Ohh - No, I was trying to be appreciative, but of course the facts are kind of bad news - PilotPTK's inputs are invaluable!!
So, I'm not sure what this means in practical terms. I also built up a UF-V3 with an XM-L2 U2 and SinkPAD - this light has a KD V2 + 1 380 7135, total of 3.42A. Now with this light, all I'm getting is 3.35A with a Pana 3400, unprotected, fresh off the charger, so it seems to be suffering some of the same issue as well. As far as lumens, it's getting a respectful 901-870, and 24 kcd for throw (low, but maybe understandable at 4.3 meter measurement). The 3.35A though drops a bit quickly, and the Pana 2900's read at 3.25A or so. So here, things seem to be at the edge, while atempting higher amps doesn't make it. Unfortunately, I didn't have good #'s on the UF-V3 before the XM-L2/SinkPAD mod, but it needed a replacement emitter (I was experimenting with dedoming, not great results).
jmpaul320 - actually was debating replacing those wires - they are a little thin and lots of them! Took out the 2 screws holding down the reflector to look at them. Replacing the emitters would be awesome too, maybe U3 1C's, but for 7, cost is high - FastTech for $32.50. IMR 26650s - I need where to find them, best price - they unprotected? Gotta do some search'n...
I don’t think an IMR cell will help here. You want a cell that holds its voltage as high as possible for as long as possible. Sanyo cells are pretty good with that.
If XM-L2 is running at a higher Vf, then at the same current in it consumes more power (obviously). Does anyone have Vf measurements comparing the XM-L2 U@ to an XM-L U2 at the exact same drive current? I’d be interested in knowing the difference. A few milliVolts do not matter, but 0.1 - 0.2V do.
Here’s a great reference that I’m sure most of you have seen
Looking at the purple section (5A), but orange, too (3A), the AWs and CGR18650CH do ok, but can’t compete with the latest 3400mah batteries - that sanyo looks great, too
source: Battery test-review 18650 summary
Thanks for pointing that out, it didn’t sink in the first pass. A 150mV difference is significant. It may be the process used (XM-L2 is Silicon Carbide, XM-L was not). I doubt it’s substrate or bond wires which would result in no gain in output other than heat. I think the added Vf is giving a little more light output, not just heat. It should not affect things in 2S or higher kights, but out ever-popular single cell lights will have an earlier roll-off in output.
This makes the buck-boost driver even more important.
OK, I just checked the datasheets:
XM-L is spec’d to be typical 50mV higher Vf than XM-L2… so our findings contradict the specs. The plot thickens.
Anyone like to do a comparison of Vf below the 3A limit? I would but I have no XM-L2 parts yet.
Dont forget that the XML is rated at 25°C while the XML2 is rated at 85°C. Someone did the math and the typical Vf of the XML2 at the same temperature is a little higher.
XM-L2 is binned at 85C, however not all specs work like that. Most typical spec values are reference only, a general guide. The min and max specs are the values that are guaranteed within the specified operating temperature. Both XM-L variants have a Vf max of 3.5V so they both are within spec. I used the typical number to show what Cree expect to see in general.
Ah, I see said the blind man.
So they did characterize the Vf at 85C. Vf will be higher at 25C, coefficient is –1.6mV/C. For a 60 degree difference that is about 100mV which is not too far from the measurement we have.
If we take the 3.35V @ 3A @ 25°C for XML and 3.3V @ 3A @ 85°C for XML2, with a coeffiecent of -2.1mV/°C respectively -1.6mV/°C that would be:
3.224V @ 3A @ 85°C for XML vs. 3.3V @ 3A @ 85°C for XML2
and
3.35V @ 3A @ 25°C for XML vs. 3.396V @ 3A @ 25°C for XML2.
Of course those are just "typical" specs, but there is a difference. Judging from the graphs in the datasheets, the Vf of the XML2 rises faster than the Vf of the XML with rising current.
great results - ive got a u3 1c going in my supfire (not copper though) - i will probably push 3.5a with a 7135 driver… maybe ill get about 1000otf lumens then…
Really wierd thing about the SupFire - it's 3.85A 7135's, but only measured 3.63A on a off-the-charger Pana 3400.. wtf? Also, when I tested the driver, I wired it up to a T6 test emitter and measured 3.93A... Again, wtf? So is the SinkPAD to blame for that? But you know, it's a crazy impressive result - I never expected to get in the realm of 1,100 lumens from the X5, specially at only 3.63A and with a XML U3 1C, where the XinTD has the XM-L2 U2 1C. The XinTD, however, is dissapointing - was hoping for higher results, and the amp measurement is low - reasons probably as discussed with a higher Vf pull by the XM-L2.
Oh - I re-tested lumens on the XinTD with a off-the-charger Pana 3400, measured 4.19v and 3.49A this time, and got 1122-1081 lumens, matching the X5 exactly!
Hardly witchcraft… Lower the temperature of the emitter die and you raise the Vf of the emitter. Every degree C lower, and the Vf goes up by .0021 Volts
Wow PilotPTK! You come through again!! This is getting really interesting!! So the SinkPAD is doing it's thing, but is total power draw from the battery staying the same? Higher Vf, lower amps?
