rhd
0
- Joined
- Dec 7, 2010
- Messages
- 8,475
- Points
- 0
The "Sun Shiner" (a triple MT-G2 build that pushes the definition of "practical")
The "Sun Shiner"
I wanted a ridiculously, impressively, and nearly uselessly bright flashlight. I didn't need something practical. I have lots of practical flashlights. You can buy practical flashlights at Costco. I wanted something impressive, and the 9x, 12x, and 15x CREE XML flashlights I've purchased online just didn't do it. So I decided to make something better
It started out as this - a 3x18650 flashlight from KaiDomain that ran 3x XML emitters and cost a not-so-cool $45. It was arguably ugly, and used a momentary switch (tied to a junky driver) instead of a proper toggle switch. When it arrived almost entirely non functional, I didn't even care... because I only wanted its body
With the addition of a proper toggle switch, and some dedication to stripping and polishing, the ugly black host, became this:
Of course, it's not just about whats on the outside, the guts had to go too. I pulled the 3x XML LEDs and replaced them with 3x MT-G2 emitters. These beasts can handle 5A at 7V each, without feeling much pain (and I'm sure others have driven them substantially harder).
I had decided to run them at *only* 3A instead of 5A, but for those keeping track, this still meant over 60W of input power. It was clearthat supplying enough current to these LEDs would require a snappy driver setup, and some good quality cells. On the driver front, I pulled from the the laser realm, where we've done a lot more to miniaturize driver tech than they have in the flashlight community. I decided to use 3x of my own buck drivers, each set to supply 3A of current. The drivers are supplied in parallel from a set of 3x 18650 cells in series. I'm using Panasonic 18650s (but not NEW Panasonic 18650s - those are still in the mail).
It was a very tight fit. I don't know why I bothered heatsinking the driver ICs in retrospect, because the build generates enough heat from the LEDs to justify a shutdown long before IC heat will be an issue.
The obvious question is "how bright is it???"
The following photos don't do it justice. I wanted something that would be brighter than my 12x and 15x XML lights, and I certainly accomplished that. From the photos, the Sun Shiners looks "a little bit brighter" than the 12x and 15x lights. In person/reality, it looks twice or three times as bright to my eye (not to mention the fact that colour is MUCH more natural).
I have some of the new 10A Panasonic cells in the mail. By my math, 3x of them in series, at 11.1V nominal, would just about be capable of supplying enough input power for buck drivers to supply 15A of output to these LEDs at a Vf of ~7V. Now that I have a working 5A buck, I'm tempted to swap out the 3A bucks in here now, and try the higher current, fed by the newer Panasonic cells. That said, the current setup is already far beyond the realm of being practical, and it's hard to imagine this getting any brighter.... so I might just leave it be
The "Sun Shiner"
I wanted a ridiculously, impressively, and nearly uselessly bright flashlight. I didn't need something practical. I have lots of practical flashlights. You can buy practical flashlights at Costco. I wanted something impressive, and the 9x, 12x, and 15x CREE XML flashlights I've purchased online just didn't do it. So I decided to make something better
It started out as this - a 3x18650 flashlight from KaiDomain that ran 3x XML emitters and cost a not-so-cool $45. It was arguably ugly, and used a momentary switch (tied to a junky driver) instead of a proper toggle switch. When it arrived almost entirely non functional, I didn't even care... because I only wanted its body
With the addition of a proper toggle switch, and some dedication to stripping and polishing, the ugly black host, became this:
Of course, it's not just about whats on the outside, the guts had to go too. I pulled the 3x XML LEDs and replaced them with 3x MT-G2 emitters. These beasts can handle 5A at 7V each, without feeling much pain (and I'm sure others have driven them substantially harder).
I had decided to run them at *only* 3A instead of 5A, but for those keeping track, this still meant over 60W of input power. It was clearthat supplying enough current to these LEDs would require a snappy driver setup, and some good quality cells. On the driver front, I pulled from the the laser realm, where we've done a lot more to miniaturize driver tech than they have in the flashlight community. I decided to use 3x of my own buck drivers, each set to supply 3A of current. The drivers are supplied in parallel from a set of 3x 18650 cells in series. I'm using Panasonic 18650s (but not NEW Panasonic 18650s - those are still in the mail).
It was a very tight fit. I don't know why I bothered heatsinking the driver ICs in retrospect, because the build generates enough heat from the LEDs to justify a shutdown long before IC heat will be an issue.
The obvious question is "how bright is it???"
The following photos don't do it justice. I wanted something that would be brighter than my 12x and 15x XML lights, and I certainly accomplished that. From the photos, the Sun Shiners looks "a little bit brighter" than the 12x and 15x lights. In person/reality, it looks twice or three times as bright to my eye (not to mention the fact that colour is MUCH more natural).
I have some of the new 10A Panasonic cells in the mail. By my math, 3x of them in series, at 11.1V nominal, would just about be capable of supplying enough input power for buck drivers to supply 15A of output to these LEDs at a Vf of ~7V. Now that I have a working 5A buck, I'm tempted to swap out the 3A bucks in here now, and try the higher current, fed by the newer Panasonic cells. That said, the current setup is already far beyond the realm of being practical, and it's hard to imagine this getting any brighter.... so I might just leave it be
Attachments
Last edited:
