brightnorm
Flashaholic
- Joined
- Oct 13, 2001
- Messages
- 7,160
Unless SF has come up with an ultra efficient LED how is this possible, and at what cost in runtime?
Brightnorm
Brightnorm
100 lumens from the U2 - really?
- Thread starter brightnorm
- Start date
Help Support Candle Power Flashlight Forum
My guess would be that if you get 65 lumens from other 2x123 5W Surefires for 60 minutes, you are probably only going to get 15-20 at 100 lumens, maybe even less. I don't think it is designed for walking around at the 100 lumens mark. Just blinding and finding (perps).
gessner17
Enlightened
- Joined
- Dec 29, 2003
- Messages
- 499
I thought 5 watt LED's were rated up to 120 Lumens....???
The 5 watt LEDs are rated at 120 lumens. But my guess is that surefire tends to underdrive them in order to prevent overheating the LED on a smaller handheld light, which can't dissipate the heat like the bigger 3D size lights. If the 5 Watt is driven fully, it will get hot very fast.
My only guess is maybe Surefire is using X-binned 5 Watters...at least, I am hoping they are.....
My only guess is maybe Surefire is using X-binned 5 Watters...at least, I am hoping they are.....
well ...
typically SF measures actual light-output of the flashlight, and that would mean there are some losses and the LuxV has to put out considerably more than just 100lm.
But then there are X-bins, and those go well over 200 lm tops!
/ubbthreads/images/graemlins/grin.gif /ubbthreads/images/graemlins/grin.gif
/ubbthreads/images/graemlins/drool.gif drooling like a madman over this light!
bernhard
typically SF measures actual light-output of the flashlight, and that would mean there are some losses and the LuxV has to put out considerably more than just 100lm.
But then there are X-bins, and those go well over 200 lm tops!
/ubbthreads/images/graemlins/grin.gif /ubbthreads/images/graemlins/grin.gif
/ubbthreads/images/graemlins/drool.gif drooling like a madman over this light!
bernhard
I get the feeling that Lumileds lumens ratings are inflated.
gadget_lover
Flashaholic
[ QUOTE ]
Pi_is_blue said:
I get the feeling that Lumileds lumens ratings are inflated.
[/ QUOTE ]
They are probably exactly right. They use very broad ranges on purpose. The problem, of course, is in managing the photons so that most go where you want them. The lumileds measurements would be without reflectors or lenses to cause loss of light. There's a lot of light coming out of the LED towards the sides.
Your household 100 watt light bulb is probably putting out close to 1000 lumens, but how much is being blocked by the lamp shade or absorbed by a reflector that's painted white? How much is illuminating the wall behind the lamp?
So I guess what I'm saying is that the optics greatly impact how much light gets out of the flashlight. Losses can be much greater than 30 percent.
Daniel
Pi_is_blue said:
I get the feeling that Lumileds lumens ratings are inflated.
[/ QUOTE ]
They are probably exactly right. They use very broad ranges on purpose. The problem, of course, is in managing the photons so that most go where you want them. The lumileds measurements would be without reflectors or lenses to cause loss of light. There's a lot of light coming out of the LED towards the sides.
Your household 100 watt light bulb is probably putting out close to 1000 lumens, but how much is being blocked by the lamp shade or absorbed by a reflector that's painted white? How much is illuminating the wall behind the lamp?
So I guess what I'm saying is that the optics greatly impact how much light gets out of the flashlight. Losses can be much greater than 30 percent.
Daniel
ubermensch
Enlightened
NX-05 optics are supposed to be ~90% efficient--reflectors are supposed to be even better (says CPFer HDS manufacturing).
And L4s are not particularly underdriven, 660mA (L4) vs. 700mA (lumileds spec).
That would be a loss of about 6 lumens (5%). I doubt the pyrex lens with anti-reflective coating is sucking up many lumens. So, all told, a loss of about 15%. Where is 65 lumens coming from, if it should be 100 lumens? It just doesn't compute. Even if you go by the MINIMUM luminous flux, (87 lumens), to get 65 lumens out the front would be a ~20% loss of light. I'm confused.
And L4s are not particularly underdriven, 660mA (L4) vs. 700mA (lumileds spec).
That would be a loss of about 6 lumens (5%). I doubt the pyrex lens with anti-reflective coating is sucking up many lumens. So, all told, a loss of about 15%. Where is 65 lumens coming from, if it should be 100 lumens? It just doesn't compute. Even if you go by the MINIMUM luminous flux, (87 lumens), to get 65 lumens out the front would be a ~20% loss of light. I'm confused.
Hasn't this been discussed several times before? /ubbthreads/images/graemlins/rant.gif /ubbthreads/images/graemlins/tongue.gif
As already mentioned there are losses because of the reflector and lens. Obviously these losses are not high enough to let the output from the Luxeon drop from e.g. 120 Lumens to only 65 or something like that. This decrese is caused by the heating of the DIE. As you can see in this chart:
At 80°C (which is realistic) the output has already dropped by 20%. The nubers specified by Lumileds refer to 20°C junction temperature.
HTH
As already mentioned there are losses because of the reflector and lens. Obviously these losses are not high enough to let the output from the Luxeon drop from e.g. 120 Lumens to only 65 or something like that. This decrese is caused by the heating of the DIE. As you can see in this chart:
At 80°C (which is realistic) the output has already dropped by 20%. The nubers specified by Lumileds refer to 20°C junction temperature.
HTH
ubermensch
Enlightened
Great work Mister T, that was what I was going to say next. This could easily be test by Quickbeam--put an L4 in the freezer, then test it in the lightbox. But doesn't it seem like if this was true, that if your light was real cold at first, you would notice a big drop in output as it heats up?
But this begs the question, how is the U2 able to circumvent these problems to get its higher output?
Also, what is the point of giving a lumen rating that is impossible to achieve in real life?
But this begs the question, how is the U2 able to circumvent these problems to get its higher output?
Also, what is the point of giving a lumen rating that is impossible to achieve in real life?
JonSidneyB
Flashlight Enthusiast
For X bin Surefires, what would I do? Hmmmm
Thinking.... I would, I would, I would....Shoot Roth if thats what it took.
/ubbthreads/images/graemlins/icon15.gif
Thinking.... I would, I would, I would....Shoot Roth if thats what it took.
/ubbthreads/images/graemlins/icon15.gif
[ QUOTE ]
ubermensch said:
But doesn't it seem like if this was true, that if your light was real cold at first, you would notice a big drop in output as it heats up?
[/ QUOTE ]
I think it depends on how fast he output decreases, because that makes it easier to recognise. Maybe you will only notice a change of the output with a luxmeter since the cold battery won't supply as much current as before.
[ QUOTE ]
But this begs the question, how is the U2 able to circumvent these problems to get its higher output?
[/ QUOTE ]
Unless they don't come up with something really spectacular like LN2 cooling /ubbthreads/images/graemlins/ohgeez.gif a high binned (and enough current) Luxeon will do.
ubermensch said:
But doesn't it seem like if this was true, that if your light was real cold at first, you would notice a big drop in output as it heats up?
[/ QUOTE ]
I think it depends on how fast he output decreases, because that makes it easier to recognise. Maybe you will only notice a change of the output with a luxmeter since the cold battery won't supply as much current as before.
[ QUOTE ]
But this begs the question, how is the U2 able to circumvent these problems to get its higher output?
[/ QUOTE ]
Unless they don't come up with something really spectacular like LN2 cooling /ubbthreads/images/graemlins/ohgeez.gif a high binned (and enough current) Luxeon will do.
oldgrandpajack
Enlightened
- Joined
- Mar 15, 2003
- Messages
- 931
IMHO, the L4, L2 on HIGH, and the U2's HIGHEST setting are designed for intermittent use. They can be left on until they go out of regulation, of course. But, there isn't enough mass, and or surface area, to deal with the heat properly, if left on for any length of time. I fear for the longevity of the LED and the circuitry, if they aren't used intermittently.
If only a test could be set up to run them for a minute and let them cool for 2 minutes, repeating until they go out of regulation. Then generate a graph, with the cooling periods taken out.
oldgrandpajack
If only a test could be set up to run them for a minute and let them cool for 2 minutes, repeating until they go out of regulation. Then generate a graph, with the cooling periods taken out.
oldgrandpajack
luxlover
Banned
[ QUOTE ]
If only a test could be set up to run them for a minute and let them cool for 2 minutes, repeating until they go out of regulation. Then generate a graph, with the cooling periods taken out.
[/ QUOTE ]
I like your thinking, Sir. Everyone is testing a light, using methods that do not emulate it's actual use. I would like to see a consensus of members, writing posts with their idea of what a practical runtime test should entail.
First to be determined, is the "average" period of time that an "average" user has the light turned on. Note that when it is turned off, the battery gets a chance to "rest", and during this time it will recharge itself to a small degree. This phenomenon is why a light that has weak batteries and is turned off and allowed to rest, will be brighter for a few seconds than when it was turned off.
Second to be determined is the real time that a light is turned off between uses. From my readings about lights over the years, I found that if you run a light continuously, clock it until the batteries are dead, and compare this number to the sum of all of the on cycles when the light gets to rest, then the sum of all of these on cycles will exceed the runtime of the light that is left on continuously. This means that it is a good practice to allow a battery to rest. It is also beneficial for any lamp, not only an LED, to not get too hot from continuous use, because this will do irreversible damage and shorten it's lifespan. At all costs, we must prevent our lights from getting overheated. Even a 30 second rest during a long run, can make a difference in the LED's lifespan, when the light's body is getting uncomfortably hot.
If we can standardize a test method to approximate actual use, then we can accurately compare one light with the same number of batteries and lamp wattage, to another one. We can not become smart shoppers, if we are all over the place in how we evaluate our lights.
Does anybody have any idea, how to implement a survey that will yield real world on and off times? I look forward to the start of this project. I hope that many members will participate.
If only a test could be set up to run them for a minute and let them cool for 2 minutes, repeating until they go out of regulation. Then generate a graph, with the cooling periods taken out.
[/ QUOTE ]
I like your thinking, Sir. Everyone is testing a light, using methods that do not emulate it's actual use. I would like to see a consensus of members, writing posts with their idea of what a practical runtime test should entail.
First to be determined, is the "average" period of time that an "average" user has the light turned on. Note that when it is turned off, the battery gets a chance to "rest", and during this time it will recharge itself to a small degree. This phenomenon is why a light that has weak batteries and is turned off and allowed to rest, will be brighter for a few seconds than when it was turned off.
Second to be determined is the real time that a light is turned off between uses. From my readings about lights over the years, I found that if you run a light continuously, clock it until the batteries are dead, and compare this number to the sum of all of the on cycles when the light gets to rest, then the sum of all of these on cycles will exceed the runtime of the light that is left on continuously. This means that it is a good practice to allow a battery to rest. It is also beneficial for any lamp, not only an LED, to not get too hot from continuous use, because this will do irreversible damage and shorten it's lifespan. At all costs, we must prevent our lights from getting overheated. Even a 30 second rest during a long run, can make a difference in the LED's lifespan, when the light's body is getting uncomfortably hot.
If we can standardize a test method to approximate actual use, then we can accurately compare one light with the same number of batteries and lamp wattage, to another one. We can not become smart shoppers, if we are all over the place in how we evaluate our lights.
Does anybody have any idea, how to implement a survey that will yield real world on and off times? I look forward to the start of this project. I hope that many members will participate.
ubermensch
Enlightened
I found these looking at the Flashlight Fan webpage today.
The L4 shows about a 20% drop in lux, Mister T was right on target. The L5 and L6 both show a 10% drop in lux, so bigger is better obviously. But what is vexing is that even with a 20% reduction from overheating, & 15% from light loss and underdriving, that should still produce ~80 lumens. Plus, I thought that they used only V-bin or higher in their lights, which would mean one should get substantially more than 80 lumens.
I love thinking about this stuff.
To complicate things further--the L5 is rated at 65 lumens as well, but it has better heatsinking, shouldn't it be rated higher? Quickbeam's lightbox shows 58 for the L4 and 60 for the L5, a 3% difference???
The L4 shows about a 20% drop in lux, Mister T was right on target. The L5 and L6 both show a 10% drop in lux, so bigger is better obviously. But what is vexing is that even with a 20% reduction from overheating, & 15% from light loss and underdriving, that should still produce ~80 lumens. Plus, I thought that they used only V-bin or higher in their lights, which would mean one should get substantially more than 80 lumens.
I love thinking about this stuff.
To complicate things further--the L5 is rated at 65 lumens as well, but it has better heatsinking, shouldn't it be rated higher? Quickbeam's lightbox shows 58 for the L4 and 60 for the L5, a 3% difference???
