Supbeam K50 V2 Mod Thread

Thanks.

Yupp. This shows how little point it is to push it further from a real life point of view. Still fun trying. Due to the battery carrier issues im planning on going back to stock.

I verifed direct thermal path in post 158. It was a long post, dont blame you for not not noticing. Maybe I should have said K50, but the mcpcb said K40 as seen in pictures.

"Also reflowed a junk XM-L2 on to the stock mcpcb for testing. Verified that the stock K40 mcpcb does have direct thermal path."



Ryan, your measurements match mine nicely (Ive never trusted the super low amp numbers on my DMM though) Did you measure what the actual current is on high after power deacrease? I forgot that bit and did not see you mention it either.


Since many seems to be into the photos. Here larger versions of the K50 driver images I have shared.

Top side photo

Spring side photo

Thanks for the reply’s. Yes, its always fun pushing anything performance related to find a bit more, and I always enjoy reading about all the grand efforts to crank it up. Its nice to know that I dont have to go gouging a trough in the mcpcb for direct thermal. With the stock light, I wonder how much total resistance might be accumulated with all those springs and if they begin to heat up and waste energy (sustained high mode use). Thoughts on that and would it really warrant a wire mod?

During my walk into high terrain (no ambient light), I would focus on a steep grassy hill on high setting and walk the beam up the hill until it was barley perceptible. Then I would lower to the second highest mode and walk the beam back down the hill until the beam was again barley perceptible. Id guess the difference was around 110 feet of range. It just makes me smile… again: 2.7A vs 5A, a lot less heat and better run time. The more I use this light, the more I like it.

A (hopefully) far less complex issue compared to yall’s driver problems… if yours happens to have a slightly off center emitter is it an easy fix in a k50v2?

I haven’t broken mine down so this is the best shot I could grab of the misalignment.

As it is, it loads up one side of the corona heavily, and presumably decreases kcd by about ~25% (based on Dale’s numbers compared to mine).

The springs are probably just seeing around 1,5A and uptowards of 2A when stock.

In mode 5 its considerably less than 1amp through each spring.. On max output, the power decease also lowers the already fairly low amps. Many (premium lights) push twice the amount of amps through fairly similar springs. I would not worry about it. I would recommend to stick with protected cells in this light though.

Unscrew the bezel take out the reflector and see what you can do. You can see pictures of how I and Tom E did it earlier in this thread.

I would not compare kcd with others and draw conclusions out of it unless you are sure how your numbers compares. Dales numbers can be more than 30% higher than a low reading meter.

Then it’s 30% lower.

The above are at 97 yds to the red oil drum. Below, 610 yds to the white barn

My TK61, by comparison

They cut, raked and baled the field today. Other than that, same pic…same place same camera settings.

Is the TK61 bone stock?

It thread off at the control ring, above it I should say. Mine is off, and I’m about to hot de-dome it.

2 big strap wrenches and a persistent nature. :slight_smile:

With the stock light de-domed, I’m getting 420.5Kcd. Now THAT is what I’m talking about! My MAXToch SN6X-2X was king of the hill last month at 365Kcd. Now it suddenly finds itself in the crowd, watching the new podium holders.

SR90 at 622Kcd, TK61 at 600Kcd and the K50V2 at 420.5Kcd. WooHoo!

sorry heggood, didn’t realize that was directed at me…no, the TK61 is anything but bone stock. Carrier has had spring bypasses, running 2 tubes for run time, all springs bypassed. New driver in the head running 7.2A, XP-L2 on a 32mm Noctigon, de-domed, making 600Kcd. The reflector base has been cut out, such that it fits completely over a 20mm Noctigon. This helps it clear everything on the star, but now that it’s got the 32mm Noctigon it’s kind of redundant.

My old S1100 with a dedomed XML2 (only a T6, and it started as a 1C, so it's pretty ugly) kicks the stock K50's ass up one side and down the other. I think the K50's focus is pretty badly off, it only shows the 'donut hole' at a distance of 3-4 inches, the S1100 is showing it out to around a foot away from the wall.

This de-domed K50 shows the donut at about 6 inches from the wall. My TK61 it’s out around 6’. Not quite that far on the SR90.

?? I'm so confused...

  • I thought donut holes were a bad thing? You would want the donut hole to be at the most minimum length? I thought?
  • not sure how comparing a mod'ed de-domed S1100 to a stock light means much? I got a mod'ed Jacob A60 and a mod'ed T20 that out throw a stock TN31, but doesn't mean the A60 and T20 are better throwers or better lights.

Tom, I’m finding that the donut hole is specific to each reflector and the amount of throw it’s capable of producing. Maximizing throw tends to take the focus further out, leaving the donut hole considerably further out…or at least in a massive reflector like the TK61 or SR90. That’s one of the things I was having trouble with on the focus with the TK61, getting the beam convergence as far out as possible so it wasn’t crossing over and enlarging out at distance.

The reflector in my Courui gives a tubular looking beam, doesn’t seem to have that convergence at all.

If you have some space where you can shine the beam out to inifinity look at the beam from the light out, you should see a darker “V” in the middle out to 3’ or a bit further on the big throwers. That should be about the distance the donut hole disappears. Unless I’m missing something completely, and that could very well be…and not for the first time.

That “v” is quite evident in this picture, with the TK61 shining across the Mississippi River…

A massive carnival searchlight, notice the men standing beside it…. this is the kind of “into infinity” I was talking about. Check out how far the “v” goes out in the center of this huge beam!

The opposite of using a magnifying glass to burn a bug. There’s a fine point that the light is in convergence, then it spreads back out on the other side of that point.

You can see where the effect comes from with a piece of paper laid over the front...

Roche F6 XML2 10* TIR:

Roche F6 XPG2 stock OP reflector:

XinTD C8 XML2:

edit: The XinTD probably needs the LED moved farther up, you can see it's not using very much of the reflector. It's got a standard ~.035" thick centering ring, where the pre-built XinTDs use the much thinner adhesive insulator.

I’ve been looking at the formula for figuring the parabolic curve, and it would seem that this “v” shape we’re seeing is due to an imperfect parabola. This is probably also due to the fact that an LED does not have a point light source so the reflector is having to gather light from a broader source and collimate it into the cone we desire. Effectively though, the differences we’re seeing are probably the result of the success rate of different manufacturers at getting the curve right.

Dunno, I'm having a tough time seeing the V - I see convergence, I don't see divergence on the other side. I know with lights like the Shocker and HD2010, The kcd #'s would go up as that donut's distance got less. Maybe you are right about the big reflectors. I recalll rdrfronty testing lux at many distances and saw no difference from 15 m on out, but that could be the specific lights he was testing at the time. I don't think I've seen any one else test lux (kcd) at significant distances - takes 2 people of course and dark open space.