Thanx ImA4Wheelr. I think the best thing to do is to actually measure the current to the LED. 
Thanx DENGOH. I thought it was the same with all K40s/TN31s. I’ll measure the voltage across the resistors of my K40 to be sure.
Went ahead and did a direct wire to my old K40. When I turned it on, the LED gave a little flash then nothing. A closer examination on the LED reveals that its three wires ‘melted’. I have killed my K40. :_(
Perhaps the stock XM-L U2 can’t handle currents these high? :~
I’ll just wait for my XM-L2s to arrive before I reflow one into the stock K40 MCPCB (yes I think the stock copper MCPCB is better because it is thicker and wider). Meanwhile, I’ll change the lead wires to beefier ones then lay my K40 into its temporary grave until the time comes when I have to resurrect it again upon the arrival of my XM-L2s. :Sp
Ouch, sorry to hear that.
I’ve heard that XM-L2’s supposedly can handle more current than the first gen LED’s, so you might want to give them a try.
Sorry to hear that NightBird. You are probably already thinking this, but just in case. Since you will be unsoldering your led, you might as well check the current to the emitter. I would remove the by-pass wire first because it sounded like you were getting a lot of current to your emitter before. You’re driver may be slightly different somehow.
If you have another light that is easy to take the driver out of, you can attach longer wires to the K40 driver and run them to the the other light’s head. Solder them up and then cut one of the wires to connect your DMM to for reading current. Put your K40 back together, except for the reflector and lens (as the wires from the driver will be comming out of your K40 there).
The reason you will want the head of another light is that you will very quickly overheat the emitter you are testing current with without it being heat sinked. If you don’t have a spare light. You can attach an emitter to a chunk of aluminum or copper instead.
You mean connect my K40’s driver to another MCPCB but this time using longer wires and taking the bypass wire first? :~
I just finished testing the K40’s driver on an XM-L2 T6 3C on Noctigon and the LED went fine on all levels. I guess its the old XM-L that can’t handle the current. I just won’t use this XM-L2 because it’ll give a warmer tint once I dedome it. :~
To get the current to the emitter, should I connect my DMM between one of the lead wires and the LED connector?
Yes my XM-L2 T6 3C on Noctigon went fine after connecting it to the direct wired K40 driver.
Yes, that way you don’t have to work in such as small space to test the current. Yeah, I would see what current you are getting without the resistor bypass wire first. You may already be around 7 amps.
Yes, you are correct about how to connect to test current to the emitter.
Crap, that sucks it smoked the emitter - djozz's test went to 8.5-10A before smoking an XM-L2. I measured 3.48A at the carrier, so about 10.5A total. I'm hoping it's losing at least 3A somewhere. This evening I'll do an emitter amp measurement on my TN31. I have two of them now wired direct:
#1 (mine) 456 kcd measured
#2 (Scotlarock) 461 kcd measured
#2 is a little better in throw and lumens, each about 1,700 after 30 secs. Mine is a 2012 XML U2 with a de-domed XM-L2 U2 1A from IOS, Scot's is a brand new 2013 XM-L2, de-domed the stock XM-L2 (should be U2 1C).
One more (#3 - rikr) to go...
Longest I've run one at was like 3 1/2 minutes, continuously - gets pretty warm...
Got messed up! 
The XM-L2 on Noctigon was doing fine on all levels but after returning to my work desk after posting here in BLF, it just goes moonlight on all levels. :~
Removed the XM-L2 and tested it with my DMM - its OK. :~
Removed the direct wire and reinstalled everything - still the same moonlight on all levels? :~
I’ll continue fixing it this weekend when my XM-L2 U2s with cooler tints arrive. :weary:
I should have said this earlier. This has nothing to do with Nightbird95, but it made me realize it should have been said.
I wouldn’t recommend anyone do this resister by-pass mod unless you have some decent experience modding lights. There are a lot of things that need to be checked/changed out before making such a change. An example would be my light. It had a couple issues that would have blown the emitter or caused other damage had I just increased current. Actually, the issues it had would have probably resulted in short emitter life without an mods. I couldn’t believe the lack of QC on my light.
An experienced person like Tom E checks out a numerous of things in the course of doing a mod that will push the light to the extreme of its potential.
Ummm, dunno bout that, thoug I do somewhat thorough checks/QC's before engaging a mod - light meter tests, carefuly eyeball reviews, etc - Yes, you do want to see if there any anomaly's in function and performance. I do take my time and am very careful, but this mod is freak'n scary!! There's several guys here much more knowledgable and check the components, heat build up on SMD's, etc., but I've never done that. I was just banking on vinh's very brief description. What I'm afraid is he posted the direct wire method, but maybe left out something else critically related - I'm thinking that's not the case though, but has been a concern of mine.
This has to be all new - he's only been doing these TN31 resistor mods since maybe June/July time frame, I'm thinking? His TN31 mods earlier in the year didn't have any changes to the resistors.
Do you think inexperienced folks should be doing this mod?
I’m willing to bet you are checking quite a few things without even thinking about. Examples would be pill level, driver seating, signs of overheating (discolored springs, burn marks, etc), bad connections, etc. I guess most of it is covered by your “eyeball” reviews. But I bet your balls are looking for a lot more than you conscientiously realize.
Got to agree with you -- it's scary in terms of power and risk. Doing a C8, blowing up a driver or LED is no big deal - plenty of spares, relatively cheap, even if the entire light is ruined. You are right - bunch of things to look for with constant double and triple checking. I've blown a couple of drivers before and lost a couple of LED's - it can be a real PITA.
Tom, if you have any doubts or questions about Vinh’s setup, just let me know and I’ll send mine up to you to dissect next week. I don’t have it yet, but should get it sometime this week. I just want to do some testing and beamshots with it this weekend first.
Well, measured amps at the emitter and I got 5.8A at best. This is weird and somewhat disappointing. I am using a alligator clip jumper with only 20 gauge wire, while I got 12 gauge leads on my meter. After soldering back the emitter wire, I re-tested at the carrier and again got the 3.5A. Wonder if the cheapo alligator clips can't handle it (they are cheap ones), or the 20 gauge wire, or poor contacts along the way, or it's real.
Well what is done is done. At least I learned a lot from it and many noobs here in BLF would also learn from my mistakes.
Anyway, I think I can ‘resurrect’ my K40. I suspect that I may have shorted the sense resistors with solder melting under them but I just don’t have the time yet to check it. :~
Because of these mod ‘projects’, I’ve been procrastinating some of my ‘to do’s’ at work. BLF is so addicting that I think noobs like me would find it hard to balance family time with work time with BLF time. :bigsmile:
I got to go back to my work as I have to finish everything before I leave this evening for a business trip. When I get back on Saturday, I’ll have all the time to sort out what happened to my K40 (I hope).
Of course I’ll still be logging in to BLF even when I’m far from home and far from my lights. 
You are so right Nightbird. I am struggling with the same responsibilities/hobby balance. Have a safe trip. Look forward to hearing back from you.
Man, that sounds so much like what I was experiencing when I was trying to mod this light. There is something we are missing. I need to get back in there and help figure this out.
Well, the more I think bout it, I'm measuring 1,700 lumens and 456 kcd, so hard to imagine it is really getting 5.8A - it must be higher, just can't measure it accurately. I know my meter setup is good - verified against others and just measured a stock TN31 XML tonight at 1,145 lumens and 125 kcd -- should be right on the money.