I would expect an XP-G to be brighter at 2A than at 1.2A - until the magic black smoke escapes. That's fairly typical, though not guaranteed. With some googling you can find examples from people who have done this and posted the results. See, e.g., this thread on CPF: http://www.candlepowerforums.com/vb/showthread.php?308569-No-real-benefits-to-overdriving-XML
The OP claims no benefit from overdriving his XM-L, but others have photos showing the benefit of overdriving both XM-Ls and XP-Gs.
Of course, this is somewhat tangential to the original question, which asked about XM-L vs. XP-G heat at a particular current level. All we need to answer that question is the knowledge that the XM-L produces more light and has Vf that is no higher.
One other thing I forgot to mention: if trooplewis meant 2A pulled from the battery, then the amount of power delivered to both LEDs would be the same, but since the XM-L has a lower Vf, the current to the LED would be higher since voltage x current = power. If we estimate that the XM-L's Vf is 10% lower, then the current would be 10% higher, and going back to those Cree data sheets, at 1.5A, an XP-G would produce about 460 lm, while an XM-L (at 1.65A) would produce about 590 lm. So the heat produced by the XM-L would be only marginally lower, but the output would be significantly higher.
I don't know what happens to Cree LEDs when driven beyond their ratings, but for sure the output flux curve won't change direction for the better. Therefore, as you overdrive one further and further, you'll get proportionately more heat than light.
ronparr, you said at the end of your last email: "and has Vf that is no higher". Are you implying that the XM-L's Vf may not be lower than an XP-G's Vf in some cases? Please explain. According to Cree's data sheets, the XM-L's Vf is definitely lower, from 100 mA and up. I guess if you compare a lower-flux-bin XM-L to a higher-flux-bin XP-G, then it could be true. I'm asking because I've seen other threads where people say that an XP-G may outperform an XM-L at lower currents, but I don't see how from the data sheets, and if I'm wrong, I'd like to know why. This is just total output in lumens I'm talking about, not brightness or throw, etc.
Trooplewis, if in your OP you meant 2A through the LED, the answer is simple: the XM-L's lower Vf, plugged into the formula power = voltage x current, would result in the power dissipated in the XM-L being at least 10% lower. Power results in light output and heat, so the heat produced by the XM-L will also be at least 10% less. This is not including any of the information provided by others. I didn't know about the die-size effects, so it sounds like the XM-L would be even more than 10% cooler.
"I would expect an XP-G to be brighter at 2A than at 1.2A - until the magic black smoke escapes. That's fairly typical, though not guaranteed. With some googling you can find examples from people who have done this and posted the results. See, e.g., this thread on CPF: http://www.candlepowerforums.com/vb/showthread.php?308569-No-real-benefits-to-overdriving-XML"
I totally agree with the OP because I got the same results with any emitter tested, being XR-E or XM-L , also my tests were performed with a heatsink which emulate a good host since you can't put a fan in a torch...sorry, I know that the naysayers , us, are unpopular...
I'd be fairly certain that electric heaters are not 100% efficient. I've not seen one for a good long while, but don't the elements tend to glow? Which would waste heat as light.
There would also be energy wasted as EMF on the elements and presumably also induced in any surrounding metals.
Any light produced will eventually become heat in your house - unless you leave your shades open and some escapes out your windows. Some of the EMF might pass through your walls before it is all converted to heat, so perhaps we can agree on "nearly 100%"?
So, you think those other folks are lying and faking their photos?
[/quote]
Yes, because at current beyond specs even when an increase is noted, say an XM-L at 4A ( the first 30 seconds or less ) the increase is very low , from 850 at 2,8 to 900L at 4A...you can't differentiate 250L from 300 L side by side, to notice something you must pass ie. 200L to 300L and is not a big difference , just a little to notice it.
Hmmm... Well, call me a sucker or an optimist, but I'm more inclined to think that manufacturing variability explains variation in tolerance to high current than to assume that all of those people are liars who faked their photos for (seemingly) no reason other than the entertainment value of deceit.
What are you implying here? Are you saying the sample size is too small? Are you saying it's small enough that it could all be part of a hoax? Really??? Before you read too much into the 2009 exif tags in one pair of photos, I'll point out that the date in the exif (6/19/2009) precedes the press release announcing the existence of that model camera (7/22/2009) , so it's clear that the exif date is not the date when the photos were taken because the camera was not for sale at that date.
Incorrect dates like this are typical when people run down the batteries on their cameras to point where the clock no longer maintains the time. The clock resets to the factory default date when this happens and users sometimes forget or don't bother to set the correct data.
Yeah - It's ambiguous because we don't know what the driver is doing. If we treat the battery voltage and current as constant, then we need to make an assumption about what is happening with the extra energy that is above Vf for the XM-L. Is the DC-DC converter in the driver providing more current at the target Vf, or is it just providing too much voltage, or is dissipating the extra energy as heat in the driver? I would guess that it's just overdriving the LED a bit, but so long as the drain on the battery is constant and we see more light coming out of XM-L, we know that we're getting more heat from the XP-G (or some combination of the LED and driver).
I didn't mean to imply this. I was just stating a minimal condition: The same current and voltage that is no higher means that the XM-L is consuming no more power than the XP-G. If it is consuming no more power and producing more light, then we know it must also be producing less heat by the first law of thermodynamics.
I don't think we disagree on anything here. I'm just pointing out that basic intro physics concepts, P=IV and the first law of thermodynamics, are sufficient to reason through this.
I think we are splitting hairs by saying the heat produced is the same, just more of it escapes the body as light... That is like saying all gas engines get the same power out of a gallon of gas, but with some, more of it escapes through the tailpipe. Yes. So what? I want to know how much us usable in powering the car.
My original question was meant to go directly to the temperature of the body of the flashlight.
If you use the same light, and seat the pill identically, will the flashlight temperature be the same after x number of minutes with both emitters pulling 2a, or will one get hotter than the other? This was not meant to be a philosophical discussion about the laws of thermodynamics, but go there if you must.
Real applications (like an electric heater) have more complications, like the efficiency or the transformer and the radiation of the heat, that is why you will see 20+different varieties of heaters at Home Depot...they are not simple, X=Y formulas.
Be careful here. Heat and infrared radiation are not the same thing. The difference betwen the XM-L and XP-G in this scenario is not that the less efficient LED is producing more IR and less visible light. The less efficient LED is transferring more energy from the battery into the kinetic energy of the molecules in the materials surrounding the LED. The emission of IR radiation from these materials is a side effect of the increased heat of these materials.
If both consume the same power and one produces more light, the one that produces more light must produce less heat. This isn't a philosophical thing; it's just a straightforward application of the one of the most basic laws of physics.
The efficiency of electric heaters is splendidly uncomplicated because it is always 100%. There are different types that vary in how much power they consume, what extra features they have (such as fans), and what how they direct the energy they do produce, but they are all 100% efficient aside from trivial amounts of energy used on fans or lights. This should be obvious based upon first principles, but if you don't believe me or don't believe the science, ask your power company or do some googling. Here are few things you might come across:
Look one of the links of the thread mentioned above as 'testimonial' of successful overdrive an XM-.L at 5A...
This guy beside getting 'extraordinaires' leds might be magician. 3 hours run-time at 16W with a 37W pack ( 7.4x5=37...
Total weight light and batteries 300 grs , astounding heat management to run an emitter at full with that mass...200 grs pack 100 grs body...
I like the bottom note, approximate... must be greatly exaggerate...
"DESTROYER 1200
This is the light to get excited about. We have utilised the latest Cree LED technology to offer 1200* Scorching Lumens! If you haven’t seen one of these in action then prepare to be blown away! The Destroyer blows our tried and trusted Illuminator to the weeds, the difference is like night and day. Using our 5000 mAh Li Ion batteries, run time is impressive for such a powerful light*. Full, half and strobe functions. Each Pack comes with Head, Battery Pack, Helmet and bar mounts, 230V charger and Extension lead.
Features:
* The LED has a rated life of over 50,000 hours and like all LED's is totally shock proof
* Light output is digitally controlled to produce constant light output for the entire charge of the battery
* 3 output modes: High 16W, 1200 Lumens* (3+ hours*), Low 8W, 600 Lumens* (8 hours*) and Strobe
* The 40 x 40 x 70mm - 8.2V Li-ion battery pack weighs just 200 grams, recharges in only 4 hrs and is protected to prevent overcharging or over-discharging. Attaches to your handlebar, stem or frame with the included pouch.
The entire assembly (light and battery) weighs in at just over 300 grams.
* All lumen ratings and run times are approximate only."
None of this is particularly strange. An 18650 is ~45 grams, 18mm in diameter and 65mm in length. The battery pack probably has 4 of them stacked 2x2, which would be 36x36x65mm for the batteries. The stated dimensions of the battery pack are 40x40x70, which leaves a few mm for the plastic case. The weight of 4x18560 would be 180 grams, which leaves about 20 grams for the plastic case given the stated 200 gram weight for the battery pack. All of this seems to fit pretty well.
This would be enough to provide ~9600 mAh at the standard voltage, assuming 2400 mAh per cell. It looks like they have the battery pack set up to provide 8.4 v, so we can cut that in half to 4800 mAh if we think of it as a 8.4v battery. Let's say the battery pack is providing 2A @8.4V, or 16.8W and that the LED sees 15W with the difference going to driver inefficiency.
Based on the assumption 2x2400 mAh 18650 cells, the batteries would be able to provide 2 A for 4.8/2 = 2.4 hours. In practice, the batteries will probably provide lower current as they start to get depleted, so the rate of draw will probably slow a bit. 3 hours to 50% seems entirely plausible. For example, my 2200 mAh 18650 cells put out over 2.2 A when fresh and driving one of my XM-Ls on high, but wind up lasting about an hour because they're putting out less than 2200 mAh at the end of the hour.
Of course, there's a bit of fudge in here. It looks like they're assuming the off-the-charger 4.2 volts for the 18650 and not the nominal 3.7 volts. It's also unclear when they consider the battery pack to be expended. Are they using the 50% of max brightness standard, or are they assuming something else? Still, this at least or more plausible and honest than the claims we from other flashlight vendors. It's nice that you can easily verify that their claims are ballpark plausible from a casual perusal of the specs. Some vendors fail that simple test.
The one year warranty seems about right too. They aren't claiming that much more brightness than one might expect from the specs: A U2 bin XML will provide between 970-1040 lumens at 3A, so they're getting a bit more brightness at the expense of a fair amount more current and, likely, a significantly shorter life for the LED.
You miss entirely the point, the issue here is how keep at 25C the junction of the emitter with a ridiculously small heat sink 100 grs are nothing to deal with 10W let alone 15W...
The Cree's figures are a lab reference , if you put the emitter over a pure copper piece the size of a car , still the junction get hot at 10W working at around 90%.
In the real world , best host , best heatsink , the emitter pushed at 10W work at near or beyond the 100 degrees mark, or 85% , lets say an U2 850 L at best...
At 15W in that host probably if survive could reach 600 L.
I don't see where the confusion is. There's at least two people with empirical results that show diminishing or even negative returns over 3A. Then there are people who think it's the epoxy that's problematic when it's stuff that's designed to deal with 100w cpu's, so they just don't know what they're talking about.
It seems CREE pushes the limits a bit closer with the XML vs the XPG or XRE, where going 25% over current is still somewhat effective.
It's also possible that some samples are better than others, just like all semiconductors.
I think I see the situation very clearly. I pointed to a thread at CPF that documented the fairly common knowledge that overdriving LEDs can produce more light - if you're a little lucky in the LED that you got, you don't destroy the LED, and you deal with the heat. One doesn't need to visit CPF for this - I could point out Foy's UF980L review, where the XM-L is clearly overdriven with good results: https://budgetlightforum.com/t/-/2284
For some reason this claim really seems to bother you - perhaps because you were unlucky and got an XM-L that doesn't accept more current. So, now we're in some kind of a game where you seem to want to attack the integrity of everybody who claims to be overdriving their LEDs and I wind up explaining that things aren't so sinister. First you went after a guy who, apparently, didn't reset the time on his camera, then you went after a vendor for making unrealistic claims about run time. When I showed that the run time claim was plausible, you then said I missed the point and turned the argument to heatsinking.
Perhaps you think that if I admit that a single person overdriving is LED is mistaken or deceitful that will somehow expose a larger lie that so many people are participating in? This game doesn't make sense to me. I have no interest in defending these people and there's no doubt that among them some are confused or deceitful. I'm just pointing out that they all can't be confused and/or deceitful.
So, let's think about what you're saying here. There are two aspects to your 600 lumen claim. The first is that the light will not stay cool and the second is that the temperature will get so high that even with the increased potential output of 1200 lumens, this will be cut by 50% due to the high junction temperature.
For the first claim, let's consider what kinds of things might be done to handle the extra heat introduced by that extra 5W of power. (Note that if the vendor is reporting the total current draw for the system as 15W, then the LED doesn't see all of the extra 5W.) In CPUs, which dissipate 65-90W in desktop form and up to 35W in laptop form, they add small fans. I guess it's kind of impractical to put a fan on your light. If only there were some way to guarantee that the user could have a constant flow of air around his light. Oh wait - it's a bike light!
For the second claim, let's extrapolate the curve you have pasted in to the point where there is a 50% reduction in output. Doing some back of the envelope calculations, this appears to be ~248 degrees. (Note that the curve is linear that it takes 125 degrees to go from 100% to 72%, so it will take a little less than 125 degrees more to go down to 50%.)
Typical solder will melt below 250 degrees, so you are accusing the manufacturer of selling a product that will melt and self destruct. Do you really think that they are selling a product with this kind of defect?
Let's take a step back: What we have here is a vendor who is overdriving an LED a bit. He is probably estimating the lumens coming off the emitter so that he can report results in the same way that so many others report their lumens rating of their lights. He is also probably stating the total amount of power driving the system - which includes loss to the driver an other inefficiencies, rather than the somewhat smaller amount that is actually hitting the LED. He is stating run times that look plausible, if very slightly optimistic, given the size of the battery pack and the assumption that he is using 2400 mAh cells. He could be using even better cells or the LED could he could be pulling less power from the cells than I estimated in my quick calculations. At worst, it seems we have some of the same (minor and well-known to the savvy consumer) marketing tricks that essentially every other flashlight vendor does.
Can we please stop accusing people who are overdriving their LEDs of being liars? Not everybody is honest, but this is a friendly board and it doesn't serve us well to clutter things up with all kinds of nasty accusations. It's not nice. It's not helpful, and I'm finding the game tiresome.
[quote=ronparr]
