InTheWorks
Newly Enlightened
- Joined
- Dec 26, 2007
- Messages
- 8
This place has been the inspiration for my build. There's a lot of cooler stuff around here, but hopefully this provides a little entertainment.
Some pics:
Light Engine - Top View
Light Engine - Bottom View
Delrin Switch Holder
Light Engine, Spacer, Switch Holder - exploded view
Light Engine, Spacer, Switch Holder - unexploded view
Detail of Switch Holder electrical connections
Light Engine, Spacer, Switch Holder - assembled in host
The spacer was required because I wanted to be able to spot focus the Ostar with either the KD or MG lense. Having a simple spacer like that makes it easy to adjust the height for more flood/less spot given the limited range of threads on the mag head. With the crenellated bezel attached, there's a limit to how much flood can be used so it worked out quite well. The die is in focus with a few threads left on the head. With the head fully screwed down, the crenellated bezel just scallops the flood. I've become a real fan of focusable aspherics.
I don't have any pics of the complete light to post, but it's using one of nanomiser's bezel and an MG aspheric now. It's definately not a looker and is still unfinished, especially with that switch held against the light body with heatshrink. Nor do I have any beamshots at this time, but it looks like most other ostar based lights with the MG lense.
Also, I did not machine any of the parts. I found a local machinist to do the work according to my CAD models. As long as my wife doesn't find out what it cost me, I think I'll live to see old age...
The story:
I had a 6D maglite sitting around and I wanted to put it to good use. After discovering CPF, I got the spark of an idea to build an aspheric Mag. So I bought the KD drop-in and aspheric lense combo. While much brighter than the original krypton bulb, it was still lacking for a light of it's size.
I then saw a few builds using the Ostar LEDs and was intrigued (and yes I read all the stuff on how they weren't as bright or cost effective as other options). I expressed interest in an Ostar LED when one of our reps was in for a visit. As we were looking at using their LED offerings in one of our products the rep saw fit to sample me one of the 6 die Ostars. Now I just *had* to build a light around it. I just didn't have time.
Fast forward several months and the annual co-worker camping trip approaches. A bunch of us from work have done a weekend camping trip for a couple of years now. The year before I brought my lights: the KD ashperic mag and my trusty EDC (Liteflux LF2). A couple of others brought theirs and we had fun in the dark ;p One of them was a surefire something or other with an LED drop-in that heavily embarassed my anaemic aspheric mag.
This year there was an outright challenge. One of my co-workers (who knew I scored the Ostar) was building a "light canon" consisting of 3 dialight 15W LEDs. He was using them for something non-flashlight related, but was putting together a "portable" light with which to compete in the flashlight contest. The gauntlet was thrown, I would not be embarrased again!
I looked around at the various methods at installing and heatsinking LEDs around here and there wasn't a 'drop-in' solution. So I decided to design my own heatsink arrangement around the Ostar and a shark driver board. It was very important to me that this thing run off standard Alkalines which definately constrained the design.
It really is quite bright and beat the pants off of the surefire. The contest between the light canon and my mag did not have a real winner. The mag was definately more impressive as it did not require a backpack (which held an inverter+power supply) and was focusable from spot to flood. Of course, the light canon had wider and brighter flood, but it wasn't *that* much brighter considering the 45W it consumed vs the 20W Ostar (measures 19.9V@1A across the LED). This is probably because the dialights are not terribly efficient. The Ostar I was sampled is not a very good bin (blue tint and half the max light output) and I'm itching for brighter...
On the subject of batteries, the cheapest no-name batteries can't seem to light this up at full brightness. My guess is that their internal resistance is too high and the converter fails to start (well fails to remain started anyway). When the converter fails to start though, it leaves the batteries connected to the LED in a sort of direct drive state. Of course, the LED is severely undervolted and would not normally light up, but the converter was on long enough to forward bias the LED and it seems to run down to 9 volts.
That said I've strictly purchased Energizers and they are getting expensive... As there are only 6 cells and 9 volts they are pushing the converter's limits. I don't think I can use NiCd batteries with their cell voltages. Though I'd love to hear any recommendations. 9 volts is a nice input voltage because it will let the LED run down to a scant 10mA @ 10V or so. When I finish this light I'd like to have it adjust down to a very low "candle" mode. Right now it's either full brightness or off.
The batteries seem to last a reasonably long time. I had it running for upwards of 30 minutes once (long story, but the light this thing put out was a blessing) and with occassional use afterwards (it's hard not to play with this light indoors) it was still plenty bright. I only change the batteries for fear of running the voltage too low and damaging the converter...
As you can imagine, the flashlight gets pretty warm during operation. I haven't measured it, but during extended (>10min) use it gets just shy of uncomfortably hot to hold near the head. The tail of the light is much cooler, but still warm. The thermal connection of the heatsink to the flashlight body seems quite good. I designed it with the intention of having the heatsink expand to the point of touching and transferring heat to the flaslight body and head. Whether this is actually ocurring or not, I don't know. The heatsink doesn't feel any hotter than the flashlight body (up near the heatsink) when it's running so there's a good deal of heat transfer regardless.
Future plans include:
I'm thinking real hard about running a 40W dialight LED with the same heatsinking arrangement, but that is a tale for another thread
Some pics:
Light Engine - Top View
Light Engine - Bottom View
Delrin Switch Holder
Light Engine, Spacer, Switch Holder - exploded view
Light Engine, Spacer, Switch Holder - unexploded view
Detail of Switch Holder electrical connections
Light Engine, Spacer, Switch Holder - assembled in host
The spacer was required because I wanted to be able to spot focus the Ostar with either the KD or MG lense. Having a simple spacer like that makes it easy to adjust the height for more flood/less spot given the limited range of threads on the mag head. With the crenellated bezel attached, there's a limit to how much flood can be used so it worked out quite well. The die is in focus with a few threads left on the head. With the head fully screwed down, the crenellated bezel just scallops the flood. I've become a real fan of focusable aspherics.
I don't have any pics of the complete light to post, but it's using one of nanomiser's bezel and an MG aspheric now. It's definately not a looker and is still unfinished, especially with that switch held against the light body with heatshrink. Nor do I have any beamshots at this time, but it looks like most other ostar based lights with the MG lense.
Also, I did not machine any of the parts. I found a local machinist to do the work according to my CAD models. As long as my wife doesn't find out what it cost me, I think I'll live to see old age...
The story:
I had a 6D maglite sitting around and I wanted to put it to good use. After discovering CPF, I got the spark of an idea to build an aspheric Mag. So I bought the KD drop-in and aspheric lense combo. While much brighter than the original krypton bulb, it was still lacking for a light of it's size.
I then saw a few builds using the Ostar LEDs and was intrigued (and yes I read all the stuff on how they weren't as bright or cost effective as other options). I expressed interest in an Ostar LED when one of our reps was in for a visit. As we were looking at using their LED offerings in one of our products the rep saw fit to sample me one of the 6 die Ostars. Now I just *had* to build a light around it. I just didn't have time.
Fast forward several months and the annual co-worker camping trip approaches. A bunch of us from work have done a weekend camping trip for a couple of years now. The year before I brought my lights: the KD ashperic mag and my trusty EDC (Liteflux LF2). A couple of others brought theirs and we had fun in the dark ;p One of them was a surefire something or other with an LED drop-in that heavily embarassed my anaemic aspheric mag.
This year there was an outright challenge. One of my co-workers (who knew I scored the Ostar) was building a "light canon" consisting of 3 dialight 15W LEDs. He was using them for something non-flashlight related, but was putting together a "portable" light with which to compete in the flashlight contest. The gauntlet was thrown, I would not be embarrased again!
I looked around at the various methods at installing and heatsinking LEDs around here and there wasn't a 'drop-in' solution. So I decided to design my own heatsink arrangement around the Ostar and a shark driver board. It was very important to me that this thing run off standard Alkalines which definately constrained the design.
It really is quite bright and beat the pants off of the surefire. The contest between the light canon and my mag did not have a real winner. The mag was definately more impressive as it did not require a backpack (which held an inverter+power supply) and was focusable from spot to flood. Of course, the light canon had wider and brighter flood, but it wasn't *that* much brighter considering the 45W it consumed vs the 20W Ostar (measures 19.9V@1A across the LED). This is probably because the dialights are not terribly efficient. The Ostar I was sampled is not a very good bin (blue tint and half the max light output) and I'm itching for brighter...
On the subject of batteries, the cheapest no-name batteries can't seem to light this up at full brightness. My guess is that their internal resistance is too high and the converter fails to start (well fails to remain started anyway). When the converter fails to start though, it leaves the batteries connected to the LED in a sort of direct drive state. Of course, the LED is severely undervolted and would not normally light up, but the converter was on long enough to forward bias the LED and it seems to run down to 9 volts.
That said I've strictly purchased Energizers and they are getting expensive... As there are only 6 cells and 9 volts they are pushing the converter's limits. I don't think I can use NiCd batteries with their cell voltages. Though I'd love to hear any recommendations. 9 volts is a nice input voltage because it will let the LED run down to a scant 10mA @ 10V or so. When I finish this light I'd like to have it adjust down to a very low "candle" mode. Right now it's either full brightness or off.
The batteries seem to last a reasonably long time. I had it running for upwards of 30 minutes once (long story, but the light this thing put out was a blessing) and with occassional use afterwards (it's hard not to play with this light indoors) it was still plenty bright. I only change the batteries for fear of running the voltage too low and damaging the converter...
As you can imagine, the flashlight gets pretty warm during operation. I haven't measured it, but during extended (>10min) use it gets just shy of uncomfortably hot to hold near the head. The tail of the light is much cooler, but still warm. The thermal connection of the heatsink to the flashlight body seems quite good. I designed it with the intention of having the heatsink expand to the point of touching and transferring heat to the flaslight body and head. Whether this is actually ocurring or not, I don't know. The heatsink doesn't feel any hotter than the flashlight body (up near the heatsink) when it's running so there's a good deal of heat transfer regardless.
Future plans include:
- a more elegant switch assembly with "modes"
- upgrading the Ostar to a brighter bin, or some other multi-die package.
I'm thinking real hard about running a 40W dialight LED with the same heatsinking arrangement, but that is a tale for another thread
