A paradox to thermal relief??!?!?

Hi guys,

One of the primary concerns for these LED lights is the thermal relief of the heat generated by the LED. We know this and have discussed it to a great extent! In a recent run time test on a driver/ battery conbination which I felt was pushing the power limit and after touching the hot light during the test, I decided to try a run time test with the light sitting in a pan filled with a significant volume of water. I reasoned that the water would provide more than ample thermal relief to the LED. This was the case as the head was not even warm to the touch during a subsequent run time test. However, the results of the test were disconcerting as the timein regulation was significantly shorter than the previous test! This has been in the back of my mind and bugging me ever since. During the last week or so, I have run a number of air and then water cooled runtime tests on various lights.

In each pair of tests, the air as well as water has been ambient and the same, within a few degrees. I have combined a few of these air/water (Hot/Cold) tests and put them on a page, HERE . Below I will include a couple of the runtime paired graphs as well. In all cases, there is a significant and noticible reduction in the primary lithium cell's capacity when it is in the water cooled environment!!!

I was under the impression that one of the real advantages to lithium chemistry is the consideration that these cells work well in cold applications. I am now of the opinion that they may work but at potentially diminished capacities!! perhaps this information has been known and discussed in the battery and electronic forums but I sure haven't picked up on the fact!! /ubbthreads/images/graemlins/icon15.gif It would seem that what is ideal for the LED (25C) is not ideal for its energy source?!?!? /ubbthreads/images/graemlins/icon3.gif From the LED manufacturers, we have cool graphs of flux VS temperature. Are there similar graphs matching a cells capacity VS temperature? With a low thermal resistance in a flashlight, I would assume that at some low ambient air temperature that the lithium battery would perform in a diminished capacity?!? Do our lithium based flashlights need body warmers in the winter time?!? /ubbthreads/images/graemlins/thinking.gif



**********




**********





There are a few more examples shown on the linked page above but I think the trend is rather clear. It would seem that not only is the capacity (amp hours) diminished but the ampacity or max power available is also diminished. I did one test shown with a Ti head and body and I figured that the thermal resistance of the metal would give the battery a better chance of remaining at an elevated temperature. It may well have helped but the reduction in run time was still way too significant and over shadows any hint of aid.

All of the CR123 cells used in the test were SF and the CR2's usd in the test were from Battery Station. I have no idea if this evidence is based on the general chemistry of lithium primaries and for all I know the fact that the batteries were standing on end may be contributory. I do know that I am ignorant and possibly stupid in regards to these tests and their results. I trust some of the battery gurus can enlighten us!?!?!

I mentioned this to Wayne Y. the other day and he suggested that this thermal "dependency" might also have some bearing on the noted results of two batteries in series having different discharge rates. /ubbthreads/images/graemlins/thinking.gif /ubbthreads/images/graemlins/icon3.gif

So, in short, what's the skinny here and do I need a neoprene/ down bootie for my light if I take it underwater or up in snow country?!?! /ubbthreads/images/graemlins/grin.gif

I am also curious as to how the delta in C will vary for the Li-Ion cells as a function of T. This last statement almost makes it sound like I know what I am talking about! /ubbthreads/images/graemlins/crackup.gif

You sre getting ~2-3% reduction in output with a hot LED, and ~25% reduction in runtime with a cold batt?

Not that I know squat about electronics, but I wonder of you could do a "backwards" PD-type arrangement with a thermally conductive liner -- insulated from the forward body -- taking heat from the LED and transferring it back. Heat would warm the battery end first, making its reaction(s) more efficient, before heat was dumped to the environment at the tail end.

I dunno. Interesting problem though.

Scott

Ladies and Gentlemen...the "professor" has entered the building. /ubbthreads/images/graemlins/grin.gif...... /ubbthreads/images/graemlins/confused.gif /ubbthreads/images/graemlins/thinking.gif /ubbthreads/images/graemlins/mpr.gif /ubbthreads/images/graemlins/au.gif /ubbthreads/images/graemlins/stupid.gif

Have you tried variations, like one in the fridge and one in warm water? Correlation/causality is the devils playground /ubbthreads/images/graemlins/smile.gif

just to cover a formality.... was there any water seepage into the lights?

...beats me.... /ubbthreads/images/graemlins/smile.gif

I also put my money /ubbthreads/images/graemlins/broke.gif /ubbthreads/images/graemlins/crackup.gif on the lower overall resistance in the circuit due to lower temperatures... /ubbthreads/images/graemlins/stupid.gif

OK, this is one for Silverfox, in case he thought he was running out of cell testing ideas. But, it shouldn't be TOO hard to pull off.

Lets see a capacity test of CR123 and R123 cells held at different temperatures, ice water and 120 degrees or so.

I can see Tom now, hunched over a pan of water with a thermometer stuck in it, occasionally adding dollops of boiling water to keep the temperature in the right range. Steam, the eery glow of some blue LEDs maybe. Real mad scientist stuff. /ubbthreads/images/graemlins/eek.gif

"OK, this is one for Silverfox . . ."

I was just thinking the same thing. I forgot about the thermal effects on the circuit.

Scott

There was no flooding in the battery chamber or LED electronics area. I do not believe there is any external circuit sapping the electricity. The water is tap water. I believe that for the most part, we are seeing the results of the battery chemistry at or close to ambient VS some elevated temperature. I have seen enough to feel that the batteries themselves should be tested with a controlled load and at a monitored and controled temperature. I considered contact resistance initially but then one would presume that if the contact resistance was high enough to have the noted effect, there should be an indication of a reduction in the light output ala two stage switch and this should actually improve the run time; predicated on the resistance at the input end. Initially these tests were confined to the Wiz2x2 converter and I was concerned that there might be some adverse effect of temperature on the circuit efficiency. At that point, I brought in the DB and NG and KL1 converters and found that this issue is not isolated to a specific circuit. /ubbthreads/images/graemlins/thinking.gif I would assume that this "whatever" would be possibly even more noticible in a DD application and even the hot wire lights. If power is hampered as it seems may be the case, a hot wire should be dimmer and if the voltage under load is hampered, a LED would be dimmer in DD as well?!? /ubbthreads/images/graemlins/help.gif

Since these tests have all been conducted in Berkeley by an admitted half witt, there needs to be some verification or similar results from tests eleswhere by someone who has a clue! /ubbthreads/images/graemlins/blush.gif

IIRC I read *somewhere* that a hot lithium cell can dissipate much more power than a cold one. Wasn't it even discussed that some "strange battery failures" would be because of this fact ... the light running happily with hot batteries and upon re-ignition ( /ubbthreads/images/graemlins/cool.gif word) with the cold batts it would be "suddenly" dead?
bernie

P.S.: Of course I don't have a clue here, just re-producing what my feeble memory tells me ... /ubbthreads/images/graemlins/grin.gif

[ QUOTE ]
McGizmo said:

So, in short, what's the skinny here and do I need a neoprene/ down bootie for my light if I take it underwater or up in snow country?!?! /ubbthreads/images/graemlins/grin.gif


[/ QUOTE ]


First in line for a neoprene booty for my McLux III- PD /ubbthreads/images/graemlins/nana.gif

Actually I really would like some sort of bite guard for it when I'm extra hand challenged./ubbthreads/images/graemlins/thinking.gif

Sorry for going off-topic.

Another question ... why don't we see a diminished output from the hot LED? /ubbthreads/images/graemlins/thinking.gif
As these are current-regulated set-ups one should expect to see reduced efficiency somewhere?

Except ... if the water doesn't do a better job of cooling, which would make everything else look right again.
Don ... can you do a test with the water moving around the light? ust thinking ...

Someone might want to alert NewBie to this thread ... /ubbthreads/images/graemlins/grin.gif

bernie

Any idea what the temperature difference was between the "hot" and "cold" runs for the batteries? For many battery chemistries, in general, the chemical reactions work better when it's warmer rather than colder. I suspect that's what's going on here, but without some temperature numbers, it's just speculation.

Graphs here and here may help to put some magnitude on the effect. Nice to see a "bomb" dropped upstairs for a change. There's going to be a lot of /ubbthreads/images/graemlins/thinking.gif over this one! /ubbthreads/images/graemlins/thumbsup.gif Don!

Larry

[ QUOTE ]
Kiessling said:
IIRC I read *somewhere* that a hot lithium cell can dissipate much more power than a cold one. Wasn't it even discussed that some "strange battery failures" would be because of this fact ... the light running happily with hot batteries and upon re-ignition ( /ubbthreads/images/graemlins/cool.gif word) with the cold batts it would be "suddenly" dead?
bernie



[/ QUOTE ]

Sorry I missed this info and it makes me wonder if the cells shouldn't be given an insulated chamber in some considerations?!?! Perhaps an electric blanket? /ubbthreads/images/graemlins/nana.gif

In terms of the LED dropping in light output, you do typically see an initial drop as the light reaches steady state in the "hot" tests. This drop is not real significant because presumably the light is operating with reasonable thermal relief.

Again, I have no idea of the temp inside the battery chamber when the lights are Air cooled but assume that the temp is much closer to ambient in the water cooled tests. I agree that it is in this delta that the diminished capacity likely results.

The questions I have are:
1) what is the optimal temp for the batteries to function in and what is their reasonable range of operation; what is the Delta C VS Delta T.

2) if the optimal range is known, can the batteries environment be inhanced for a practical matter and if so, Is this enhancment viable for the typical ambient ranges one would use the light in?!?!? I expect the solution for the Sahara at night in summer may be different than the north pole in any season! /ubbthreads/images/graemlins/icon15.gif

Yep - nothing surprising here. Hot battery = more efficient chemical reaction.

I've run some high powered (750mA+ drivers) from a single CR123 and they run for a certain amount of time - the battery does get hot. If allowed to cool the battery is no longer capable of running the driver at full current. Start the driver at lower current and slowly ramp up the current as the battery starts to warm up (causig it to warm up more) and it then runs at full current again.

So, what this means is that you can do a runtime test and get say 30 minutes (battery really hot at the end).

Do the runtime test again, but in shorter intervals (that allow the battery to cool down) - at some point (well before the previous 30 minute total) the driver won't be able to start up again because the battery is too cold to deliver the required current.

This effects probably most batteries, but a single CR123 being on the teetering edge of being able to run a single Lux3 via a high current boost converter is much more problematic. That'll teach us for using a battery that was meant for point and shoot cameras (low current app) in high drain applications instead!

Bring on the li-ion...

george.

Larry,
Thanks for the graphs! It is interesting that they look at RT but in the case of our lights, the batteries are most often operating at well above RT if the light is on for any length of time. It would be instructive to see if the graphs continue to shift up at these "elevated" temperatures.

EDIT: George slipped in there on me! /ubbthreads/images/graemlins/smile.gif So George, have you done any testing or evaluation of the Li-Ion cells at various operating temperatures? Is their chemistry less effected by their opperating temperatures?

[ QUOTE ]
McGizmo said:
EDIT: George slipped in there on me! /ubbthreads/images/graemlins/smile.gif So George, have you done any testing or evaluation of the Li-Ion cells at various operating temperatures? Is their chemistry less effected by their opperating temperatures?

[/ QUOTE ]

Not yet, I've been doing some runtime tests of the bbFlex (buck/boost li-ion) driver and have finished the basic tests. They are very consistent (I cycle the 2 same cells), so it should be easy to change the temperature of the cell and re-run a couple of the shorter (higher current) tests and see what happens.

I would 'guess' that the li-ion cells wont be as affected as CR123's since their internal resistance is a lot lower (hence capable of more efficiently delivering higher current to the driver) - but we'll see. Should have some numbers for you early next week.

It'll probably just prove that we should all move to the australian desert to get the best runtimes /ubbthreads/images/graemlins/grin.gif

george.

Great George! I have a couple Pila's and Copias I could test but I would not be confident in the recharge, that I in fact had the cell starting each test with the same bank balance. /ubbthreads/images/graemlins/blush.gif

I seem to remember some posts by SilverFox, and or Ginseng, about improved effeciency of CBP 1650 AA cells, when they are really hot vs cold. Maybe same issue with Li primary cells,, rechargeable LiON cells,, and NiMH cells. All more efficient at high temps.

Bill