition, but from the semi-conductor. Blue color, requires electrons to excite the PN-junctions and form, not photons. Throw photons back at the semi-conductor, and you will make electrons. (Hook a volt meter to an LED and shine another light at the die. You have a photo-diode now. 
) So, tint shift will always be relevant to how much blue light is recycled, and the phosphor matrix. As the other colors strike the die upon their reflection, they are not converted any more to longer wavelengths, which I have found in testing by use of monochromatic light.
Mem,
You know that I would took measurements but I sold longer brothers and now I am waiting for smaller ones 
( probably you'll get them faster than me).
OK,
I have sourced some really good acrylic 50mm aspheric lenses for the 1406, precision ground B270 glass 50mm for 1406, and also 67mm precision ground B270 glass for the 1405/1504. Also the 30mm and 35mm models should be covered as well with precision ground glass.
Does anyone know the back focal length (lens rear plano side to emitter focal length) for the 1405 and 1406?
If anyone has these lights yet and can get the BFLs (the maximum BFLs at full extension will do), and they are within lens spec, I will go ahead and purchase a couple of the lenses for some test runs.
I will be home tomorrow night, if no-one beats me to it I will measure both lights (within a mm or so).
If i understand what BFL’s is correctly i think my 1504’s measurement is 42mm, from the edge of the o-ring to the emitter surface.
But because i can’t place the calliper right on the emitter and the edge of the o-ring, i have to just hold it in air & make an estimate, and by my calliper & eyes it looks like 42mm.
But to get a perfect focus on a soldered XP-L on a sinkpad in the brass pill, or should i say as best as this stock lens will get me, i need to unscrew from max zoom about a quarter turn. So maybe the stock lens have an BFL of about 40mm, again if i understand how this works right that is
I would wait for djozz to confirm my measurement just in case i made a mistake.
I measure it out at 40.13mm to the surface of a noctigon. So the Cree sheets show the LED package base to LED die surface at about .73mm so that would make the total 39.40mm. I also compressed the O-ring a tiny bit to correct for the pressure the lens would likely put on it. But put me in the same camp as cajampa and wait for djozz.
The .68mm added thickness of the noctigon over the stock pad seemed to throw my focus off just a hair and my plan to correct this was to machine the back of the pill just slightly. This of course is not easy to do without a lathe (although I happen to have a lathe) but the opposite of shimming the head lock point for a shorter FL can be done with just a piece of tape so for an easy to fine focus adjustment a little too shallow is better than too deep. Or at least that’s how I look at it.
Also it seems I have the opposite focus problem as cajampa which is interesting, I wonder how the sinkpad vs noctigon thickness compares or maybe it’s that I have an aluminum pill and the brass pills are slightly different, or more likely you can’t expect that much precision from these. Oh and to that point you can probably make your focus land dead on cajampa if you put some tape or something on the part of the pill that stops the head.
Interesting comparison
Yeah it is weird we seem to have the opposite problem, i wonder if it is because i have soldered my mcpcb & you used heatpaste instead.
I have some aluminium pills coming from n10sivern so i will see if there is any difference on the pills later.
I am not sure i understand how & where i should place the tape to get it to stop at the right focus.
And i very much agree that it is much preferable to have the ability to overzoom than not even be able to get a proper die spot.
Maybe you could use a little less heatpaste or double up on the o-ring.
When you’re zooming out the head it’s stopped by the back of the pill, how the pill overhangs of the size of the body so if you put something on the back side of the overhang or on the area of the head that contacts it, it will stop the zoom when those two faces contact. Might as well be something thermally conductive since heat can move from the pill to the head through that surface.
Pictures borrowed from mholdenattorney.
I’ve thought about another O-ring but it should be pretty easy to machine back the pill but if you were picking a lens a mm or two of adjustment room in one direction would make it so anyone could tune the focus without any machining. The retention ring is quite tight so the paste layer is probably quite thin and I lapped the pill a little to give a smoother hopefully leveller area for better thermals.
Ok, i got where i should be putting some tape or something.
I tried some but one layer or 2 layers didn’t seem to make any difference, i guess i can try adding it on but so far 0,5mm did hardly change the focus at all.
And i am not sure but, i wonder if focus changes by the distance i want to focus at, it seems like it. Maybe a little adjusting room at the spot point is good, not as much as a quarter turn but a little.
My presents arrived.
I ordered brass and AL host pill forms of 1504 to start.
My first impressions of the 1504s: Overall, not a bad light/host to build off of. The glass is OK, though there is much to be desired. There are notable casting inconsistencies on asphere side.
Definitely grease threads.
The pill configuration is the biggest limitation out of the box, though not something a workaround can’t fix. Once the LED can be brought to a level plane with pill top, a recycling aperture could be used—not until then. A shorter focal length lens would allow the LED to come up on the pill for it to work.
I’m going to whip an idea up for the 3D printer and try something.
UPDATE: Focus falls short of FL in ALL factory 1504 setups I’ve now tested. I’m sure you guys are discussing that already. My solution is to fix this problem at the bezel, not at the pill, for maintaining the factory lens. So I am going to first print a lens spacer-ring.
So we will need to mod pill a bit like adding copper chunk to make LED level with upper part of the pill? Other thing will be be a little problem (swapp main lenses with shorter focal length) cause they are hard to find.
It would be great if MEM will manage to made plug and play upgrades for this light but even if that will not be possible I already see 1mcd 1504 versions made by MEM :)
KKW made a small RA that fit inside the pill, and got a 20% gain. So it can be done but, with much less performance than the big RA’s you have been making & getting 50-70% more performance from.
A spacer + new lens + an RA sound complicated, but i am sure that is needed to get max performance out of the 1405/1504 platform.
If you can whip up a upgrade kit i am interested 
I look forward to see what you come up with MEM
Well I’m thinking go down to a better 60mm lens and use spacer. Just because that’s the good lens I see and it could outperform this stock lens potentially. The even larger one has a F# of 1+ so I think FL will be way too long with that lens, which leaves a 60mm.
Inside front bezel is about 61.25mm ID average. With a 60mm lens, lens holder covering 1.5mm around lens edge, puts you at 57mm. Not too far away from starting useable aperture. Close enough to make it worth the trial anyways.
I haven’t had focus issues with my 1405’s, maybe I’ve been lucky.
That was freaking great. Every detail about what I am seeing with the factory lenses I had just written, and detail of the following processes to arrive at their expected gains explained.
I write out upgrade paths #1, #2, and #3 for 1504 that we can get to, with about 30 member (or 30 light) commitment, and apparently whatever button I clicked opened an HTML debugger I had no idea I had in Moz FFox. Which sent me back to writing a completely a blank post. Yeah, I need a moment before keyboard keys fly.
Will be back to re-write everything I just did. Lesson learned, never type the message you intend to send inside of the browser. |(
How about this as my post recap:
Stage #1. Short and sweet drop-in, factory lens AR-coated via vacuum desposition, 30 people needed at roughly ~$30 to get their lens processed and coated for 9-12% lux increase (plus shipping back). Top-notch AR process with “oleophobic” top-coating layer to repel hand oils as well.
Stage #2, 15-20%+ lux gain, 3D printed lens adapter, precision lens, same AR-coating as above, $100 drop-in kit boost.
Stage #3, includes stage #2, plus custom copper pill, plus 3D printed reflective aperture mount stage + broadband coated reflective aperture, ~$175. 50-60%+ lux gains (~750kcd 1504s). This would also require you to take the non-optical hardware components up to maximum drive potentials with your driver+de-domed LED to reach such high proclaimed lux values. Without the optical additions, though, and your 1504 limitation will remain around 400-500kcd average as it has been. Again, the magic number is about 30 commitments for this.
These are approximations. For stage #2 to be final, I have to complete testing with the lenses I ordered earlier today to be sure FL is no issue in the 1504s, which based on specs—it shouldn’t be. But I still must confirm this. I will get more accurate details, but I am just about certain these will be close to actual costs to go forward with such upgrade paths.
P.S. I will be updating the “Poll thread” I started on the multi-head multi-LED light for tripod use that was in question a few days back. I will update later tonight with further details about those projects. I will have some photos showing some progress and a comparison aspheric build next to a little 1504. Stay tuned if interested. Would you be interested in a powerful multi-LED spotlight if it was somewhat large and required a tri-pod?
Hi Mem,
I think your link ( http://www.budgetlightforum.com/node/39815/results )does not work for me? Would like to see all mentioned described with pics. Picture says more than 1000 of words.
Language barrier bothers me so it is hard to me to follow completely... So I will have few questions?
- Is expected throw(750kcd) in 1504 with XM-L2 or XP-G2 emitter?
- Does this upgrades needs to be fitted by us or you will sell complete upgraded 1504(all done by you).
- Will zoom out function work in ra modded 1504?
- I don't know if I understood this well but if we go straight to stage 2 build(new AR coated lenses with adapter) than we don't need stage 1 build (factory lens coated with ar vacuum desposition, I see problem for non US residents in posting lenses to you forward/backward post which should rise stage 1 to about 70$.
This also confuses me: This would also require you to take the non-optical hardware components up to maximum drive potentials with your driver+de-domed LED to reach such high proclaimed lux values. Without the optical additions, though, and your 1504 limitation will remain around 400-500kcd average as it has been. Again, the magic number is about 30 commitments for this.
I really don't understand so I really looking forward for your new thread "1504 upgrades"...
I fixed the link and I have the “MEMorized” J20 posted there for size next to a 1504.
Yes, when I lost everything that I had typed, I was in a scrambled panic to re-type what quick detail I could, as I had to get going shortly after. My apologies for the lost detail.
Good questions. 
For those in the USA, shipping would not be very much going back to you, in a well padded small box via USPS Priority, probably $6 or so. If you want to stick with the 1504/1405 build of light which you have, but are looking for a cheap ~10% boost in lux, we need to get a bulk of these lenses into a vacuum deposition chamber to be ultrasonic cleaned and then coated. That is the cheapest route beginning an optical upgrade path.
What I mean when I say “take your hardware up to maximum drive potentials”, is that I am proposing a lens system upgrade which will effect the optical-side of things, but to get your light to 750kcd+, you would need to drive the LED properly with the proper driver, correctly de-dome and use that LED. Basically I am stating that these mods are LENS mods for efficiency, not driver/electrical mods. Both have to work together to reach such high numbers of output.
If you are asking me about LED size, this host size is the size for an XP-G2, all the way. XM-L2 should be for 75mm+ lens and bigger heatsink sized host. BUT, that is strictly personal preference when going there. If you want less kcd and more light coverage, the L2 is entirely your option. Don’t expect continuous lux to be as high with the L2 as it would be with a G2, though. Less power to cover less die area is going to lead to a cooler LED and light, therefore average kcd numbers will be higher with XP-G2.
For option #2 I was going to explain much more about that. Currently it is my prediction those will be the gains in that setup but I can’t be too certain. The new ground lens should be here in a couple days for me to yield final test results against the factory lens. I do know that the lens I am replacing it with is much higher quality, without drastic visible mold voids on the glass. Which you do have to look for closely to see. Once you find them though, you can see why the stock ~67mm lens is a poor lens when you have a quality one by its side to compare surface quality to. Ripples and waviness go away. One common test I do with all of my lenses is look at image formation on a surface from the lens, taking light from far away into the lens, and focusing the image close (at FL) on a white paper. The image formation is very poor with the stock lens. Lacks great deals of contrast. This shows me that the lens doesn’t have a very optimized focal point for all incoming incident rays. Coating will not fix this physical problem. Regardless, whatever output you are getting with the factory lens—it will only go up in output once coated. So while it may not be the best lens after hearing me speak of it, coating it for some gain is cheaper than buying another lens. If you are really trying to stay on the budget side, that is the way to go.
Option #3 will be whole overhaul basically on light. New pill design with new lens so that no matter what, output will go far higher than stock using RA setup. I’m not offering to assemble each one per se, but if someone needed it, I’ve always built lights for members, so I’m not against it.
First I await the new lens, then I will go from there upon posting those results with output data .
I take it the RA would be printed with an SLA style 3D printer and not a filament style? Do you have a plan for how to polish them already? It sounds like you’ve made them before, if you’re willing to share any details with me here via PM or the thread in regards to the geometry, I might see if I can machine one on my lathe. The 20% limitation on the small collars I’ve made so far has me confused as to why bigger RA’s work better when a small one should be better at returning more light to the LED surface without loss.
I have been wondering the same thing about the size, to me it seems like it would be the opposite because any imperfections in a larger dome would be more likely to miss the phosphor due to the longer distance they have to travel. I could of course easily be overlooking something.
Also Kloepper have you documented your RA attempts anywhere, my search fu is failing me. I’ve been thinking about trying to build one and was planning on using a known sphere like a ball bearing to lap the dome. If that method works would be pretty easy lap a plastic RA like MEM’s.