Longer runtime on low beam for Cree and Seoul LED lights than on the Luxeon III LEDs?

[FaithinGod]

Dear CPF Members,

Do any of you have any runtime graphs to verify if the runtime for the current Cree and Seoul LED lights have significantly greater runtimes than the current Luxeon III LED lights when driven at low levels of light output (such as when comparing the same flashlight with the Cree LED in it to the same flashlight with the Luxeon III LED in it, or comparing the same flashlight with the Seoul LED in it to the same flashlight with the Luxeon III LED in it, at, for example, 2 to 3 Lumen output)?

I have seen several charts verifying that the Cree and Seoul LED lights do have greater (about double) runtime at maximum levels of light output, but one engineer who I spoke to said he did not believe the low level output of the Cree or Seoul LED lights was significantly longer at the same lower level of light output as the Luxeon LED lights. But I did see a runtime chart on a McGizmo Seoul LED light which was run at a low light output level (about 2-3 Lumen, if I recall) and it ran about 3 days.

My assumption would be that if the output of the Cree and Seoul are twice that of the Luxeon III when at maximum output while also using the same amount of energy, then the low beam output of the Cree or Seoul should also be twice the runtime of the Luxeon III when driven at the same low power level.

If any of you are able to offer help in this regard I would be most grateful.

Thank you sincerely in advance.

FaithinGod

 

[2xTrinity]

I have seen several charts verifying that the Cree and Seoul LED lights do have greater (about double) runtime at maximum levels of light output, but one engineer who I spoke to said he did not believe the low level output of the Cree or Seoul LED lights was significantly longer at the same lower level of light output as the Luxeon LED lights. But I did see a runtime chart on a McGizmo Seoul LED light which was run at a low light output level (about 2-3 Lumen, if I recall) and it ran about 3 days.

This depends on how "low" you're talking about. In just about any situation, the answer is yes, but the difference in efficiency is not going to always be 2 to 1.

The problem with a lot of these LEDs is that because the Cree is so efficient, it takes a lower current to even reach a 2-3 lumen low level. Since below a certain point (IIRC around 80mA current) efficiency will start to drop off, and tint will significantly shift. This means that while it will be trivial to swap your emitter and get a "low" level with the same runtime and double the brightness, dimming the Cree to reach the old low level will take a bit more work.

To get dimmer, it will be necessary to use a combination of current limiting (with the current set right at the efficiency "peak") and pulse-width modulation (flashing the light off and on rapidly). At low levels like that, a lot of the power at super low levels like that is eaten up by driver inefficiency (which would affect both the Lux and Cree about equally), not by the LED, so doubling LED efficiency won't double overall efficiency. As the drive level increases, so will the difference in runtime between the Luxeon and the Cree. The biggest difference should be when the Luxeon at 1A max, and the Cree is driven comfortably at around 450mA more than 2 to 1 difference in runtime (the harder you push a battery, the more power is lost to internal resistance).

A lot of high-end flashlights use this hybrid method of dimming (current limiting combined with pulse-width modulation). If you used this technique I just described both with a Cree, and with a Luxeon, the Cree would win. That's because its peak efficiency is higher than the Luxeon's peak efficiency, so it will appear just as bright with a shorter PWM duty cycle ("on-time" of the pulses)

 

[yellow]

what kind of "engineer" would believe that the double efficient Cree/SSC do run longer on high but NOT on low output?
:thinking:
if anything is more efficient, then its more efficient at any time, right?

Problem to proof this:
1. with an emitter-swap light, there is (roughly) the same current as before and thus no increased runtime (but more brightness),
2. the hell of a light when it comes to Led and mass production (the Fenix) has a WAY TOO BRIGHT low output and thus, again, runtime on low is not too long (compared to, say, a homemade mod with a real low)

 

[FaithinGod]

Re: Longer runtime on low beam for Cree and Seoul LED lights than on the Luxeon III LEDs?

Thank you 2xTrinity and yellow for your replies.

I remain hopeful that someone is able to offer a Runtime Chart (or just personal visual witness, even though the human eye can't see that accurately), as inquired above, to provide a conclusion to the question, specifically in regards to two identical Lights but one with the Luxeon III in it and one with a Cree or Seoul in it and both driven at the same low electrical power level (such as 50mA for example, keeping it simple, as with a resisitor rather than with Pulse Width Modulation or Current Regulation or the Hybrid of the latter two).

Sincerely,

FaithinGod

 

[yellow]

Re: Longer runtime on low beam for Cree and Seoul LED lights than on the Luxeon III LEDs?

(such as 50mA for example,

that way of entering the solution to the "problem", is the problem:
at 50 mA there is NO difference in runtime at all, why should? Its all in the battery then.
but
the Cree/SSC light will be double bright

--> means You will only need 25 mA for the same brightness and thus runtime will be double with the more efficient emitters

 

[Long John]

I'll throw antoher important factor into the game, the vf of the Led.

Low vf = less consumption.
I.e. swapping a low vf LuxIII T--H to a high vf SSC achieves possibly shorter runtimes at the same driving level.

Best regards

____
Tom

 

[2xTrinity]

I'll throw antoher important factor into the game, the vf of the Led.

Low vf = less consumption.
I.e. swapping a low vf LuxIII T--H to a high vf SSC achieves possibly shorter runtimes at the same driving level.

Best regards

____
Tom

While a Seoul may have a higher Vf, it will still be producing more light than a Luxeon for the same amount of power (Vf * If), as Seoul/Cree has been shown to be twice as efficient in terms of lumens per watt -- and both are about the same as each other. Seoul emits more light at 1A, but Cree usually has lower forward voltage so is actually not being driven as hard (less power consumption). While the point about Vf might matter if you simply swap a new LED into an older light that is current-regulated, you must also remeber that you'll be getting a lot the light at the same time. If you drop a Seoul with a higher Vf into a light that is voltage regulated or direct-driven, you'll end up with less light and longer runtime, as the driver won't push nearly as much current through the higher Vf LED.

If comparing a light designed to use Seoul/Cree at low levels to one designed to use Luxeon that way, the Seoul/Cree will win for runtime. The biggest problem is that for the sake of convenience (less complicated driver circuits) most purpose-built Cree and Seoul lights, including the upcoming ones from Surefire, have chosen "low" levels that are 3-5 times as bright as older "low" levels.

 

[nerdgineer]

A mathematician (in the mathematician joke sense) might say that at zero current, their run times are the same....

However, I think most engineers would observe that the lumens/watt rating of the Crees/Seouls are about double the Lux's across the board, from about 5 mA on up...

 

[FaithinGod]

Re: Longer runtime on low beam for Cree and Seoul LED lights than on the Luxeon III LEDs?

Thank you Long John; nerdgineer, and thank you again yellow and 2xTrinity for you replies.

I apologize for my error, I meant to say 50 Candlepower rather than 50 mA in my inquiry in post #4, with the hopes that the Seoul or Cree would be much longer (if not double) in runtime when giving out (for example) 50 Candlepower compared to a Luxeon III also giving out 50 Candlepower.

Sincerely,

FaithinGod

 

[RustyKnee]

that way of entering the solution to the "problem", is the problem:
at 50 mA there is NO difference in runtime at all, why should? Its all in the battery then.
but
the Cree/SSC light will be double bright

--> means You will only need 25 mA for the same brightness and thus runtime will be double with the more efficient emitters

hi yellow,

read long johns post. same drive current doesn't mean same run time. if for a given drive current led A has double the Vf of LED B, led A will have half the burn time.

Stu

 

[cratz2]

I almost feel like I've taken crazy pills, having read through this post.

My knowledge of physics and chemistry stops at my second year in college, but it seems that comparing two identical boards powered by two identical cells druving two LEDs with identical Vf, with the same current pull from the cell, the longer said board/cell/LED combo would run the same length.

In the real world, when comparing as similar as possible LED lights where one has a Cree and one has a Lux III, generally the exact same board is being used because of cost factors and the desire to be as quite as possible to market... This means that rather than the Cree-based light having a longer runtime on low, it will have 'about' the same runtime, but be 60%-100% brighter than the Lux III version of the same light.

The Liteflux LF-2 and the VB-16 come to mind as lights that has a a truly LOW available low setting and should allow for very long runtimes. Also, any light modded with a Flupic allows the user to have a pretty darn low low setting.

 

[RustyKnee]

cratz2 what you says makes perfect sense.

Back to the orignal question again, looking at the curves in the datasheets I can't see a reaon why a cree/seoul wouldn't last significantly longer than an lux 3 at the same brightness. The datasheets don't use candlepower but lumens, so I didn't so any rough calcs specific to the question.

Stu