Small Bright Lights: Too Hot or Not?

Cmoore

Enlightened
Joined
Aug 27, 2003
Messages
236
Location
Northern VA
I've noticed that there appears to be a wide variation regarding how hot some small, bright lights get after a few minutes of constant on runtime. Among my lights, here are some examples:

Peak Caribbean, pocket body: gets warm but not hot at all
KL4 on single 123 tube when using R123: gets hot as hell
HDS 60 on high: gets pretty warm, but short of too hot
Orb Raw: gets pretty warm but not hot at all
10 different 3W direct drive, single R123: mixed bag; some get hot as hell others just warm to warm plus

I've read many reports from CPF members that also suggest that there is a significant difference in heat buildup among similar lights. I realize my perception of warm vs. hot may be different than others; but, I don't think any differences here are enough to account for what I believe to be real and significant differences in how hot some small, bright lights get despite being the same lights and having the same components.

My question is; what can this apparent difference in heat buildup be mainly attributed to?
 

Macaw

Enlightened
Joined
Oct 8, 2005
Messages
565
I'm no expert but I would say that the efficiency of the LED would be a major factor in the temp differences, especially within the same model of flashlght. Luxeon lottery again.

As an example, My Peak Kino bay gets significantly warmer than my Fenix L1P, even though the Fenix is the brighter of the two and with longer on runtime. Both being single AA Flashlights.
 

paulr

Flashaholic
Joined
Mar 29, 2003
Messages
10,832
The heat depends on the power level, the converter and LED efficiency, and the light's ability to transfer heat to the outside world. So the 5 watt L4 naturally gets hotter than the 3 watt lights, etc. To keep a small light from heating up too much, hold it tightly in your hand when you use it. Your hand is a natural liquid cooled heat sink.
 

joema

Flashlight Enthusiast
Joined
Aug 14, 2005
Messages
1,189
Location
Nashville, TN
Cmoore said: "...I've noticed that there appears to be a wide variation regarding how hot some small, bright lights get after a few minutes of constant on runtime....My question is; what can this apparent difference in heat buildup be mainly attributed to?..."

This is a good question. After running 15 min. untouched on a table, my Caribbean is very warm but not uncomfortably hot. Yet other people report theirs is scalding hot.

People vary greatly in their heat perception so that's one possible answer.

Alternatively maybe overall efficiency varies due to mfg and component variation, but it's hard to imagine the magnitude of variation being sufficient to account for the described differences, esp in a given light that's non calibrated like the Caribbean.

With HDS maybe a little more understandable since each light is calibrated for output, sacrificing whatever runtime (hence efficiency) necessary to meet the output spec. Thus some might be significantly hotter than others. But that's not the case with most other lights.

If we could get some objective temperature measurements at certain run times for various examples of a given light, maybe that would help determine.
 

Cmoore

Enlightened
Joined
Aug 27, 2003
Messages
236
Location
Northern VA
My guess is it's mainly differences in LED efficiencies -- even given the same binned LEDs.

But, this is merely a guess on my part. I'd like to hear from some of the builders who deal with this issue.
 

TrueBlue

Flashlight Enthusiast
Joined
Mar 5, 2004
Messages
2,373
Location
Central CA
This is an interesting subject. I thought I was the only one that wondered about thermal dissipation of LED emitters in a light.

Some heads on lights get warm and some lights get unbearably hot. What is the difference? I'm sure I can safely say the brighter the light output means more heat is generated in the head. Any excess energy that is used to light the emitter is instantly converted to heat. Where does the heat go? If the light has a decent heat sink the hat is quickly transferred to the head of the light to be dissipated in the air, water or the hand holding the light. A bad heat transfer design looks the heat into the light then slowly destroys the light emitter. Drive levels of the light, heat sink transfer design, material used and the mass of the light determines how well the light transfer heat away from the hot emitter.

Thanks to CPF member Dbedit I now own a great and accurate digital temperature tester. The new unit is made by Fluke and called the Raytek Foodpro thermometer. It uses a single AA battery and is very accurate to within a half a degree. You can read more about the digital thermometer here.

Fluke Raytek digital infrared thermometer

I did a quick test of my SF L4 using a Li-Ion 17650 battery. I took the temperature of the head of the light then turned on the light. I left the light standing up so all the heat would be concentrated in the head and only thermal transfer from the heat sink would be used. In real life the light would be horizontal and be moving in the air so this was a very severe test. Every 30 seconds for 5 minutes I took temperature readings of the head of the light with the thermometer.

Start = 70.5 degrees
30 seconds = 75.5
60 seconds = 80.0
90 seconds = 83.0
120 seconds = 87.0
150 seconds = 90.0
180 seconds = 95.0
210 seconds = 96.5
240 seconds = 95.5
270 seconds = 99.5
300 seconds = 99.0

It looks like in three minutes the light has virtually reached it thermal capacity. At that point the body of the L4 started to warm up. So the heat of the head reached saturation levels and continued to transfer the heat to the body of the light. If I had continued to test the light I would think the head would remain the same temperature or if the body was efficiently linked to the head the head might actually lower its temperature.

A high output light with a small head like the L4 should heat up fast. A high output light with a bigger head, a larger mass, and a good heat sink capacity would not physically feel as hot as fast. I think a high output light with a small head that didn't heat up would worry me.

To figure out if there is a good heat sink capability then you'd need more information on the mass of the head, the design of the heat transfer and the drive current going to the LED.

I need to test (play) with the digital thermometer more. Thanks for the great toy, Dbedit!

 

paulr

Flashaholic
Joined
Mar 29, 2003
Messages
10,832
WRT the Caribbean, Peak's voltage boost circuit is not very consistent from one unit to another. So some Caribbeans will use more power, and dissipate more heat, and presumably be a bit brighter, than others. It will also depend on the exact battery voltage.
 

peskyphotons

Newly Enlightened
Joined
Jan 1, 2005
Messages
88
Location
Northern Ca.
I found that my Fenix runs quite cool. Why? Efficient? Underdriven? My SL jr. which I consider to have a poor runtime runs warmer. I have a 3 C cell, 3 watt light from Amondotech that has some sort of regulator in it that that became quite warm when the batteries where almost depleted. The battery end of the light was warmer than the led end and when I pulled the batteries out they were hot. The only thing I thought of was when the voltage in the batteries dropped the regulator pulled more amps, running up against the high resistance of the alkalines.
 

Robban

Enlightened
Joined
Jul 6, 2004
Messages
849
Location
Sweden
TrueBlue said:
Start = 21.5 degrees Celsius
30 seconds = 24
60 seconds = 26.5
90 seconds = 28
120 seconds = 30.5
150 seconds = 32
180 seconds = 35
210 seconds = 35.5
240 seconds = 35
270 seconds = 37.5
300 seconds = 37
I took the liberty of converting your values to C for us international folk :) I rounded the numbers off quite roughly.
 

TrueBlue

Flashlight Enthusiast
Joined
Mar 5, 2004
Messages
2,373
Location
Central CA
No problem. That is how it works. One person builds up on the last person's thoughts. Good job. :)
 

tvodrd

*Flashaholic* ,
Joined
Dec 13, 2002
Messages
4,987
Location
Hawthorne, NV
This one still cracks me up, and I am still surprised I didn't kill the LS. It was a BadBoy 500 on a CR2 and peaked at ~1 hour.










site1025.JPG


Larry
 

Trashman

Flashlight Enthusiast
Joined
Mar 15, 2005
Messages
3,544
Location
Covina, California
The KL4 on the 2 cell tube will get a lot hotter using a 3.7v li-ion (168s, 17670) than it will using 2 CR123's. So, with a 1 cell body, it's going to get even hotter faster! The 3.7v batteries have to work a lot harder to supply the necessary current to the KL4, hence the heat. The lux 5 in the KL4 needs between 6 and 7 volts to run; just think how much harder than 3.7v rechargeable must be working!

(note - I'm not expert here, so I won't be suprised if anything I just said gets shot down)
 

TrueBlue

Flashlight Enthusiast
Joined
Mar 5, 2004
Messages
2,373
Location
Central CA
I just tried testing my KL4 head on my variable power supply. At 6V the KL4 head draws 910mA. At 3.7V the current draw is increased to a strong 1550mA. That is 70% more power draw at 3.7V. No wonder the head heats up so fast. Less voltage means more current has to run through the KL4 head to maintain the bright beam.

Yes, this was a dangerous test to the KL4 head. On the PS I had the current regulation set for wide open and the PS could supply amps to the head to easily blow the emitter or electronics. In my PS there is no voltage sag light a battery would. The KL4 has tough components…thankfully.

The L4 would probably live a longer life by using 6V. The head would have more voltage and less power dumping into the light. Either 3V or rechargeable 3V batteries would extend the life of the light.

Here is the original temperature test with a single 3.7V battery.

Start = 70.5 degrees
30 seconds = 75.5
60 seconds = 80.0
90 seconds = 83.0
120 seconds = 87.0
150 seconds = 90.0
180 seconds = 95.0
210 seconds = 96.5
240 seconds = 95.5
270 seconds = 99.5
300 seconds = 99.0


So now that the light is cooled down I should try a temperature test using 6V batteries.

Start = 75 degrees (can you tell I live in CA; it was a warm day today)
30 seconds = 78.5
60 seconds = 81.5
90 seconds = 86.0
120 seconds = 90.0
150 seconds = 92.5
180 seconds = 97.5
210 seconds = 101.0
240 seconds = 100.5
270 seconds = 102.0
300 seconds = 103.5

At first it would seem that the KL4 gets hotter with 3.0V batteries. But check the spread of temperature for 5 minutes and you will see they are both 28.5 degrees hotter at the end of the test. We have just proven the physics principle of Conservation of Energy; energy is neither created nor destroyed. I'm proud!

By the way- I think tvodrd is more light suicidal than I am! :)

Are we off thread yet?
 
Last edited:

powernoodle

Flashlight Enthusiast
Joined
Feb 25, 2004
Messages
2,512
Location
secret underground bunker
Trashman said:
The KL4 on the 2 cell tube will get a lot hotter using a 3.7v li-ion (168s, 17670)

I did my own test of the thermal properties of a KL4/168S, though not on purpose. Took a Powernoodle nap one day, and the KL4/168S combo on my belt decided to start producing light without my consent. Don't know how long it was on, but it was trapped between my glorious Powernoodle body and the bed. The thing got so hot that I couldn't touch it, but seemed to suffer to long term ill effects.

Make your own joke.

best regards
 

balazer

Enlightened
Joined
Aug 22, 2005
Messages
486
LED efficiency does not play much of a role in how hot parts of the flashlight get. Even the best LEDs have something less than 25% efficiency, so roughly 80-90% of the power gets converted to heat no matter the LED type.

The most important factors in how hot parts of the light get are how much power the light draws, and how heat energy is conducted away from the LED. Variations in LED forward voltage can have a big impact on how much power a light draws, as can battery voltage in some cases. Heat transfer has to do with the volume and shape of the metal between the LED and the air, what kinds of metal they are, what kinds of gaps or seams are in the metal, and the amount of surface area touching the air.

Copper and aluminum are the two best common metals for heat conduction.

Air gaps between metal surfaces, even very narrow gaps between pieces of metal pressed tightly against each other, do not conduct heat nearly as well as the metal itself. That's why heat conducting compound is important. The heat conducting compound has much better thermal conductivity than air, but still not nearly as good as the metal. That's why it's important to keep the gaps narrow (have the metal pressed tightly together) with a thin layer of compound in between.

Good heat transfer keeps the LED cooler, which makes it brighter for any given current and makes it last longer.

A hot or cool head is not necessarily a sign of good or bad heat transfer. Heat is making it from the LED to the head, obviously, but without knowing how good the heat transfer between the LED and the head is, you don't know how hot the LED is to make the head that hot. It's best if the head has good thermal conductivity to the body of the flashlight, so that the head doesn't get too hot and the body can help dissipate the heat.
 

LEDcandle

Flashlight Enthusiast
Joined
Aug 15, 2005
Messages
1,943
Location
Mushroom S'pore
TrueBlue said:
Thanks to CPF member Dbedit I now own a great and accurate digital temperature tester. The new unit is made by Fluke and called the Raytek Foodpro thermometer. It uses a single AA battery and is very accurate to within a half a degree. You can read more about the digital thermometer here.

The Fluke is no doubt super accurate, but if your flashlight body has a low emissivity, the readings will be off.

Aluminium, depending on its finish, has a super low emissivity in most caes. *edit : Ok found it, anodized black aluminium is around 0.82. HAIII theoretically should be lower than this, but should not be that much worse off as anodized clear aluminium is 0.76.

So your readings should be quite accurate :D

I'm thinking of getting an infrared thermometer too.. but wondering about complications in measuring a chrome body. haha. might have to paste tape or use marker to mark a black spot on it.

Have you used it to measure the Luxeon slug directly?
 

offroadcmpr

Enlightened
Joined
Feb 3, 2005
Messages
810
Location
CA
Macaw said:
As an example, My Peak Kino bay gets significantly warmer than my Fenix L1P, even though the Fenix is the brighter of the two and with longer on runtime. Both being single AA Flashlights.


My guess for that is that the fenix seems to have a more efficient driver. Drivers can up to 40% of the power given to them through heat. Even though the fenix is brighter, it also lasts longer on a alkaline battery, which means that it is actually using less power than the peak, therefore making it cooler.
 
Top