Two primary "lenses", at this point, yes. A reverse reflector (recycler) I'm calling lens #1, and a large plano-convex as lens #2. More testing is being done with different lens pairs, currently. The largest task, machining, is ahead. I have to adjust some of the internal part parameters, and have new part runs performed on a CNC for me.
Mega Kcd Aspheric
- Thread starter matt304
- Start date
Help Support Candle Power Flashlight Forum
The_Driver
Flashlight Enthusiast
The SST-90 ist 9mm^2 without the dome. The dome magniefies the die. Basically the "perceived size" (how the aspheric "sees" it) is just about doubled because of the dome.
The SBT-70 doesn't have a dome. This means that the "perceived size" ist the same as the acutal DIE size: 7mm^2.
Try removing the dome of the SST-90. It is fairly easy. The intensity will increase noticeably and the spot will become smaller.
The SST-90 dome is removed, it was just there in the original photos of the test heatsink. You can see the increased size on all LEDs compared to a de-domed one, you're right, so to directly compare them, I had to remove the SST-90 dome. With it, the square was huge. The SBT-70 still contains the flat glass, however, in the photos above. It seems to lower the color temperature after light is recycled, more when with the glass left on. I believe it has to do with multiple image planes that are very close to one another, the wavelengths are shifted upon focus. This could be completely wrong, or have to do with the A/R coating scattering the light back onto the die differently, etc, but from tests, the light looks to become lower k when it reflects from the Luminus glass window, after that much I am not certain.
This 75mm lens is not precision ground in this test so this image is somewhat abberated. I have new large, precision ground aspheric lenses and a DCX condenser lens on the way, they should both be here weds. The lenses also have higher F#s without the usual loss of light associated with higher F# lenses, because I've found that I can get the beam very narrow and maintain the same output flux, before hitting the lens. So on a lens with 45mm FL that was 75mm diameter only 50-55mm diameter of light is actually striking that lens. Because of the lower F# here, the chip image has more chromatic abberation shown above than the new lenses will, at over 60mm focal lengths. 60mm focal length is about the sweet spot for this light head. If I move the driver to an external mount on the light, like a small machined aluminum housing box mounted on the body, I could achieve near an 85mm FL with no beam obstruction and a lot of LED cooling. But that changes the design entirely. The Ergonomics vs Function game.
This 75mm lens is not precision ground in this test so this image is somewhat abberated. I have new large, precision ground aspheric lenses and a DCX condenser lens on the way, they should both be here weds. The lenses also have higher F#s without the usual loss of light associated with higher F# lenses, because I've found that I can get the beam very narrow and maintain the same output flux, before hitting the lens. So on a lens with 45mm FL that was 75mm diameter only 50-55mm diameter of light is actually striking that lens. Because of the lower F# here, the chip image has more chromatic abberation shown above than the new lenses will, at over 60mm focal lengths. 60mm focal length is about the sweet spot for this light head. If I move the driver to an external mount on the light, like a small machined aluminum housing box mounted on the body, I could achieve near an 85mm FL with no beam obstruction and a lot of LED cooling. But that changes the design entirely. The Ergonomics vs Function game.
I had no idea whatsoever the DEFT-X used such a small aspheric lens. Based on the curvature of the lens looking at it, the F/# is probably like a 0.5-0.6. Running the base numbers, lux over chip area, chip diameter vs chip width (circle vs square), etc., this light is on the order of 2 DEFT-Xs, in forgiving simulations. The beam will start larger, but the beam angle will be lower in return. 1.5 million cd sounds too high to take in I know, but on the low-end (assuming imperfect optics, heat, efficiency losses), it should be achievable from the setup. Keep in mind though, the project scale. This light is physically much larger than the DEFT-X, so even if it hits 2 million cd, it's apples to oranges comparing them directly, because the bigger light should naturally hit a much higher Kcd when scaled-up optically like this. Also a total lux increase of 4x being squeezed into the lens being far higher than a G2 (sbt-70 vs xp-g2)--it's using much more wattage. This should get interesting pretty quick. 
I'm sure no one wants to carry around a 4" diameter light.
My recycling aperture design allows an SBT-70 to achieve the same emission angle of chip light as that of a smaller emitter (like xp-G2/xm-L2). I also am seeing higher benefit from recycling the light when the die is larger and circular like the SBT. With a square die, light scatters around more under the recycling aperture. Although a square image is "redrawn" back on the LED, a smaller LED is also now a smaller target to hit, meaning even a small misalignment of the collar ends up producing unwanted coma, more artifacts, as well as Kcd figures moving all around from light to light in higher percentages due to this effect.
I'm sure no one wants to carry around a 4" diameter light. My recycling aperture design allows an SBT-70 to achieve the same emission angle of chip light as that of a smaller emitter (like xp-G2/xm-L2). I also am seeing higher benefit from recycling the light when the die is larger and circular like the SBT. With a square die, light scatters around more under the recycling aperture. Although a square image is "redrawn" back on the LED, a smaller LED is also now a smaller target to hit, meaning even a small misalignment of the collar ends up producing unwanted coma, more artifacts, as well as Kcd figures moving all around from light to light in higher percentages due to this effect.
Last edited:
The_Driver
Flashlight Enthusiast
Have you seen this? (it's in German and you need to be registered to see full size pictures).
It's a de-domed XM-L2 with a very large aspheric lens (diameter = 208mm). It does 2.366.823 lux (3077m ANSI throw).
And that is without a Wavien collar and without a pre-collimator.
Here can find a beamshot comparison that includes it ("Optik Scheinwerfer XM-L2 de-domed (Photon Mod)") and a lot of other very powerful throwers including the Maxabeam in 50W mode (3 Million lux).
Note though that the "Optik Scheinwerfer" was apparently defocussed a bit (accidently) and only did around 2.5 million lux (they meareud all the lights on that day at a distance of 70m).
It's a de-domed XM-L2 with a very large aspheric lens (diameter = 208mm). It does 2.366.823 lux (3077m ANSI throw).
And that is without a Wavien collar and without a pre-collimator.
Here can find a beamshot comparison that includes it ("Optik Scheinwerfer XM-L2 de-domed (Photon Mod)") and a lot of other very powerful throwers including the Maxabeam in 50W mode (3 Million lux).
Note though that the "Optik Scheinwerfer" was apparently defocussed a bit (accidently) and only did around 2.5 million lux (they meareud all the lights on that day at a distance of 70m).
Last edited:
The_Driver,
I made an account and logged in, but the admin hasn't fully activated my account. I translated the site to english, it's just giving me a "you're not allowed here yet" message.
Is there anyway you can host photos of the build, or point to some which aren't blocked?
The beam he is making with the XM-L2 looks to be around 4500K, nice and warm. Truthfully, it looks like the beam I'm making with only a 75mm lens, but I can tell his spot is more energy dense with the lens he used. But, without a focal point known, just knowing a lens diameter has me curious about FL on that 208mm lens.
Also, The_Driver, what was the "Optik Scheinwerfer" war? Or are you just saying his light, the Photon Mod defocussed that day?
I made an account and logged in, but the admin hasn't fully activated my account. I translated the site to english, it's just giving me a "you're not allowed here yet" message.
Is there anyway you can host photos of the build, or point to some which aren't blocked?
The beam he is making with the XM-L2 looks to be around 4500K, nice and warm. Truthfully, it looks like the beam I'm making with only a 75mm lens, but I can tell his spot is more energy dense with the lens he used. But, without a focal point known, just knowing a lens diameter has me curious about FL on that 208mm lens.
Also, The_Driver, what was the "Optik Scheinwerfer" war? Or are you just saying his light, the Photon Mod defocussed that day?
Last edited:
The_Driver
Flashlight Enthusiast
The_Driver
Flashlight Enthusiast
Ok, I asked the owner's of the pictures for permission to post them here (Photon and tino79). Here they are:
The "Optik-Scheinwerfer":
Here is a picture from the front. That was lmost y year ago when the light still had an XM-L with dome on. It now has a de-domed Xm-L2 with a Wavien Collar.
Perfectly flat heatsink surface for the LEDs PCB:
The epoxy cured under light pressure to keep it as thin as possible:
Here a wall-shot of the spot:
I hope this helps you.
The "Optik-Scheinwerfer":
Here is a picture from the front. That was lmost y year ago when the light still had an XM-L with dome on. It now has a de-domed Xm-L2 with a Wavien Collar.
Perfectly flat heatsink surface for the LEDs PCB:
The epoxy cured under light pressure to keep it as thin as possible:
Here a wall-shot of the spot:
I hope this helps you.
Walterk
Enlightened
Inspiring projects.
Last edited:
Hey guys...I've been meaning to reply back here. Was staring at those photos the yesterday of the Photon mod.
First things first.
The_Driver, Thank You very much for those pictures! Truly, "candela porn" in its finest. Is that really a 208mm diameter lens? It looks smaller in diameter, judging by his hand--but photos can be misleading. One thing I know, that lens does not have a large spherical radius that I can see protruding--again though, maybe the pictures are misleading me. Looks like an F/# of 0.8 or so, trying to make out the curvature. Is it possibly a double convex lens?
I Just finished a whole day of optical testing with about 10 different lenses. I finally got the condenser system to work for me. That is, a negative followed by a positive lens. The equations above are the ones I had been following. But I was playing too close to the same absolute focal points to see any gains. This is where some confusion was occurring. I was using negative lenses that were too short on the negative, and losing light due to the cone being expanded so much. The Galilean equations state a shortening of focal length when a |-| and |+| lens are paired. Yet I was seeing an increase in focal length. This was because my focal lengths were so close to one another, the thickness of the lens itself was this difference offset, it was adding a few mm in BFL, confusing me. I tried a highly negative value lens, and I got the beam size to shrink in between the two lenses. This meant that the hotspot became bigger after the 2-lens system, vs using one lens. But the light was falling on a smaller area of the lens. This made me believe that maybe, just maybe, with a large enough diameter negative lens, I could collimate light to the second lens better than by using an aperture alone (the first problem I found fast was that negative lenses need roughly 2-3x the diameter initially "required"--or the diameter of light entering them x3 I would say, to keep from scattering that light everywhere within the lens). So far the negative+positive lens system is hit and miss for me, I'm not so sure it's worth the hassle or that there's any real Kcd gain when said and done. But if the light is contained within the needed path to the secondary lens (aspheric), it seems like more lux would be the end-result. Anyways, those are just some thoughts, more on this topic still needs to be understood by me. I have been working separately on a more interesting "Kcd project", the injection of more light into the phosphor layer itself, which is showing tremendous lux boosts in the amplified chip areas. But at the cost of total consumed input power.
Anyways, yeah, I'd love to buy that lens. I even have a host for it right now I can fit a 170mm OD lens into. With a lens that big, I don't think I would use an XM-L2, I think I would use the higher outputting round-emitter SBT-70. What are your thoughts on that idea?
My only real problem now is, the 10A CC driver. There were some problems with it. In short, it's not going to work for the specific intended build. Do you guys know who to talk to about custom drivers around here? If I was to fully drive the SBT-70, I would need all 10A, and then some. I have the cooling available, and I keep saying that as there are two different builds sort of going on, a large mod and a "giant" mod. Portable water-cooling is what I plan to attempt to use in the "giant" build, and if not portable water cooling (an intended computer CPU water cooling unit), I am going to use a forced fan induction copper CPU heat-sink. For the large mod I am focusing heavily on now, I need to find an 85-87mm OD aspheric, with a 55mm BFL--what my testing is telling me after going through a bunch of lenses ordered. Any ideas where to find an 85mm x 55mm FL aspheric? I was going to go down the custom road and have them ground for me, but I'm not so sure I want to dive into a custom lens run, until I can do some testing with an existing lens in that size range.
First things first.
The_Driver, Thank You very much for those pictures! Truly, "candela porn" in its finest.
Is that really a 208mm diameter lens? It looks smaller in diameter, judging by his hand--but photos can be misleading. One thing I know, that lens does not have a large spherical radius that I can see protruding--again though, maybe the pictures are misleading me. Looks like an F/# of 0.8 or so, trying to make out the curvature. Is it possibly a double convex lens?I Just finished a whole day of optical testing with about 10 different lenses. I finally got the condenser system to work for me. That is, a negative followed by a positive lens. The equations above are the ones I had been following. But I was playing too close to the same absolute focal points to see any gains. This is where some confusion was occurring. I was using negative lenses that were too short on the negative, and losing light due to the cone being expanded so much. The Galilean equations state a shortening of focal length when a |-| and |+| lens are paired. Yet I was seeing an increase in focal length. This was because my focal lengths were so close to one another, the thickness of the lens itself was this difference offset, it was adding a few mm in BFL, confusing me. I tried a highly negative value lens, and I got the beam size to shrink in between the two lenses. This meant that the hotspot became bigger after the 2-lens system, vs using one lens. But the light was falling on a smaller area of the lens. This made me believe that maybe, just maybe, with a large enough diameter negative lens, I could collimate light to the second lens better than by using an aperture alone (the first problem I found fast was that negative lenses need roughly 2-3x the diameter initially "required"--or the diameter of light entering them x3 I would say, to keep from scattering that light everywhere within the lens). So far the negative+positive lens system is hit and miss for me, I'm not so sure it's worth the hassle or that there's any real Kcd gain when said and done. But if the light is contained within the needed path to the secondary lens (aspheric), it seems like more lux would be the end-result. Anyways, those are just some thoughts, more on this topic still needs to be understood by me. I have been working separately on a more interesting "Kcd project", the injection of more light into the phosphor layer itself, which is showing tremendous lux boosts in the amplified chip areas. But at the cost of total consumed input power.
Anyways, yeah, I'd love to buy that lens. I even have a host for it right now I can fit a 170mm OD lens into. With a lens that big, I don't think I would use an XM-L2, I think I would use the higher outputting round-emitter SBT-70. What are your thoughts on that idea?
My only real problem now is, the 10A CC driver. There were some problems with it. In short, it's not going to work for the specific intended build. Do you guys know who to talk to about custom drivers around here? If I was to fully drive the SBT-70, I would need all 10A, and then some. I have the cooling available, and I keep saying that as there are two different builds sort of going on, a large mod and a "giant" mod. Portable water-cooling is what I plan to attempt to use in the "giant" build, and if not portable water cooling (an intended computer CPU water cooling unit), I am going to use a forced fan induction copper CPU heat-sink. For the large mod I am focusing heavily on now, I need to find an 85-87mm OD aspheric, with a 55mm BFL--what my testing is telling me after going through a bunch of lenses ordered. Any ideas where to find an 85mm x 55mm FL aspheric? I was going to go down the custom road and have them ground for me, but I'm not so sure I want to dive into a custom lens run, until I can do some testing with an existing lens in that size range.
The_Driver
Flashlight Enthusiast
The "Ampere!" Driver from the German company pcb components is a buck driver that has been tested up to 11A. Is is basically made for Luminus LEDs.
I will post more pictures of Photon's light when I have time in the evening. He sent me quite a few more.
I will post more pictures of Photon's light when I have time in the evening. He sent me quite a few more.
Last edited:
Sorry, been meaning to reply.
The_Driver, you are a lifesaver man! The Ampere is right! Thanks for that link. That is exactly what I needed to find. I'm trying to order a few of them to the USA to try out and test an SBT-70 at full power 37W. I wonder if the Ampere can support parallel operation for bigger Luminus chips...?
Borg,
I am not sure if you are asking the right question--or if I am understanding your question. Because it's a near paradox the way you word it below:
"...what effect on the beam will have a thicker lens compared to a thinner lens, assuming both lenses have the same diameter and the same focal point?"
If both lenses have the same diameter and the same focal point, that merely means that the sum of both their faces will bend light in a very similar way from the same incident ray angle striking them both. If the lens is "thicker", well then something else would have to change on the other side (plano side becomes concave, etc) for the focal point to be the same, if the diameter is the same. An exception is a lens like the Fresnel design, where that angular difference is replicated over and over by producing rings, which thus yields a thin lens (one also with poorer imaging quality, due to the many places where light refracts improperly). I don't believe this is the question you are really trying to ask, or answer, though.
The_Driver, you are a lifesaver man! The Ampere is right! Thanks for that link. That is exactly what I needed to find. I'm trying to order a few of them to the USA to try out and test an SBT-70 at full power 37W. I wonder if the Ampere can support parallel operation for bigger Luminus chips...?
Borg,
I am not sure if you are asking the right question--or if I am understanding your question. Because it's a near paradox the way you word it below:
"...what effect on the beam will have a thicker lens compared to a thinner lens, assuming both lenses have the same diameter and the same focal point?"
If both lenses have the same diameter and the same focal point, that merely means that the sum of both their faces will bend light in a very similar way from the same incident ray angle striking them both. If the lens is "thicker", well then something else would have to change on the other side (plano side becomes concave, etc) for the focal point to be the same, if the diameter is the same. An exception is a lens like the Fresnel design, where that angular difference is replicated over and over by producing rings, which thus yields a thin lens (one also with poorer imaging quality, due to the many places where light refracts improperly). I don't believe this is the question you are really trying to ask, or answer, though.
I needed to choose between these 2 lenses for the best throw : this one : http://www.aliexpress.com/store/pro...-led-lens-10W-100W-high/219000_876628128.html
and this one : http://www.aliexpress.com/store/pro...r-80-glass-led-lens-LED/219000_878205960.html
Inner diameter 71,1mm first lens, 72 mm the second one
Focal length 50mm first lens, 52 mm the second one
Thickness 30,5mm first lens, 38,7mm the second one
In the specifications there is a difference of 8mm in thickness and both lenses have almost the same focal length and interior diameter. Maybe is a error in the specifications.
I already ordered the 38,7mm thick one.
and this one : http://www.aliexpress.com/store/pro...r-80-glass-led-lens-LED/219000_878205960.html
Inner diameter 71,1mm first lens, 72 mm the second one
Focal length 50mm first lens, 52 mm the second one
Thickness 30,5mm first lens, 38,7mm the second one
In the specifications there is a difference of 8mm in thickness and both lenses have almost the same focal length and interior diameter. Maybe is a error in the specifications.
I already ordered the 38,7mm thick one.
Order them both and find out? My guess is, they measured EFL/BFL differently, or, the specs are in fact wrong. I've noticed translations of products that sure aren't great ones to the point it's gotten out of hand. Some probably done on purpose, like every older Chinese "XM-L T6" light suddenly became an "XM-L U2" labeled light when XM-L2s were released--what a coincidence.
Anyways, I ordered another lens that looked intriguing to mess with. It showed up finally--the 66mm concave-convex. (I've been ordering a LOT of lenses just to see the different product and MFGR abilities.) Well, it's a heck of an aspheric dome, one that doesn't focus worth a nickle. The "concave" side looks like a wavy plano. I don't mean a misinterpreted slight-concave, I mean a surface that looks like they just poured the glass mixture in and polished enough to get it to look clear, the concave grind wasn't completed or something went wrong. Someone's first day on the job grinding/polishing maybe? :laughing: It is this lens, mount is exactly the same and all dimensions the same, but I can assure you, it is not a true concave on that side like they claim, or like it looks from the photo: http://www.fasttech.com/products/0/10007135/1583800-replacement-jgf-66-66mm-optical-glass-concave
I'm still waiting on the German hi-amp buck drivers. Can't wait to get them. :naughty:
Anyways, I ordered another lens that looked intriguing to mess with. It showed up finally--the 66mm concave-convex. (I've been ordering a LOT of lenses just to see the different product and MFGR abilities.) Well, it's a heck of an aspheric dome, one that doesn't focus worth a nickle. The "concave" side looks like a wavy plano. I don't mean a misinterpreted slight-concave, I mean a surface that looks like they just poured the glass mixture in and polished enough to get it to look clear, the concave grind wasn't completed or something went wrong. Someone's first day on the job grinding/polishing maybe? :laughing: It is this lens, mount is exactly the same and all dimensions the same, but I can assure you, it is not a true concave on that side like they claim, or like it looks from the photo: http://www.fasttech.com/products/0/10007135/1583800-replacement-jgf-66-66mm-optical-glass-concave
I'm still waiting on the German hi-amp buck drivers. Can't wait to get them. :naughty:
Does anyone know a member(s) at CPF known for having good light testing equipment credentials here that I might be able to send a custom light off to, to verify its output in Kcd, and to report his findings to the forum as an unbiased tester of my design? If you think you might know who that would be, please PM me if you might know or are that person. I need to establish valid results here with the forum. Thanks guys. This thread has been running on for months I know, you all have probably lost interest by now, but, our progress is pushing a new type of boundary. Testing is slow and meticulous, but we want to optimize the design, and get it to some members in the end.
I lumped my testing and thoughts into this one thread on some various projects, which I should have probably split up so it's more of an on-point thread. Sorry that it has become so bloated with different subjects.
I lumped my testing and thoughts into this one thread on some various projects, which I should have probably split up so it's more of an on-point thread. Sorry that it has become so bloated with different subjects.
I am subscribed to this thread and never lost interest for it.
I am also interested in a good design for the best aspheric thrower that can be squeezed in a classic flashlight form factor.
Do you have any "unofficial" test numbers?
Can you provide more details of the final design ?
I am also interested in a good design for the best aspheric thrower that can be squeezed in a classic flashlight form factor.
Do you have any "unofficial" test numbers?
Can you provide more details of the final design ?
So far the bodies (bezels) except 86mm convex lenses. That's if I make ring holders thin to hold them. It could accept from 78-85mm aspherics as well with a flange ring lens design. The focal length is long, can get 60-62mm roughly depending on height of heatsink. The body provides this by its length inside the head available. I have the hosts, but the host needs some better machining, and some custom parts for a design I could run some batches of. I am going to try to turn the sinks on my lathe, but the light has two design models I have nearly finalized. A thrower in one design, very low beam angle. Then, a very large internal copper sink design, higher wattage, and 12x XM-L2 as a variant in another form, with different heatsink internally, but same exterior design. One just throws a massive amount of lux out, one is capable of very high Kcd. It is not close to being final for production, though, so I don't want to pretend these are weeks away from appearing. The body is final, about it (like I said, design is too, just needs to be completed). The circuit boards are covered, the German drivers deliver over 100W each, so I'm good there on power delivery. To produce the huge wattage model, I need a bigger lathe chuck, I simply cannot turn the copper myself on my current lathe. So I have to have the heatsinks for that version made by a CNC shop/owner or someone with a bigger lathe than me. The thrower is still in a technical phase with the collar system, one that further amplifies light, more so than a bare wavien collar, focuses the light to a smaller image. The collar needs to be modified, and another type of optic needs to be centered and glued into the light collar to make it durable in a drop or bash to retain full alignment. There are some production steps, basically, that I still need to find the right machinist/shop for. But, I have the right host for the needs, and I have the right heatsink designs figured out. I just need to get batches of them machined for me to get the price down, and that will take some searching. They are solid copper, the driver attachment that is, and the internal LED sink. I'm using one primary light design to cover two completely different throwing needs. The aspheric is a tougher project to pull together, so it's been a lot of testing, and throwing ideas out, bringing new ones in, getting closer to a good design for real use. It's a long process.
I cannot give any unofficial numbers. I did some initial calculations, but it could vary so much depending on the final lens mfgr design, that I feel it would be false advertising right now. Kcd may fall shorter than expected, and then it would be a let down and require design change. The tests are showing that the light is going to reach some really high values, that is based on its sheer size. This may not be appealing to some, but the goal is runtime, not just a blast to dead batts. So, I chose 3x 32650 cells as the handle size. The Feilong cells pack a lot of energy in 3S, 12.6V configuration. The good thing is, the drivers do have a soft start of amperage, so immediate load is not placed on the switch click (also a beefy switch, in its own right).
I cannot give any unofficial numbers. I did some initial calculations, but it could vary so much depending on the final lens mfgr design, that I feel it would be false advertising right now. Kcd may fall shorter than expected, and then it would be a let down and require design change. The tests are showing that the light is going to reach some really high values, that is based on its sheer size. This may not be appealing to some, but the goal is runtime, not just a blast to dead batts. So, I chose 3x 32650 cells as the handle size. The Feilong cells pack a lot of energy in 3S, 12.6V configuration. The good thing is, the drivers do have a soft start of amperage, so immediate load is not placed on the switch click (also a beefy switch, in its own right).
Last edited:
lumen aeternum
Enlightened
- Joined
- Sep 29, 2012
- Messages
- 907
How do you apply current preferentially to certain areas of the chip?
What happens if you grind a depression in the back of an aspheric lens, deep enough so the LED chip is "inside" the lens? Maybe fill the empty space with heat conducting grease. The lens might need to be Pyrex to handle the heat on the small area touching the LED; maybe not.
electronFarmer
Newly Enlightened
I have seen a lot of reviews by this gentleman, all done in the same manner, same testing AND he has his own Lumen meter
http://www.candlepowerforums.com/vb/member.php?16087-selfbuilt, there are others. This sight has Google search built so "flash light reviews" entered there resulted in this page. Hope that helps.
http://www.candlepowerforums.com/vb/member.php?16087-selfbuilt, there are others. This sight has Google search built so "flash light reviews" entered there resulted in this page. Hope that helps.
