After whetting my appetite with a repurposed Darcy Metal Gear build (that pulls so much from 3 AAA's that I had to track down some NiCD's to bear the brunt of it), I started looking at options for a 445 build. While Jayrob, Flaminpyro and others showed us plenty of host options, I decided that I would examine the prospects of repurposing a PGL-III-C greenie with a dead driver I have sitting around.
Here is the host.
It shows the scars of a retaining ring that seized up in what I presume was excess paint or anodizing at the mouth, that I eventually extracted by cutting it up with a Dremel. Suffice it to say that I learned a lot about how to disassemble things properly from this little struggle. In any case, it nevertheless held promise for my goal, which was to build a 445 with some staying power.
Side goals were to fully preserve the original FDA safety mechanisms of the host, which include a key interlock, shutter and butt plug (whatever that thing is called
My first step was to call in reinforcements, in the form of a Jayrob custom heatsink, machined to match the dimensions of the main 532nm module and set up with a Microboost maxed out with Jayrob's "Hot" option. The original module is the one with the brass section.
Since there was a quite large volume of metal, Jay and I elected to simply mount the driver at the back end of the sink and drill a channel through the center to the top where a standard Aixiz module would sit.
Stage 1: assembling the sink and diode. I installed my waiting module with 445nm diode into the sink, wired it up to the driver, checked connections nd hooked it up to my bench supply which has variable maximums for voltage and current. With a 3.9V max, I elected to vary the current for testing.
With everything set up, I turned the current up to a few hundred mA, just enough to get it lasing, and -- flicker-flicker BRIGHT dead.
I yanked the current back down. Dead already? Checked everything, turned it back up. A sickly purple glow told me everything I needed to know -- the diode was fried. Aw, dammit. Time to log into LPF and see if daguin's got any more of these lef----
wait a minute. Purple? I applied power again. That's not 445, it looks like 405! I realized that I'd just shoved about 1A through a spare PHR .
I thought I'd already swapped in the new diode.
So I just needed to fetch the 445nm and install it in the module. But where was the blasted thing? I had it in my head for weeks that it was in the module! OMG, what happened to it, did I throw out the little envelope Dave had sent it in? Oh No, WHERE IN THE NAME OF DAGUIN IS THAT 445NM -----
oh there it is.
Carefully install it in the module, and once again set things up in the sink. Double-check all the connections, discharge the driver caps, solder. Apply power, and --
... all is right (and rich blue) in the world!
The first design assumption had not passed muster. My Darcy setup, which likely tops out around 300mW with freshly charged batteries, seems hardly to warm up the heatsink, which is roughly a 1" by 1/2" cylinder. I had it in my head that something over 2" long would be able to soak up that heat for quite a while.
Not so. Run at full power, even this sink became too hot to hold after about a minute(3.8V @ 1.5A into the Microboost). The host would help bleed it off, but even so I was still blown away by how much power was involved here.
As I said to Jay, I'm going to need a bigger boat -- er, set of goggles. I have a pair, but they are out of their league here. Anything caught in the beam within five feet was set to smoking in seconds. More than once my wife asked me what was burning as I labored away in the lab
This test target started smoking before I even brought it to full power -- and burned right through before I could snap these shots.
All that remained by this point was installing it into the host.
I elected to retain the original main switch rather than try to match something else in. After an abortive attempt at trying to epoxy it onto the plastic mounting bracket in its original position, I elected to strip the components off the dead driver board and remount it there, which had the advantage of being the original mounting setup.
Here's the assembly, showing the original plastic mounting hardware.
[/URL]
See that stranded wire splayed around the side? That's the first huge flaw in the setup as it now stands. Positive power was provided for in the original setup, but drew the negative power off the case via the aluminum cap (seen in the pic with the blue pads). Here my limited knowledge brought me to an impasse: how do I attach the driver wire to the case? Nothing but aluminum as far as the eye can see -- solder won't stick. I ended up attaching the negative wire to the aluminum cap (minus the blue pads in the pic) with some electrical tape, and bolstering the connection with some foam compressed between the switch board and the new sink.
As this needed to have some steady pressure applied to that electrical tape to ensure a decent connection, I had to retain the sink with pressure. And here we come to the second major flaw in this build -- no retention solution. With the host threads unusable, I had no immediately available method of retaining the sink other than this hack:
Yes, that's a length of wire wedged into the threads.
And if that weren't enough, once I'd addressed all the glitches that were my fault, the host decided to throw up one of its own: the "butt plug" began malfunctioning. After some irritated searching for potential opens around the driver, I finally tracked it down to the host, and decided to sacrifice the "butt plug" safety with a soldered short.
Once all that was done, however, the beast roared to life. Here are two shots comparing it with the 300mW Darcy build with freshly charged batteries.
The wooden bedframe which was the target for the second shot was smellably smoking after scant seconds, from four feet away.
Here are some final build notes:
An unprotected LG Chem 18650 cell is able to handle the average 1.5A pulled by the Microboost. This current pull gets up to around 2A before the driver starts flashing, and then drops to half power. Don't be running it by then, that's 2A at 2.2V and will kill your unprotected 18650.
I found out that feeding the Microboost at 2.2V (from my bench supply) allowed me to turn the current as low as 10mA, taking it down to Class IIIA levels, and it still lases.
A big thanks to jayrob, who took the time to include some notes and instructions with the sink, in particular the polarity -- which is more than I could say for CNI, hence the availability of this host
Finally, I would like to Ask The Experts some advice on solving these remaining issues with this build:
1. How to reliably wire up the driver negative to aluminum (specifically that aluminum cap)? I've heard that aluminum is solderable, but it's beyond my limited skills. I don't recall how it was done with the original driver; I suspect a direct connection, since the driver sits in it. There is a hole in the cap, I've thought of using a screw for a tight mechanical connection, or a clip.
2. Ideas for a retention solution? I originally planned a set screw, but I don't know where such screws can be had (Home Depot?) and I think the housing may be too thin unless I can tap a very small hole. (By way of note: the set screw Jayrob used in the sink, fitting a 1.5mm allen wrench, is probably too big). Most of the outermost threads at the host mouth are destroyed and unusable.
With solutions to those issues in place, this will be a finished build.
Here is the host.
It shows the scars of a retaining ring that seized up in what I presume was excess paint or anodizing at the mouth, that I eventually extracted by cutting it up with a Dremel. Suffice it to say that I learned a lot about how to disassemble things properly from this little struggle. In any case, it nevertheless held promise for my goal, which was to build a 445 with some staying power.
Side goals were to fully preserve the original FDA safety mechanisms of the host, which include a key interlock, shutter and butt plug (whatever that thing is called
My first step was to call in reinforcements, in the form of a Jayrob custom heatsink, machined to match the dimensions of the main 532nm module and set up with a Microboost maxed out with Jayrob's "Hot" option. The original module is the one with the brass section.
Since there was a quite large volume of metal, Jay and I elected to simply mount the driver at the back end of the sink and drill a channel through the center to the top where a standard Aixiz module would sit.
Stage 1: assembling the sink and diode. I installed my waiting module with 445nm diode into the sink, wired it up to the driver, checked connections nd hooked it up to my bench supply which has variable maximums for voltage and current. With a 3.9V max, I elected to vary the current for testing.
With everything set up, I turned the current up to a few hundred mA, just enough to get it lasing, and -- flicker-flicker BRIGHT dead.
I yanked the current back down. Dead already? Checked everything, turned it back up. A sickly purple glow told me everything I needed to know -- the diode was fried. Aw, dammit. Time to log into LPF and see if daguin's got any more of these lef----
wait a minute. Purple? I applied power again. That's not 445, it looks like 405! I realized that I'd just shoved about 1A through a spare PHR .
I thought I'd already swapped in the new diode.So I just needed to fetch the 445nm and install it in the module. But where was the blasted thing? I had it in my head for weeks that it was in the module! OMG, what happened to it, did I throw out the little envelope Dave had sent it in? Oh No, WHERE IN THE NAME OF DAGUIN IS THAT 445NM -----
oh there it is.
Carefully install it in the module, and once again set things up in the sink. Double-check all the connections, discharge the driver caps, solder. Apply power, and --
... all is right (and rich blue) in the world!
The first design assumption had not passed muster. My Darcy setup, which likely tops out around 300mW with freshly charged batteries, seems hardly to warm up the heatsink, which is roughly a 1" by 1/2" cylinder. I had it in my head that something over 2" long would be able to soak up that heat for quite a while.
Not so. Run at full power, even this sink became too hot to hold after about a minute(3.8V @ 1.5A into the Microboost). The host would help bleed it off, but even so I was still blown away by how much power was involved here.
As I said to Jay, I'm going to need a bigger boat -- er, set of goggles. I have a pair, but they are out of their league here. Anything caught in the beam within five feet was set to smoking in seconds. More than once my wife asked me what was burning as I labored away in the lab
This test target started smoking before I even brought it to full power -- and burned right through before I could snap these shots.
All that remained by this point was installing it into the host.
I elected to retain the original main switch rather than try to match something else in. After an abortive attempt at trying to epoxy it onto the plastic mounting bracket in its original position, I elected to strip the components off the dead driver board and remount it there, which had the advantage of being the original mounting setup.
Here's the assembly, showing the original plastic mounting hardware.
See that stranded wire splayed around the side? That's the first huge flaw in the setup as it now stands. Positive power was provided for in the original setup, but drew the negative power off the case via the aluminum cap (seen in the pic with the blue pads). Here my limited knowledge brought me to an impasse: how do I attach the driver wire to the case? Nothing but aluminum as far as the eye can see -- solder won't stick. I ended up attaching the negative wire to the aluminum cap (minus the blue pads in the pic) with some electrical tape, and bolstering the connection with some foam compressed between the switch board and the new sink.
As this needed to have some steady pressure applied to that electrical tape to ensure a decent connection, I had to retain the sink with pressure. And here we come to the second major flaw in this build -- no retention solution. With the host threads unusable, I had no immediately available method of retaining the sink other than this hack:
Yes, that's a length of wire wedged into the threads.
And if that weren't enough, once I'd addressed all the glitches that were my fault, the host decided to throw up one of its own: the "butt plug" began malfunctioning. After some irritated searching for potential opens around the driver, I finally tracked it down to the host, and decided to sacrifice the "butt plug" safety with a soldered short.
Once all that was done, however, the beast roared to life. Here are two shots comparing it with the 300mW Darcy build with freshly charged batteries.
The wooden bedframe which was the target for the second shot was smellably smoking after scant seconds, from four feet away.
Here are some final build notes:
An unprotected LG Chem 18650 cell is able to handle the average 1.5A pulled by the Microboost. This current pull gets up to around 2A before the driver starts flashing, and then drops to half power. Don't be running it by then, that's 2A at 2.2V and will kill your unprotected 18650.
I found out that feeding the Microboost at 2.2V (from my bench supply) allowed me to turn the current as low as 10mA, taking it down to Class IIIA levels, and it still lases.
A big thanks to jayrob, who took the time to include some notes and instructions with the sink, in particular the polarity -- which is more than I could say for CNI, hence the availability of this host
Finally, I would like to Ask The Experts some advice on solving these remaining issues with this build:
1. How to reliably wire up the driver negative to aluminum (specifically that aluminum cap)? I've heard that aluminum is solderable, but it's beyond my limited skills. I don't recall how it was done with the original driver; I suspect a direct connection, since the driver sits in it. There is a hole in the cap, I've thought of using a screw for a tight mechanical connection, or a clip.
2. Ideas for a retention solution? I originally planned a set screw, but I don't know where such screws can be had (Home Depot?) and I think the housing may be too thin unless I can tap a very small hole. (By way of note: the set screw Jayrob used in the sink, fitting a 1.5mm allen wrench, is probably too big). Most of the outermost threads at the host mouth are destroyed and unusable.
With solutions to those issues in place, this will be a finished build.
