Experiences with XM-L on copper board

I just wanted to share my findings. Today I modded my (old) Trustfire X8. I just changed the LED for a XM-L U2 on a copper heatsink. I have got the LED from LED-tech. The driver was left as is.
For those who don’t know that light: It is originally equiped with a T6 and runs on 2 18650 batteries. The pill is threaded into the reflector and the reflector is threded into the head. So, the heat transfer is expected to be good by this design. I added arctic silver to all threads in the head and also under the LED.
Before the emitter change it was driving at 1,58 A at the tailcap and it did exactly the same amps with the new setup.
Light output was increased by 40 % on fire up and settled to about 35% after one minute. Mesured this with a lightmeter at 85 cm distance (from the tabletop to the floor). That’s much better than I expected. The batteries I used were 3000 mAh Trusties that had 4,15 Volts at the first test before modding the light and I didn’t recharge them in between.

Nice results of 35-40%! I've only been doing over-driven setups with XM-L2's on SinkPADs, so wasn't sure how it would do on a straight stock under or ok driven driver. What copper board, a SinkPAD 20mm?

With the over-driven setups, it seems like the low resistance batteries (Panasonic PD's, AW IMR, Samsung INR) do better.

I’s a 16 mm sinkpad. Since it is a 2x18650 light. I don’t think a better battery type would change the output dramatically.

It looks more and more like that the introduction of the copper boards with direct heatsinking to the core has done more to improving output of custom lights than the new generation of Cree leds.

Your increases are incredible. Can you please post your lux readings before and after the mod? Thanks for posting.

fyi, just completed an XM-L2/SinkPAD update of a Small Sun T08, here's the results: https://budgetlightforum.com/t/-/16808#comment-349221. When you combine XM-L2/SinkPAD, some heat mgt improvements, upping the amps with regulation, you can get about 100% improvement. Basically I did the same mod to an HD 2010, same result.

Still like I said, what you got at 1.58A is pretty good. Are you sure you don't mean 20mm, not 16mm? Haven't heard of a 16mm - in fact we are trying to get them made for us/BLF by SinkPAD.

Its 8.4v, 1.58a at tail cap is getting on for 3a at the emitter. :slight_smile:

Oh, yes, forgot about the 2 batteries mentioned above! 3.16A effectively. Thanks Gords.

I believe that since the pill is both threaded into the head, and part of the electrical negative pathway, that the r&r of this pill is partly to credit for the gains.

The before and after measurements were taken at the tail cap, rather than at the LED, so any increase in current to the emitter would not have been measured by this method.

Any decrease in resistance in the circuit will provide improvements, and I think we are witness to that here.

Aren’t we seeing some improvements from a better sinked u2 emitter over a t6 emitter? Its even posdible it was a u2 at the upper end of binning and a t6 at the lower end, and better heat sinking and a drop in circuit resistance. Or am I being anal? :bigsmile:

Sorry, I’m still smarting from a brits = baddies reference joke :open_mouth:

Totally mis-read this before. So all that was changed was XML T6/alum star, to XML U2/SinkPAD. Well a few things goin on, nothing is that simple:

- this light has been out pretty long, so maybe the OP had it for 1 1/2 years under steady use and maybe the emitter is showing it's age?

- T6 --> U2, maybe 10-15% I'm guessing?

- Alum star to SinkPAD? This is the big mystery - I'm thinking at 3.2A, maybe another 10-15%?

- LED wires - did they stay the same or were upgraded? A few percent there possibly?

- maybe better star mount, or maybe just fresher thermal grease or epoxy would help if it's an old light?

It all could add up, Oldienea would know best of the history and mod work.


It’s 16mm. The light was barely used but still it’s possible the original LED was a T6 at the lower end of the binning and the new U2 is at the upper. I added new leads, because in fact I have destroyed the insulation of one of them. But the new ones are just barely thicker than the original were. Sur when I was up to it i also cleaned the lens, before reassembling, but it was not leally dirty. The reading of the luxmeter was 32000 before the mod and 44500 after at switch on, which is 39% increase. One minute later it was 43500.

I don’t think the emitter current changed significantly if that driver was regulating in both cases, which I’m quite sure it was.
Measuring output using a lux meter (especially at 85cm) with direct exposure can be quite variable in itself.
The better way to do comparative tests on the same light would be a ceiling bounce test in a small room (i.e. integrating bathroom). It may be too late for that test now, however.
Either way, you definitely have a brighter light, regardless of numbers. Congrats to that!

Ohhh, ok, it's not a SinkPAD but it is a direct copper connection. Yes, saw that - what a waste though - why they dont' sell the star board on their own, and why just a U2, not even a U3, and pricey. Almost like they are stuck in a time warp 6-8 months ago, but still, you got good results with it.

Nice result oldienea, its great when a plan comes together :slight_smile:

They did have U3 a few months back, they sold them all in about a week…that was before the sinkpad frenzy!