Supbeam K50 V2 Mod Thread

308 posts / 0 new
Last post

Pages

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US
Supbeam K50 V2 Mod Thread

Hey guys, this is the mod thread for the Supbeam K50 V2 from the recent (and fast paced) group buy. Thanks a lot Justin. Wink

I’ll definitely me modding my light, dedome, current boost, PCB lapping, copper braiding, etc. and I’ll post the progress below.

If you guys do your own mods, please feel free to share the process and steps. Photos and measurements are always welcome too.

I’ll made an index on the OP, so that the mods by each member are easily accessible.

 

MemberInfoPicsDedomeEmitter CurrentPCB LappingThicker WiresCopper BraidingOutput (lm)Throw (kcd)
ryansoh3Post #1

Unboxing #82

Gasoline

Stock (5.0A)

Measurements #206

     
Tom EPost #101

General #101

Driver #109

 Stock (5.0A)   1,656208
RaceR86

Post #104

Driver #158

General #104 

8.0A (RIP) #129

6.9A → 5.6A #158

   

1722 (start)

1638 #158

 206 #158

 

General Information:

The following information is from Vinh's K50 thread. It has helpful information on modding the light and is definitely recommended. 

I'll try to distill the load of information. 

Driver tuning:

Current can drain up to 9A if desired. I will likely set current less than 9A...Will see which current tune the best brightness.

Reflector focus tuning:

I tried getting the led closer in which usually work well but in this case the led has to be moved down away from the reflector a bit. The reflector is so big that a tiny horizontal drift could throw the beam out of focus. 

BLF ≠ B-grade Flashlight Forum

 

Edited by: ryansoh3 on 08/01/2014 - 10:53
ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US

This post is intentionally left blank.

(Sorry comfy.)

BLF ≠ B-grade Flashlight Forum

 

zeremefico
zeremefico's picture
Offline
Last seen: 7 hours 1 min ago
Joined: 03/27/2012 - 02:44
Posts: 1390
Location: Greece
Love

₪₪₪₪ ΟΥΔΕΝ ΚΡΥΠΤΟΝ ΥΠΟ ΤΟΝ ΗΛΙΟ ₪₪₪₪

My YouTube channel

Flashlights & edc gear

K40M F16

aoeu
aoeu's picture
Offline
Last seen: 9 months 2 weeks ago
Joined: 02/16/2014 - 04:44
Posts: 793
Location: Sydney, Australia

How did you find the dedome? Vinh said after he dedomed his he had to adjust the focus somehow because it was out after that.

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US
aoeu wrote:
How did you find the dedome? Vinh said after he dedomed his he had to adjust the focus somehow because it was out after that.

I haven’t done it on the K50 V2 yet, but I have dedomed my Supbeam K40 and ThruNite TN31.

I think dedoming does throw the focus out a little bit, but I didn’t mind the beam profile in my lights so all I did was some reflector tuning (with my tongue at the right angle) so that the “flower” around the hotspot was evenly spaced.

Others shave a couple of mm off the base of the reflector so that the reflector will be deeper in.

The review SelfBuilt breifly touches on what Vinh does on his modded lights about focusing the reflector.

BLF ≠ B-grade Flashlight Forum

 

RaceR86
RaceR86's picture
Offline
Last seen: 5 years 9 months ago
Joined: 02/22/2012 - 08:32
Posts: 3777
Location: Norway

Can you sort the list from highest to lowest emitter amps??  J)

That would make the list and thread more interesting to watch. Especially if more people starts pushing the limits.

Ill just put the link to Djozz emitter testing here.

And here is the video of frying an XM-L2. Skip to 3,40-3,45 for the exiting stuff..

Let the games modding begin! Big Smile

J)

BLF LED database – collaboration spreadsheet and latest news about where to buy LEDs
http://budgetlightforum.com/node/19342

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US

Sort by emitter current sure does sound fun, I think I’ll do just that.

I think 8.5-8.7A is the record to beat. 0:)

BLF ≠ B-grade Flashlight Forum

 

Slewflash
Slewflash's picture
Offline
Last seen: 4 months 1 week ago
Joined: 07/05/2012 - 08:43
Posts: 1977
Location: Australia, Melbourne

ryansoh3 wrote:
Sort by emitter current sure does sound fun, I think I’ll do just that.

I think 8.5-8.7A is the record to beat. 0:)

This will be interesting. If someone decides to push their XM-L2 above 8.5A they sacrifice reliability and longevity. Note that he didn’t drive the LED for 20 minutes at each interval, but only a few seconds. So while it may survive 8A for more than 30 seconds, who knows when you leave it on for longer.

With that being said, I love his crash tests and have it bookmarked and I want to see people breaking records! What’s the highest amps in an XM-L2 or even XP-G2 light that someone has ever gone? I mean I think 7A would be the highest with ryan here going to push his K50 to that level.

Slewflash 

aoeu
aoeu's picture
Offline
Last seen: 9 months 2 weeks ago
Joined: 02/16/2014 - 04:44
Posts: 793
Location: Sydney, Australia

ryansoh3 wrote:
Sort by emitter current sure does sound fun, I think I’ll do just that.

I think 8.5-8.7A is the record to beat. 0:)

I don’t think it would be hard to beat with active cooling! Even submerge the heatsink in iced water will keep the temps down easily. Purely for experimental purposes!

DB Custom
DB Custom's picture
Offline
Last seen: 2 weeks 3 days ago
Joined: 01/13/2013 - 22:28
Posts: 20733
Location: Heart of Texas

I will say that I’ve used the gasoline de-doming method quite a lot. But I’ve recently started hot de-doming in the light. I use a welders plate designed for a welders mask to view the emitter while it’s on in Turbo mode. Before turning the light on, I use a scalpel to cut around the base of the dome, very carefully of course. Then with the welders plate I view the emitter while in Turbo and use the tip of the scalpel to apply light upward pressure on the side of the dome opposite the bond wires. After about 30 seconds you can see the dome release and start to lift. Gently continuing the pressure the dome comes off clean in about 1 1/2 seconds after it starts that initial release. The remaining silicone can be left on the substrate, leaving a circle of clean substrate surrounding the emitter. This process really actually takes about 60 seconds all told. Put the reflector back on, replace the bezel and enjoy. This method seems to have a benefit in that the green hue so often associated with de-doming is greatly reduced or simply not present at all. The 6 or 8 I’ve done in the past few days have been very nice…warmer, but in a pleasing way.

Now, that being said, sometimes the reflector needs to be repositioned to gain accurate focus for the newly doubled throw. This has to be accomplished in whatever way the actual light in hand demands. This might mean that the emitter needs to come further into the reflector. In my TK61, it had to be pulled back actually. So each one will have it’s own requirement. Vinh states that the K50 V2 is a tough one to figure out, and calls himself a genius when he get’s it right. So it must be especially difficult to get right in this particular light, as it is in the TK61.

This remains to be seen and of course I’ll share my own findings when I get my K50 V2, even as y’all are doing yours.

DB Custom
DB Custom's picture
Offline
Last seen: 2 weeks 3 days ago
Joined: 01/13/2013 - 22:28
Posts: 20733
Location: Heart of Texas

I’m running my TK61 at 7.15A with the de-domed XM-L2 U2 1A. I don’t have it properly focused, got tired of working on it to be truthful, and it’s doing 541 Kcd.

Slewflash
Slewflash's picture
Offline
Last seen: 4 months 1 week ago
Joined: 07/05/2012 - 08:43
Posts: 1977
Location: Australia, Melbourne

Interesting Dale.. Looks like 7.15A is the new highest.. do you think that by leaving a small layer of the dome on it increases the throw less than fully de-doming in petrol? Vinh’s lights were tested by rdrfronty to be ~620kcd at only 5.75A. But then again, the difference between 541 and 620 is definitely within the margin of error especially with the inconsistencies with light meters. I know mine reads a little on the high side.

Slewflash 

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US

Slewflash wrote:

This will be interesting. If someone decides to push their XM-L2 above 8.5A they sacrifice reliability and longevity. Note that he didn’t drive the LED for 20 minutes at each interval, but only a few seconds. So while it may survive 8A for more than 30 seconds, who knows when you leave it on for longer.

With that being said, I love his crash tests and have it bookmarked and I want to see people breaking records! What’s the highest amps in an XM-L2 or even XP-G2 light that someone has ever gone? I mean I think 7A would be the highest with ryan here going to push his K50 to that level.

Yup, according to Cree’s datasheet for the XM-L2, it supposedly has an unlimited life span at its rated specs:

We all know that’s not true when we push our LED’s to 3x their rated current, and the previous gen LED’s are rated for 50,000 hours, which translates to 5.70 years.

Hopefully we can find a sweet spot between 5.70 years and insta-poof. Big Smile

aoeu wrote:
I don’t think it would be hard to beat with active cooling! Even submerge the heatsink in iced water will keep the temps down easily. Purely for experimental purposes!

What I have in my mind is setting up the K50 to a ridiculously high level 7-8A but only use it for bragging rights. Even with my modded K40 or TN31, Level 3 out of 6 is bright enough for me for “normal” uses.

DBCstm wrote:
I’m running my TK61 at 7.15A with the de-domed XM-L2 U2 1A. I don’t have it properly focused, got tired of working on it to be truthful, and it’s doing 541 Kcd.

Thanks for chiming in Dale. I’m not sure if a K50 can beat the TK61 when both modded to their max potential (7.0 – 7.5A) because the TK61 has a slightly bigger reflector. I couldn’t make up my mind but I decided on the K50 because if its superior build and stunning fins. Should definitely help with the heat.

You’re one of the people who say that the hot-knife deodming method yields better tints than the gasoline method. Interesting, I’d love to get to the bottom of this.

BLF ≠ B-grade Flashlight Forum

 

DB Custom
DB Custom's picture
Offline
Last seen: 2 weeks 3 days ago
Joined: 01/13/2013 - 22:28
Posts: 20733
Location: Heart of Texas

And I test at 5M, probably not ideal for this kind of thrower. I need to try a 15M test.

That said, I got an Olight SR90 from rdrfronty that I put an XM-L2 in, de-domed, at 6.62A I’m measuring 612Kcd interpolated from the 5M test. My new Courui big head has a beautiful beam profile, doesn’t seem to have convergence but appears to have a tubular hot spot that goes on forever…341.25Kcd. So each reflector is going to have it’s own anomalies.

DB Custom
DB Custom's picture
Offline
Last seen: 2 weeks 3 days ago
Joined: 01/13/2013 - 22:28
Posts: 20733
Location: Heart of Texas

And he claimed 670Kcd from his. That would indicate he’s gotten more throw from the K50 than the TK61. Not sure if he’s redone the TK61 though, as I understand he’s been working on raising the power output.

rdrfronty’s TK61 is making, they suspect, in the 6.5A range at the emitter.

Oh, by the way, that 7.15A is an emitter amp reading, taken off the negative lead about an inch away from the emitter with a clamp meter while in the light.

I’ve even got a little Eagle Eye X6 doing 6.77A to a de-domed XM-L2 for 1556 OTF lumens. Smile Even that little light is doing 82Kcd! :bigsmile:

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US

That’s good news, I almost got the Niwalker Vostro on sale but I’m glad that I got the K50 instead.

I wonder how the driver will handle solder bridged sense resistor pads.

Edit: Hoping that the head isn’t glued…

BLF ≠ B-grade Flashlight Forum

 

comfychair
comfychair's picture
Offline
Last seen: 5 years 6 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

I think I'll turn mine into a K50M. (scan their product lists, you'll figure it out easy enough)

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US
comfychair wrote:

I think I’ll turn mine into a K50M. (scan their product lists, you’ll figure it out easy enough)

With an MT-G2?

I wonder what the max voltage of the driver output is.

If it’s higher than the K40/TN31 we’re used to, then the soldier bridging method won’t work and will likely fry XM-L2’s. Precisely calculated SMD resistors (1206 sized?) would be necessary.

If not, it’s going to be almost impossible to drive a MT-G2 with the stock driver.

BLF ≠ B-grade Flashlight Forum

 

Fern
Offline
Last seen: 2 days 1 hour ago
Joined: 09/13/2011 - 04:29
Posts: 198
Location: New Zealand

What’s the typical max current the emitter sees in a standard K50 V2?

Tom E
Tom E's picture
Offline
Last seen: 1 hour 15 min ago
Joined: 08/19/2012 - 08:23
Posts: 14956
Location: LI NY

There are TK61's out there doing ~8A, at least tested in that realm - just fyi, probably the soon-to-be TK61vn V2.

Also, the 5 meter to 15 meter throw measurements will make a difference for some, maybe most lights, not all though. I measured a TN31 at 435 kcd or so at 5m, then rdrfronty measured the same light at 5m - about dead-on with me, then measured it at 15m and got ~580-585 kcd, so with that light, 15m does help. According to rdrfronty, once you get to 15m, further out distances don't make much of a difference, so he sort of standardized to measure his throw at 15m because of his numerous test trials at various distances.

Not sure what the stock amps is for a K50 V2, but must be up there to produce 1,600 OTF lumens. I'd be guessing at 4-4.5A.

Fern
Offline
Last seen: 2 days 1 hour ago
Joined: 09/13/2011 - 04:29
Posts: 198
Location: New Zealand
Tom E wrote:

Not sure what the stock amps is for a K50 V2, but must be up there to produce 1,600 OTF lumens. I’d be guessing at 4-4.5A.

Yes thought it must be close to the 5A mark which is I think is sensible and advantageous for emitter longevity and cell choice.

Modding not out of the question though Smile

18sixfifty
18sixfifty's picture
Offline
Last seen: 2 years 1 month ago
Joined: 12/25/2012 - 20:19
Posts: 4161

I didn’t manage to get in on the group buy, they were gone by the time I saw it. But one of our members has asked me to mod his for him. So I’ll be posting my results here. I really won’t know what I’ll be doing to it until I get into it. Most likely just a de-dome, focus work, heavy wires, lap the star, best thermal grease and resistor mod. I would think that 6 or 7 amps should really be enough for this light. Keep it reasonable so it’s not going to kill it anytime soon. Anything more and you end up in that area where it’s really diminishing returns. I would think this will come really close to the TK61 and should beat my TK61 as it’s not running that high of amps. I’ll be able to post the results of the two up against each other. That should be an interesting match up.

I really wish I had gotten in on this group buy but that’s the way the cookie crumbles and at least I get to mod one.

I’m a junky, I mod lights so I can sell lights so I can buy more light to mod so I can sell lights to buy more lights to mod.

Slewflash
Slewflash's picture
Offline
Last seen: 4 months 1 week ago
Joined: 07/05/2012 - 08:43
Posts: 1977
Location: Australia, Melbourne

I’m going to be following this thread closely. It’d be fun to see how high people push their K50s. Of course some emitters will fry at 8A, some might fry at 8.5 or even 9A. It definitely has better cooling fins than the TK61, but also has a slightly smaller reflector.

Slewflash 

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US

Sorry to hear that 18sixfifty. It was indeed a very nice deal and the lights did sell out quite fast. Seems like these are limited edition Supbeams indeed.

I can’t wait to get the light. Does anyone know the differences between the K50 and the K50 V2?

BLF ≠ B-grade Flashlight Forum

 

bibihang
bibihang's picture
Offline
Last seen: 1 week 2 days ago
Joined: 11/10/2011 - 09:32
Posts: 2445
Location: Malaysia

Subscribed! Thanks everyone. Smile

aoeu
aoeu's picture
Offline
Last seen: 9 months 2 weeks ago
Joined: 02/16/2014 - 04:44
Posts: 793
Location: Sydney, Australia

So what would be the amp draw per battery for this light in standard form?

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US
aoeu wrote:
So what would be the amp draw per battery for this light in standard form?

I’m assuming 1.5-2.0A per cell, so it’s quite manageable for any cell.

Things get a little different when modded, I’m guessing about 3.0-4.0A as the driver’s efficiency falls.

BLF ≠ B-grade Flashlight Forum

 

comfychair
comfychair's picture
Offline
Last seen: 5 years 6 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

4.5A output is around 16-17 watts, and with 4 cells in series the current draw is tiny. It should be well under 2A even when the cells are dead empty.

Slewflash
Slewflash's picture
Offline
Last seen: 4 months 1 week ago
Joined: 07/05/2012 - 08:43
Posts: 1977
Location: Australia, Melbourne
ryansoh3 wrote:
aoeu wrote:
So what would be the amp draw per battery for this light in standard form?

I’m assuming 1.5-2.0A per cell, so it’s quite manageable for any cell.

Things get a little different when modded, I’m guessing about 3.0-4.0A as the driver’s efficiency falls.

I’m a bit new to modding drivers so I don’t really know much, but why does the driver efficiency fall as you mod it? I’m wouldn’t have thought adding a resistor or bridging connections would cause efficiency to fall so low.
Pushing 12-16A to the driver when the emitter itself will only be getting 7-8A?

Slewflash 

comfychair
comfychair's picture
Offline
Last seen: 5 years 6 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

With the cells in series, amps is amps. It's just at a higher voltage (more power in watts).

Max voltage is 16.8v, min voltage is 12v. High voltage low current input, to get a low voltage high current output. That's a buck driver.

Driver efficiency falls because more current through the driver means more heat, and heat is wasted electricity. The actual efficiency might not change much, but the amount of wasted electricity goes up.

ryansoh3
ryansoh3's picture
Offline
Last seen: 10 months 1 week ago
Joined: 08/04/2012 - 08:21
Posts: 3890
Location: US

Slewflash wrote:

I’m a bit new to modding drivers so I don’t really know much, but why does the driver efficiency fall as you mod it? I’m wouldn’t have thought adding a resistor or bridging connections would cause efficiency to fall so low.
Pushing 12-16A to the driver when the emitter itself will only be getting 7-8A?

That’s a good question. I’m not entirely sure, but as you push the inductor, diode, etc. beyond their rated specs, the efficiency of the driver seems to drop noticeably. When I started pushing 6.0-6.5A to my K40 and TN31, I noticed that the driver heats up pretty hot and sucks a lot of current from the cells. If the cells are almost depleted, the driver would try to suck more current for the lack of voltage. This caused my springs to heat up and lose their tension. Copper braiding them helped a lot.

Here’s a video of a voltage sweep with my modded K40. Note how the current draw shoots up to 6.69A when the voltage is about 7.5V.

BLF ≠ B-grade Flashlight Forum

 

Pages