Yeah I just missed the battery thing for a bit 
Yes, you will have to compare with my battery tests.
It might also be interesting to compare to 3xAA cells.
That was a lot of drivers 
I will order some them and see about building a 3xXML test rig.
XP-G2 would be nice too.
Thinking: I can build the 3*XML with plugs for 1, 2 and 3 leds, and replace the led on this rig with a XP-G2 = 100% use of the test rigs and parts.
Excellent info and work. I think the answer to why Comfy was getting lower current driving MG-T2’s with 7135’s is here. I don’t know about his set up, but my 7135’s were soldered to a copper pipe. So they had lots of heat sinking and I was getting close to rated output. I later learned I was getting as much as the cells I had connected could give.
since each led will have some variation in Vf it might be a good idea to test the led bare and post its voltage/current curve, and indicate which particular led was used in each driver test
Great tests HKJ! Very useful. You have enough suggestions for now. 
Adding the additional measurements you list would be good. 
You can sort of work it out from the curves marked "led", but you are right, it would be nice to have a voltage-current curve for the leds.
I will, of course, tell if a particular test is done with XPG, 1xXML, 2xXML, 3xXML
I do not know if there are enough suggestions, it is a questions about what people would like included the test.
i thought of that too, if the same LED is used for all single LED drivers then it makes the tests a lot more comparable and we can discern slight anomalies in drivers when comparing different tests 
Can i also suggest the Qlite 3.04A driver to test, i assume it will give identical current shifted results to the 2.8A 105C but it would be nice to have confirmation
I was planning on using the XPG for all drivers with 1.5A or less in led current and the XML for anything up to 3A.
Please: shop and sku or a link.
That is completely reasonable
I was thinking it would be a massive undertaking, but a chart of voltages for XP-G, XM-L and XM-L2, i have long suspected different tints have different forward voltages (and maybe different bins might as well for the same tint), however instead of testing doens of chips, how about a few common ones, an XP-G2 R5, XM-L T6 3C and U2 1A and U3 1C, and XM-L2 T6 4C and U2 1C
Links:
xp-g2
http://intl-outdoor.com/cree-xpg2-r5-3c-led-20mm-mcpcb-p-703.html
http://intl-outdoor.com/cree-xpg2-r5-2b-led-16mm-mcpcb-p-704.html
xm-l
https://www.fasttech.com/products/1609/10001906/1136804-cree-xm-l-t6-3c-5000k-white-led-emitter-with-16mm-
https://www.fasttech.com/products/1609/10001903/1287502-cree-xm-l-u2-1a-6300-7000k-white-led-emitter-with-
http://intl-outdoor.com/cree-xml-u3-1c-led-16mm-mcpcb-p-593.html
xm-l2
http://intl-outdoor.com/cree-xml2-t6-4c-led-14mm-mcpcb-p-757.html
http://intl-outdoor.com/cree-xml2-u2-1c-led-12mm-mcpcb-p-743.html
http://intl-outdoor.com/noctigon-xm16-mcpcb-cree-xml2-u2-1c-led-p-746.html
i know this would be a different project, so its just an idea
http://intl-outdoor.com/qlite-reva-71358-multiple-modes-circuit-board-304a-p-710.html
No.
To get a good idea of Vf it is necessary to test a couple of leds of each type (It varies due to production tolerances).
I will not expect any Vf differences due to tint (That is variations in phosphor), but bins might have an impact on Vf.
I will put it on the list.
i had the same conversation a few hours ago with another forum member, there is even an off chance you get a high end of tolerance T6 and a lower end of tolerance U2, we would never know without several of each chip tested.
Perhaps a similar test, 3 to 5 XM-L2 U2 all the same tint just to see what the forward voltage is, or perhaps any chip of the same bin/tint that you may already have
“The driver will reduce the current when it gets to hot (8 linear drivers on a small PCB do get hot very easily), as can be seen at 4 volt and above.”
Yes, this is why must to solder the driver to the pill. I always wonder when I see a driver mod like this:
You should open Paypal donations 8)
Your site is the place to go when I need info on batteries or chargers and I already feel like I O U
Thank’s a lot for your effort, HKJ. This provides great information for choosing drivers, and for understanding them.
I’d like to see charts for the LD-2C which seems to be the successor to the LD-29, the latter is already on your list.
http://intl-outdoor.com/ld2c-3a-12-cell-circuit-board-p-732.html
It has similar specs, nice modes and is smaller. Most interesting: whereas the LD-29 seems to underperform in single cell mode (only 2,5-2,6A), this is not mentioned with the LD-2C. So there’s still hope…
If you could test both, LD-29 and LD-2C, in comparison and in both voltage-ranges, would be great.
I’d gladly provide you with my new, unused, untouched LD-2C from IOS that came the other day. PM me your address if you want to use it. I wanted to fiddle with it myself but your rig is much better and results are comparable that way. Or PM me your PP-address, I second the idea of Matjazz.
Greetings
Again, thanks for a good, informative report.
Strange that you have a 3.04A variant of the 105C. A variant that Fasttech should not have in stock (8x0.38A).
Thank you HJK for doing all this work! Your results for the Nanjg 105C confuses me though, could anyone please help me understand what’s happening here?
I have read forum members here (Old Lumens springs to mind for one, when he was modding the 3 and 4 AA to “D” battery carriers) trying to use three AA Eneloops to achieve 2.8A at the LED and finding that they weren’t getting anywhere near 2.8A but, when they used four AA Eneloops, they hit the 2.8A easily, and it remained “regulated” at 2.8A for a reasonable time.
Their driver would have been seeing something roughly around the 4.8 volt mark from four Eneloops under full load, but your graph shows that by 4.5 volts the output has apparently dropped back down to 1A? Why do your results differ so greatly from their real world results? Is it really just due to the heat build-up of the driver you tested?
This exact driver is rated from 3 to 4.5 volts. I realise that 4.5 volts is not their ideal input voltage, and a lot of that will get burnt off as heat, but why would it be rated to 4.5 volts if it couldn’t produce the required current at that voltage?
Please understand I am not trying to devalue the work you have done here!!! I just can’t quite make sense of it all.
HKJ, Did you get a sense for how fast the 7135’s overheat when not properly heat sinked?
I know it would depend on a lot of factors, but just wondering how fast current would drop in a worst case situation after light if flipped on with fully charged cells.