Crash-testing a XM-L2 and a XP-G2 on copper Sinkpads

109 posts / 0 new
Last post
ryansoh3
ryansoh3's picture
Offline
Last seen: 2 months 2 weeks ago
Joined: 08/04/2012 - 08:21
Posts: 3876
Location: US
comfychair wrote:

They don’t pop or even turn blue at 9 amps, though based on my one sample I can’t say it puts out more light at that much current.

hmm…. they don’t have a max current that ultimately kills them?

BLF ≠ B-grade Flashlight Forum

 

comfychair
comfychair's picture
Offline
Last seen: 2 years 5 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

Almost certainly, but I was using an actual driver in a light with normal batteries. I don't know of a driver that does more than 9A and I don't have a big power supply to drive it directly (I have a few, but they're all in the range of 11-16v).

djozz
djozz's picture
Offline
Last seen: 1 hour 16 min ago
Joined: 09/07/2012 - 17:04
Posts: 13285
Location: Amsterdam

djozz wrote:

mattthemuppet wrote:
awesome! There's nothing a bit of science that's both informative and destructive at the same time! I wonder if you dedomed those LEDs and bridged the bond wires with a dab of solder they'd work again? Might be fun to try out if you have some spare time :)

I tried to repair a broken bond wire once on a dedomed xpe going wrong and I found it impossible to do. But well, these leds are a bit bigger so I'll give it a try Smile, just dumped them into some Coleman Fuel:

Ok, I tried to repair the broken bond wires and it was a complete failure. But what is better than looking at other peoples failures Laughing? So here we go, enjoy!:

 

The leds with most of the domes removed (left xm-l2, right xp-g2, same magnification), still in the Coleman to see through the silicone well. You can clearly see the incinerated bond wires (strangely they go two at a time Undecided)

 

I cleared away more silicone from the xp-g2 and also the bond wire debris and bended the bare wire ends are close to each other, added some solder paste and touched it with the iron. The solder paste melted and got immediately sucked away by the iron, no connection Frown.

 

Added more solder paste, same thing, and also destroyed the wires. Went over to the xm-l2 led, with loads of solder paste.

Failed miserably again, destroyed the wires again, now I cleared all the debris and tried to solder what was left of the connections with big blobs of solder wire:

Actually when connected to a led tester I got some very faint light, so I tried soldering again for a better connection. Well, it ended with nothing at all, end result: dead and ugly looking led:

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

ryansoh3
ryansoh3's picture
Offline
Last seen: 2 months 2 weeks ago
Joined: 08/04/2012 - 08:21
Posts: 3876
Location: US
comfychair wrote:

Almost certainly, but I was using an actual driver in a light with normal batteries. I don’t know of a driver that does more than 9A and I don’t have a big power supply to drive it directly (I have a few, but they’re all in the range of 11-16v).

Ah, I see what you mean, thanks for the explanation.

djozz, awesome work and nice try repairing the wires. Smile

Do you think a dedomed LED will handle less current?

BLF ≠ B-grade Flashlight Forum

 

RaceR86
RaceR86's picture
Offline
Last seen: 2 years 7 months ago
Joined: 02/22/2012 - 08:32
Posts: 3777
Location: Norway

comfychair, are you using this 9A driver?
Does it give output 9A to an MT-G2 on high?

You do have a point though. If if there are no other drivers out there that can handle 9+ out there it is a bit limited use in knowing. But its always nice knowing how much “head-room” you might have. J)

Nice attempt djozz. At least you tried, which is more than I have though of in the past and with my recently fried XP-G2. Smile

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

comfychair
comfychair's picture
Offline
Last seen: 2 years 5 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

There are two planes on the substrate, the die is sitting on the negative plane & the bond wires jump over the little isolation trench to make the positive connection. Looks like the solder bridged across the trench. Did it act like there was a dead short, or just total open ckt?

I think the way to do it would be to try and stick a wire to the connection points on the die first, then lay them over and tie in to the + plane. The issue you'll run into trying to do it with solder is that with such small tolerances enough heat to melt solder at one end will also melt it at the other at the same time and the whole thing falls apart.

comfychair
comfychair's picture
Offline
Last seen: 2 years 5 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

RaceR86 wrote:
comfychair, are you using "this ":http://intl-outdoor.com/9a-3mode-5512v-circuit-board-p-561.html 9A driver? Does it give output 9A to an MT-G2 on high?

Yes and yes. http://budgetlightforum.com/node/21921#comment-411218

djozz
djozz's picture
Offline
Last seen: 1 hour 16 min ago
Joined: 09/07/2012 - 17:04
Posts: 13285
Location: Amsterdam

comfychair wrote:

There are two planes on the substrate, the die is sitting on the negative plane & the bond wires jump over the little isolation trench to make the positive connection. Looks like the solder bridged across the trench. Did it act like there was a dead short, or just total open ckt?

I think the way to do it would be to try and stick a wire to the connection points on the die first, then lay them over and tie in to the + plane. The issue you'll run into trying to do it with solder is that with such small tolerances enough heat to melt solder at one end will also melt it at the other at the same time and the whole thing falls apart.

Ah, never knew that the two wires are parallel, thought that current was going from one to the other, that explaines easy why they both go.

Perhaps an idea is to start with a functioning led and 'enforce' the bond wires with solder blobs.

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

comfychair
comfychair's picture
Offline
Last seen: 2 years 5 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

Hot air would work better, probably from underneath to keep from knocking anything out of place. If you could take a suitably small piece of wire formed to mirror the bond wires, give a light coat of paste, get it positioned and then add heat...

zeremefico
zeremefico's picture
Offline
Last seen: 1 hour 6 min ago
Joined: 03/27/2012 - 02:44
Posts: 1289
Location: Greece
RaceR86 wrote:
I have a driver (from STL- V6) that I can easily resistor mod from about 2,8-5A as I wish.. Modded it to 4,7A yesterday, but did not to dare try it on an XP-G2 after I had fried one at 5A with another driver. Dialed it down to 3,7A just to keep it moderate. 0:) Im already feeling inspired to go higher now.. J)

Can you show a pic with your resistor mod?
Did you use a Trim-pot?

₪₪₪₪ ΟΥΔΕΝ ΚΡΥΠΤΟΝ ΥΠΟ ΤΟΝ ΗΛΙΟ ₪₪₪₪

My YouTube channel

Flashlights & edc gear

K40M F16

RaceR86
RaceR86's picture
Offline
Last seen: 2 years 7 months ago
Joined: 02/22/2012 - 08:32
Posts: 3777
Location: Norway
zeremefico wrote:
Can you show a pic with your resistor mod? Did you use a Trim-pot?

djozz sorry about the OT.

Added one R120 for about 4,7.
Two R120 in series (on top of the stock resistor) was about 3,65 I think. Stock output on mine was quite close to 2,8 I think..

Its basically the same driver that I modded in this messy thread. But there are a few differences. If you have the same looking driver, yours might have differences too. I have not tested it for more than a short period though, so mod at your own risk… Wink

Edit: I have no intention of using the huge stock aluminium mcpcb in that light. Ill kinda make a cooper “pill” for a copper mcpcb to sit on… Ill might make a mod thread some time in the future…

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

Tom E
Tom E's picture
Offline
Last seen: 1 month 2 days ago
Joined: 08/19/2012 - 08:23
Posts: 10899
Location: LI NY

I'm seeing consistent issues with not being able to get 3.5A+ out of an XP-G2/SinkPAD combo, whenre I can with an XM-L2, in a single cell Nanjg driver. I think you're right - XP-G2's have a higher Vf demand.

Tom E
Tom E's picture
Offline
Last seen: 1 month 2 days ago
Joined: 08/19/2012 - 08:23
Posts: 10899
Location: LI NY

Firelight2 wrote:
DENGOH wrote:
What an experiment. I am more confident to do 3A for XPG2 now. Thanks.
Yes. Tonight I stuck an XPG2 neutral on a Noctigon inside a Sipik 68. I replaced the driver with a Nanjg 105c with 3 extra regulator chips attached for 4 amps total, running on IMR 14500. It's bright, but unfortunately, the focal length is wrong for the new star. I'll need to file down part of the outside of the pill in order to make it focus properly.

What is the amp tailcap measurement? I've been unable to get above 3.3-3.4A I believe. It still does well and bright, but just not getting the full amps.

RaceR86
RaceR86's picture
Offline
Last seen: 2 years 7 months ago
Joined: 02/22/2012 - 08:32
Posts: 3777
Location: Norway
Tom E wrote:
I’ve been unable to get above 3.3-3.4A I believe.

Basically the same as with me and others.. XP-G2 seem to need two cells (and a buck driver) in order to go higher..

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

djozz
djozz's picture
Offline
Last seen: 1 hour 16 min ago
Joined: 09/07/2012 - 17:04
Posts: 13285
Location: Amsterdam

RaceR86 wrote:
djozz sorry about the OT.

That's perfectly alright, I am guilty of that myself all the time.

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

mattthemuppet
Offline
Last seen: 4 days 21 hours ago
Joined: 01/04/2012 - 16:20
Posts: 1259
Location: WA

fair play for trying Smile I’d suggest some enameled copper wire to replace the bond wires, but making suggestions is easy when you’re not doing the work!

ImA4Wheelr
Offline
Last seen: 1 hour 5 min ago
Joined: 02/03/2013 - 14:51
Posts: 7768
Location: SC

I wonder if dedoming with gasoline weakens the wires. I know gasoline weakens some metals. I learned the hard way when I soaked a motorcycle chain in gasoline to clean it.

Or maybe the dome just physically supports the wire and helps prevent it from sagging and shorting at a lower amperage.

Werner
Werner's picture
Offline
Last seen: 3 months 3 days ago
Joined: 10/19/2012 - 15:00
Posts: 3677
Location: Germany

I have tried once to repair a bond wire on a XML and the results were like yours…one drop solder on the phosphor and the wires still not connected…

Firelight2
Firelight2's picture
Offline
Last seen: 15 hours 29 min ago
Joined: 04/08/2011 - 15:17
Posts: 2716
Location: California

There are conductive epoxies that are advertised as a replacement for solder. No idea how they’d do in a high-current, high-temperature application. But it’s possible something like that might work better at repairing bondwires if you can find one with the right specs.

Hikelite
Offline
Last seen: 3 months 4 weeks ago
Joined: 07/13/2011 - 16:18
Posts: 3614
Location: RO

djozz wrote:

Slim Pickens wrote:
I wonder how an LED like the "Luxeon T":https://www.google.com/url?sa=t&source=web&cd=1&ved=0CDQQFjAA&url=http%3... would perform on a Sinkpad board.

I'd love to see the Luxeon T for sale on a Sinkpad, it seems excellent for flashlight use and I understand it has very low internal thermal resistance. It is just that it has a very different solder pattern and needs its own Sinkpad.

MT-G2 is 1.5C/W which is pretty low, Luxeon T is 3C/W.

 

 

djozz
djozz's picture
Offline
Last seen: 1 hour 16 min ago
Joined: 09/07/2012 - 17:04
Posts: 13285
Location: Amsterdam

I took some time to put the video data into graphs, and in these graphs the numbers sure look official Surprised, watch out for good-looking graphs!!! In each video I turned the current up to a fairly high level and then turned the current down to minimal and started over again but now until *poof*, I did this to see if any 'burn-in' effect (lowering of the Vf at the same current) could be detected. It turned out that the second run both with the XP-G2 and the XM-L2 gave almost exactly the same numbers, so no burn-in effect here (Perhaps a burn-in effect will take place when the led is driven way over its maximum output).

Take notice that the output lumen estimates were calculated from ceiling bounce numbers that are derived a bit different with the two leds, the XM-L2 had a big 49mm reflector on top without lens, the XP-G2 had a small 20mm reflector on top with a AR-lens. The reflector size should not make a difference for the lux-reading, but the lens does a bit. (sorry I messed that up a bit in the experiment)

I believe that the voltage reading coming from the power supply in the XM-L2 graph -although a bit rough- does reasonably represent the actual Vf of the emitter at that current. The voltage in the XP-G2 graph was measured more accurately and directly at the (thick) led wires.

I just used the current reading of the power supply for the graphs, I did not check that with a DMM (should have done that). But for now I see no reason to distrust the current numbers.

Here we go:

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

bdiddle
Offline
Last seen: 2 months 2 days ago
Joined: 12/14/2012 - 14:36
Posts: 811

Do you think you could do that with a Gen 1 XML? I’d donate one if needed. I am curious how an XML U3 scales, and if it can take more current at the top end.

And thanks for taking the time to do this!

Newb

djozz
djozz's picture
Offline
Last seen: 1 hour 16 min ago
Joined: 09/07/2012 - 17:04
Posts: 13285
Location: Amsterdam

bdiddle wrote:
Do you think you could do that with a Gen 1 XML? I'd donate one if needed. I am curious how an XML U3 scales, and if it can take more current at the top end. And thanks for taking the time to do this!

That would be interesting, and I could do that, and I do have one spare (albeith a T6), but what I lack most is time for the hobby. And to be honest: these kind of tests should be done a bit less leisurely than I do them, I am not a that accurate kind of person I'm afraid. I am hoping for someone else with time, equipment (calibrated stuff, integrating sphere) and precision to do this kind of testing (texaspyro for instance is much better equipped than me in doing these kind of tests Wink).

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

Tom E
Tom E's picture
Offline
Last seen: 1 month 2 days ago
Joined: 08/19/2012 - 08:23
Posts: 10899
Location: LI NY

Pretty clear from the graphs the Vf tracks higher for the XP-G2 across the spectrum of amps -- wish I knew this sooner, but this sure does explain a lot.

Thanks djozz!! Great work/info!!

comfychair
comfychair's picture
Offline
Last seen: 2 years 5 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

Dedomed XPG2 R5 3D/Sinkpad/big CPU heatsink, 1-2-3-4 cells in parallel, direct drive

Samsung INR-15Q, unprotected
1cell - 2.56a 3.61v
2cell - 2.93a 3.75v
3cell - 3.15a 3.81v
4cell - 3.24a 3.85v

Panasonic NCR 2900mAh, unprotected
1 - 2.53a 3.59v
2 - 3.07a 3.77v
3 - 3.28a 3.85v
4 - 3.42a 3.89v

Tom E
Tom E's picture
Offline
Last seen: 1 month 2 days ago
Joined: 08/19/2012 - 08:23
Posts: 10899
Location: LI NY

Wow, that's interesting #'s. I got the 15M's and 20Q's, and the 20Q's are very good - seem identical to the 20R's, but to see them beat out by standard Pana's with qty 2 or higher in parallel, is kind of mind blowing -- not sure why? Do you know?

Also, the 2.56A for a single 15Q seems rather low - I'm definitely getting higher #'s with a Nanjg driver and 22 gauge wires with a 20Q. Mine are tailcap measurements though...

comfychair
comfychair's picture
Offline
Last seen: 2 years 5 months ago
Joined: 01/12/2013 - 05:39
Posts: 6198

The 15Qs have been sitting a while, they were all 4.08-4.09v at the start, the grey Panasonics were just charged yesterday. Also, I only have one meter so I had to change the wiring around between the voltage & amperage measurements, could be something got jiggered. I have big fat wires on my battery box but the leads after that were single 22AWG, and the test leads for the meter are full length, something like 3 feet each, stock Fluke leads. Mostly it's only interesting in showing how splitting the load across multiple cells and the resulting higher voltage affects things.

I figured more than one INR would have killed it dead instantly, I was surprised by those numbers.

djozz
djozz's picture
Offline
Last seen: 1 hour 16 min ago
Joined: 09/07/2012 - 17:04
Posts: 13285
Location: Amsterdam

Hey, I changed my mind. I just ordered a new MT-G2 for a flashlight built I am going to do when I have the time.

This leaves me with one spare (5000K 6V) for yet another crash-test on a Noctigon board that I may do one of these days. But please give me a reason not to do this test (like: he or she has done that already and the results can be found here, or: relax !, I will do this test for you and I am doing a much better job at it than you do) because I am still reluctant to demolish a perfectly good MT-G2 Frown.

link to djozz tests 

“I used to think that top environmental problems were biodiversity loss, ecosystem collapse and climate change. I thought that thirty years of good science could address these problems. I was wrong. The top environmental problems are selfishness, greed and apathy, and to deal with these we need a cultural and spiritual transformation. And we scientists don’t know how to do that.”   (Gus Speth)

garrybunk
garrybunk's picture
Offline
Last seen: 1 hour 19 min ago
Joined: 10/31/2011 - 09:25
Posts: 6038
Location: Johnstown, PA

djozz wrote:
This leaves me with one spare (5000K 6V) for yet another crash-test on a Noctigon board that I may do one of these days. But please give me a reason not to do this test (like: he or she has done that already and the results can be found here, or: relax !, I will do this test for you and I am doing a much better job at it than you do) because I am still reluctant to demolish a perfectly good MT-G2 Frown.

Umm . . . I don't think anyone is going to say those things!  We're waiting for you to do it! Smile  Thanks for these tests by the way!  This is very helpful information to many of us!

-Garry

My Bike Lights Thread, Optics (TIR) Comparison Beamshots, Diffusion Techniques


NOTE: Moving my photos from PhotoBucket to Flickr. PM me or post in the thread if you can’t see my images and need me to fix or send a gallery link. PhotoBucket images should remain visible until November 2018.
RaceR86
RaceR86's picture
Offline
Last seen: 2 years 7 months ago
Joined: 02/22/2012 - 08:32
Posts: 3777
Location: Norway

Very nice and useful graphs!
I have bookmarked this page! Wink

Djozz, would be nice have a chart going to 10A or slightly beyond on the MT-G2. You dont have to kill it just for the sake of killing it. But if output seem to continue to rise, then I would not mind seeing what 12A would do, despite not knowing of any drivers that will do that kind of output.
There is not that much point in going far beyond 10A just for the sake of killing it. But it would be nice to know if there is some headroom left at 9A. 0:)

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

Pages