Someone posted about a new light from Electronic GoldMine, one of my favorite places to find unusual stuff. Here is a picture of it from their site:

There are some comments on another thread about how the light has almost no front end room for a decent collimator, but I foolishly go ahead and buy a couple anyway (and for what you get, $11.95 is actually not a bad deal) and I must say that the no-room comment is absolutely true, there is a chrome plated plastic shelf around which eight Nichias (okay maybe they're not even Nichias) sit, but they seem to only be a millimeter or two below the lens.

Undaunted, I put three fresh AAAs into the quite well built battery holder and they pop in very smoothly. When I activate it by clicking the tailswitch (a reverse momentary type) I am kind of surprised that it has quite a bright output, casting a good useful blue-white flood beam around the immediate area. Getting into the light later will reveal that these eight LEDs are actually being direct driven off the 3AAAs. But the output is very diffused and the light meter confirms this:

A hair over 41 lux from a meter away.

So I take it all apart:

The best feature is the third item from the left, the threaded aluminum retaining ring. It is useful since there is a small ledge all around the inside of the hole which will help expand some of that precious room at the top of the head once reversed and in combination with a sized heatsink that drops in from the top:

There are three holes because in addition to the threaded grounding screwhole I originally planned to pass a separate wire through to the SYAK-binned Lux3's neg tab. But in the end I simply tightened the screw directly onto the tab itself since some CW2400 was used to affix the sink to the ledge:

An IMS 20.5mm reflector was dropped down the head followed by the above shown "pill" built using the retaining ring, the whole assembly of which was threaded down the head with a small pair of needlenosed pliers, one tip in each hole, until the Luxeon's plastic base shoulder was just snug against the bottom of the 20.5mm reflector:

Despite inverting the retaining ring to create more space, the whole light when screwed back together was still an eighth of an inch or so longer than the original since the head would not screw in as far due to the relatively long reflector now being used, so the exposed thread gap was filled with a thick o-ring:

Time for some brightness testing. We will be putting this modified GoldMine light (or GM, as there is no name nor even documentation that came with the light) up against a light some of you may already know, the VIP. But this VIP is unlike any other, because this VIP has its Own 20.5mm IMS reflector as well:
And note how very elegantly mounted it is, too. Seriously though, the 20.5mm IMS reflector was too long (and a bit too wide) to fit inside the regular VIP head. But the reflector's end hole was sized so perfectly that it made for a very snug friction fit onto the Luxeon3's dome, allowing the light to be turned sideways or even pointed straight down if necessary.

With the VIP on Medium, both lights were pointed at the ceiling and angled about 20 feet up and away. The GM is on the right in all beamshots:

And note how very elegantly mounted it is, too. Seriously though, the 20.5mm IMS reflector was too long (and a bit too wide) to fit inside the regular VIP head. But the reflector's end hole was sized so perfectly that it made for a very snug friction fit onto the Luxeon3's dome, allowing the light to be turned sideways or even pointed straight down if necessary.

With the VIP on Medium, both lights were pointed at the ceiling and angled about 20 feet up and away. The GM is on the right in all beamshots:

Yes, it actually outperforms the VIP on Medium, even with the VIP using the same 20.5mm reflector. So the VIP was switched into High and another beamshot was taken:

Slightly different story now. But this just illustrates how bright the GM mod really is, especially with the 20.5mm reflector. And to put it into an objective format, here is the "after modification" lux reading:

Yup, over 900 lux. Wonder what the results would be if I had used a T-binned emitter...?