Actual Lumen Readings in 10.5in Sphere.

[Yoda4561]

Folks have run them unresistored off a 4.2v IMR18650, pushing the upper limits, but with a 3-something volt LiFe they shouldn't have any trouble at all. Pretty sure we lost a number of pertinent posts, but here's one. http://www.candlepowerforums.com/vb/showthread.php?306783-XM-L-and-an-IMR-18650 I think some folks tried the 26650 batteries, but there wasn't enough voltage sag and they instaflashed the LED. Off the IMR18650's I think a few went blue (if the thermal path wasn't ideal) but didn't most didn't seem to suffer permanent damage.

 

[ti-force]

I was really temped to purchase the S12 myself to see what it did in OTF.

MOre importantly; I wanted to take out the driver and run it direct drive.

bigC

This S12 is a loaner, so no modding this one . CPF member Got Lumens? was kind enough to send his light to me so we can all know how it performs :thumbsup:. I didn't have a chance to finish testing last night (life called), but I hope to finish up tonight and post my results. I will say this, the S12's output after 3 minutes runtime on high is greatly affected by the temperature of the the light (pretty much expected from something that's thermally controlled). The cooler you can keep the light, the better the light sustains its high output. I'm also testing to see if added resistance has a negative affect on output.

This light would get extremely hot if it didn't have thermal regulation, so if you decide to dd one, I would expect it to get real hot within a few minutes, maybe less. It would be extremely bright for a short period of time, especially for such a compact light.

 

[ma_sha1]

I was thinking that a 3 volt LiFePO4 would make a good power source for a DD XM-L. What do you think?

The Lifepo4 is 3.3-3.4v normally w/o load, too close to XML-vf. When under load, the battery voltage will drop & Led vf will rise. Single Lifepo4 will not poof the XML, just the opposite, it'll not be able to provide enough Amps to get to the 3A spec, probably a lot less. It's not a good solution due to under drive not over drive XML in DD.

Single IMR 26500 & 26650 are 4.1v os so normally, DD get 3-4 Amp on several XMLs I tested & for a few other XMLs, it went up to 6 Amp. It takes very small vf. difference to see huge current variations under single IMR.

Since XMLs are not binned by Vf like P7 used to be, one can not manage Direct Drive as confidently as P7.

 

[recDNA]

7

The Lifepo4 is 3.3-3.4v normally w/o load, too close to XML-vf. When under load, the battery voltage will drop & Led vf will rise. Single Lifepo4 will not poof the XML, just the opposite, it'll not be able to provide enough Amps to get to the 3A spec, probably a lot less. It's not a good solution due to under drive not over drive XML in DD.

Single IMR 26500 & 26650 are 4.1v os so normally, DD get 3-4 Amp on several XMLs I tested & for a few other XMLs, it went up to 6 Amp. It takes very small vf. difference to see huge current variations under single IMR.

Since XMLs are not binned by Vf like P7 used to be, one can not manage Direct Drive as confidently as P7.

Thanks for the explanation. I thought that th voltage of the lifepo being close to the Vf of tbe xml made it ideal. I thought it was powerful enough not to sag much but Ilack the experience that you have. I've been trying to think of a build that would suit lifepo so I could try a safer chemistry. I'll have to wait for a qualified builder to come along. I figured DD would be easier to build but that imr cells would toast the led.

Sent from my Droid using Tapatalk

 

[ti-force]

S12 OTF measurements can be found here.

 

[Yoda4561]

7

Thanks for the explanation. I thought that th voltage of the lifepo being close to the Vf of tbe xml made it ideal. I thought it was powerful enough not to sag much but Ilack the experience that you have. I've been trying to think of a build that would suit lifepo so I could try a safer chemistry. I'll have to wait for a qualified builder to come along. I figured DD would be easier to build but that imr cells would toast the led.

Sent from my Droid using Tapatalk

Just going by the Cree vf:current spec sheet and assuming 3.0-3.2v under load for the cells, you "should" run between 280-600 lumens depending on the particular LED's exact vf and battery condition. This puts the drive current between 1amp (3.0v) and 2 amps (3.2v) As long as you're not looking for max output, it should be a nearly ideal direct drive battery, since there's almost no way it should overdrive the LED.

 

[bigchelis]

Its been a while, but to revive this sticky thread I have a cool light for testing. I will post the number on Saturday.

The VaraPower1000 by Lamdalights.com = whats different about it?????

• The Hosts is a 3C Mag now. No more D size.
• The heatsink is pure copper
• The emitter is a XM-L 6T
• Direct Copper to LED reflow
• Driven hard...reall hard
• 60~80K lux at 1M is predicted
• 1000 OTF at least is expected
• Uses Deep Rebel Reflector

There is a 2C Mag VaraPower 2C in testing and specs are not finalized, but I hope to test that one too. Please note: these are usually sold out quick, so if you want one you have to check his site regularly.

There is a 2C Mag XM-L that will not have the 33 VaraPower modes, but will offer 3A to the LED from 1.2V~2.8V; I have been waiting since late last year for this one too. I have my paypal ready for this one for sure.

 

[recDNA]

People in this thread know as much about flashlight output as anyone anywhere so could someone answer this question? How is it possible that the thrunite catapult v3 is rated under 770 ANSI lumens even though the emitter is driven at 3.5 amps yet the fenix tk35 is rated over 800 ANSI lumens though very likely driven at lower amperage? Is ANSI bogus or has Fenix discovered something nobody else knows?

Sent from my Droid using Tapatalk

 

[bigchelis]

Hi all,

The Lamdalights 3C VaraPower had a loose ground connection. Which is fine since they are not for sale yet, till Kevin feels they are perfect.

Instead, I got the 2C XM-L low voltage input. This one is designed to run 3A at the LED with 1V~3V inputs and is fully regulated.

I did a quick test and get 620 OTF lumens on a fresh pair of NiMH C cells by Tenergy with 4.25A at the tailcap.

I will be doing alot of testing on this one. So, stay tunned.

• 3A at the LED
• XM-L 6T
• Copper heatsink
• Copper to LED bonding
• Full power from 1 or 2 NiMH C cells.
• Can use Alkalines at reduced output.

 

[recDNA]

I don't know if anybody is still .doing tests these days but I can think of one I'd LOVE to see. I'm very curious if running a finned P60 head makes any real difference in the output of a high powered drop in in a room at room temperature. find it hard to believe that the increase in surface area is great enough to dissipate enough heat to matter without adding a fan or cold temp to help carry the heat away. Still air is a very poor conductor of heat. I don't like the looks or feel of heavily finned heads but if they really make a significant difference that is translated into the output curve I'd love to see it.

 

[bigchelis]

I don't know if anybody is still .doing tests these days but I can think of one I'd LOVE to see. I'm very curious if running a finned P60 head makes any real difference in the output of a high powered drop in in a room at room temperature. find it hard to believe that the increase in surface area is great enough to dissipate enough heat to matter without adding a fan or cold temp to help carry the heat away. Still air is a very poor conductor of heat. I don't like the looks or feel of heavily finned heads but if they really make a significant difference that is translated into the output curve I'd love to see it.

I honestly think those type of bezels provide Zero benefit to maintain lumen output. I think adding copper to the base where the LED sits is where all the attention should be put. Like the new coppper P60 hosts from Fivemega. More not less!!!

Maybe they do get rid of the heat, but that means the LED has to get hot for the bezel to work right? I rather keep it cool or spread the heat over more material.

If anybody has a bezel like this send it in please. I can do a before and after test to prove or disprove if they actually work.

bigC

 

[ma_sha1]

Agree with Big C, Fins are for "marketing", may be better looking but they are useless for heat dissipation w/o air flow, unless one wave it fast back & forth . Useful for Bikelight as it's in fast moving air.

One of the old Thermo profile thread had busted the myth on finning of mag headThe finned one actually performed worth as it lost mass)
http://www.candlepowerforums.com/vb/showthread.php?115232-Mag-Thermal-Profile

P60 is no different, best is to hold it where it's hot & let it transfer heat away via. blood circulation, in such case a smooth surface provide larger contact area, thus faster heat transfer.

 

[bigchelis]

Now; regarding the Low Voltage input 2C XM-L by Lamdalight.com. This light has almost flat output and with such low voltage input I think its a record.

4.25A at tailcap at 1sec and 5.20A at tailcap at 10min. I used 2 NiMH C cells by Tenergy. I still need the alkaline OTF numbers and 1 NiMH C cell test too.

The 2C XM-L should be able to operate at similar output with just 1 NiMH C cell, we will have to see.

639.3 1 sec
639.3 30 sec
638.1 1 min
650.0 2 min
640.5 3 min
633.3 4 min
629.8 5 min
629.8 6 min
632.1 7 min
632.1 8 min
629.8 9 min
632.1 10 min

 

[ti-force]

I'm actually surprised that the output isn't higher, since the light has a larger head/reflector and is being driven at a higher drive current. My Quark Turbo made 600 OTF at 2.1A drive current.:thinking:

 

[bigchelis]

I'm actually surprised that the output isn't higher, since the light has a larger head/reflector and is being driven at a higher drive current. My Quark Turbo made 600 OTF at 2.1A drive current.:thinking:

I think these type of drivers are usually less efficient.

I have a PCC built 1D Mag with 2.8A shinning beam driver and it nets me only 560~580 OTF lumens with Deep Reflector.

Then I tested another 3D Mag by PCC that is powered by the same 2.8a driver and deep reflector and its 680ish OTF. Both builds pull 2.88ish at the tailcap with identical voltage input.


Maybe some of the light escapes from the base and doesnt make it out the front since these builds use Deep rebel reflector.


bigC

 

[recDNA]

I honestly think those type of bezels provide Zero benefit to maintain lumen output. I think adding copper to the base where the LED sits is where all the attention should be put. Like the new coppper P60 hosts from Fivemega. More not less!!!

Maybe they do get rid of the heat, but that means the LED has to get hot for the bezel to work right? I rather keep it cool or spread the heat over more material.

If anybody has a bezel like this send it in please. I can do a before and after test to prove or disprove if they actually work.

bigC

Thanks for the explanation!

 

[recDNA]

I think these type of drivers are usually less efficient.

I have a PCC built 1D Mag with 2.8A shinning beam driver and it nets me only 560~580 OTF lumens with Deep Reflector.

Then I tested another 3D Mag by PCC that is powered by the same 2.8a driver and deep reflector and its 680ish OTF. Both builds pull 2.88ish at the tailcap with identical voltage input.


Maybe some of the light escapes from the base and doesnt make it out the front since these builds use Deep rebel reflector.


bigC

It's really nice to have such large output with safer battery chemistry. I wish more manufacturers would build lights capable of over 600 lumens with no fear of explosions (RCR) or poison gas venting (IMR).

 

[bigchelis]

It's really nice to have such large output with safer battery chemistry. I wish more manufacturers would build lights capable of over 600 lumens with no fear of explosions (RCR) or poison gas venting (IMR).

This is the only light in existence or to ever exist that I know of, that will allows 10 plus watts of power from just 1V~3V input. I have seen other drivers that can generate 5 watts off .8V~3V, but thats about it.

So, for me to see over 10watts off just 2 or 1 NiMH C cell I am excited, but nature of the beast makes it less efficient or I been told.

bigC

 

[Nitroz]

This is the only light in existence or to ever exist that I know of, that will allows 10 plus watts of power from just 1V~3V input. I have seen other drivers that can generate 5 watts off .8V~3V, but thats about it.

So, for me to see over 10watts off just 2 or 1 NiMH C cell I am excited, but nature of the beast makes it less efficient or I been told.

bigC

I can imagine that the cells are not to happy, but true flashaholics tend to abuse things.

I am amazed to see this so soon. I thought it would be a little longer before 1 or 2 NiMh would be pushing this kind of output.

I wish I had one of those boost boards. I would love to run a 1 D Saft cutdown with a rear clicky.

 

[elho]

I am giving up the Spheres. It just eats up too much of my time and I neglected other important things in my life.

Aww, sad news. But I do understand the reasons for it very well (just look how long it has taken me to find the time to pick up reading this thread again).

Many, many thanks for the massive amounts of work and time that you invested into providing us with the numbers we were dying for!