10440 in Fenix LD01

LED Flashlights – General Info 1 x AAA Flashlights
1

I know nothing about LED flashlights, can anybody please tell me if I could use a 10440 battery in a Fenix LD01

Thanks :slight_smile:

2

f e n i x l d 0 1 r 4

ENELOOPS:

tailcap readings.

tailcap readings with freshly charged Eneloops (1.54V)...vs...depleted Eneloops (1.20V)

Med: 0.207...0.657A

Lo: 0.022...0.041A

Hi-1: 0.655...1.750A

Hi-2: 0.448...0.551A

comment: as expected, amperage rises with falling cell voltage. this way the brightness is regulated constant.

actual runtime measurements.

Eneloop runtimes on Hi:

1.54V, cell#2, 2h0min, @0.83V ->1.1494V, N/A

1.55V, cell#8, 1h44min, @0.83V ->1.1691V, 798mAh

1.54V, cell#6, 1h37min, @0.88V ->1.1848V, 749mAh

1.54V, cell#7, 1h35min, @0.98V ->1.2008V, 745mAh

1.54V, cell#5, 1h35min, @1.01V ->1.2103V, 772mAh

1.54V, cell#6, 1h34min, @0.91V ->1.1934V, 761mAh

1.54V, cell#2, 1h33min, @0.96V ->1.1966V, 742mAh

1.55V, cell#7, 1h32min, @1.13V ->1.2268V, 728mAh

1.54V, cell#8, 1h31min, @1.15V ->1.2356V, 737mAh

Eneloop runtimes on Med:

1.54V, cell#8, 3h24min, @0.83V ->1.1577V, 811mAh

1.55V, cell#2, 3h15min, @0.88V ->1.1761V, 755mAh

1.54V, cell#8, 3h12min, @1.09V ->1.2000V, 740mAh

1.55V, cell#7, 3h8min, @1.11V ->1.2039V, 753mAh

Eneloop runtimes on Lo: TBA

1.55V, cell#2, 31h45min, @0.85V ->1.1617V, 772mAh

note:

down-step from Hi-1 to Hi-2 occurs after ~7mins (6min53sec, to be exact!)

no matter which voltage the cell has, in High-mode the torch comes on always on Hi-1, for example in series you'd get something like this:

#1, 0.655...0.768A, then down-step (from 0.768A) to 0.448A

#2, ...0.823A, then down-step to 0.xxxA

#3, ...0.838A, then down-step to 0.481A

#4, ...0.862A, then down-step to 0.487A

#5, ...0.872A, then down-step to 0.xxxA

#6, ...0.926A, then down-step to 0.510A

#7, ...1.055A, then down-step to 0.551A

#8, ...1.750A, then drops out of regulation

If you sum these intervals, the (max. possible) accumulated runtime on Hi-1 is about 49min (7*7min), or in other words, if you need max brightness you could turn on and switch to Hi-1 (period#1), then after ~7mins, cycle thru the modes again to High-mode (Hi-1; period#2), for a total of 7 periods. After that, when you're in your 8th period, the cell is becoming depleted and you get only a few more minutes (< 7min) on Hi-1 until the light completely drops out of regulation: the light would dim steadily within minutes from 1.750A to low light output.

10440's:

The older PWM-regulated LD01 versions (Q4, Q5, R2, R3) were known to support 10440's, see the many CPF references. So what about the current version, the LD01 R4 which is not PWM-regulated but current-regulated (no PWM confirmed!)? First let's start off with this link from authorized Fenix dealer Lighthound:

http://www.lighthound.com/Fenix-LD01-Cree-R4-AAA-Flashlight-HA-III-Black-Finish-can-use-10440-batteries_p_3377.html

The product sales webpage states clearly that the LD01 R4 black aluminum version can use 10440 LiIon cells! There are "hundreds" of Fenix LD01 threads on CPF and often the 10440's topic is included in the discussion, most of which date from 2008-2010 and cover the PWM-regulated versions. i've selected the most relevant or interesting threads, may they be helpful related reading (in no preferred order):

I did enough reading, i've had it! :tired:, now i just go ahead and use 10440's for the sake of this review:

tailcap readings.

tailcap readings with freshly charged Protected Ultrafire 10440 (4.215V resting voltage)...vs...depleted cells (3.600V resting voltage; within seconds switftly drops/sags down to 2.750V under load, shortly before the protection trips):

Med: 0.890A...0.105A

Lo: 0.169A...0.019A

Hi: 1.316A...0.187A

comment: on Hi, for a cell with 300mAh capacity, the initial discharge current is 4.4C which is way beyond the 2C safety specification of 10440 LiIon cells which says that the continuous discharge rate of a 10440 cell should not exceed 700mA. Assuming that the actual capacity of 10440's (often rated with a nominal capacity of 500mAh or 600mAh, see Trustfire or Ultrafire 10440 cells lol) is somewhere between 250mAh and 350mAh the initial discharge current calculates to between 3.7C and 5.3C. 320mAh would be a generous bet, and this calculates to 4.1C. However, since the light is not regulated on Li-Ion's, amperage decreases with decreasing cell voltage, the light gets dimmer and dimmer on Li-Ion's and there is no continuous (or constant) 4C discharge rate to be feared. In fact, on Hi the discharge rate continuously declines from 1.3A (4C, 4.215V) to 0.2A (0.6C, 2.750V).

FYI, even if a torch (the PCB and the LED) can withstand for example 4C with proper cooling, it doesnt mean that you or the torch are safe. Continuous constant 4C is a hazardous discharge rate for normal 10440 LiIon cells and it could lead to poof or kaboom. in practice you are safe to discharge at 4C for short periods of time, e.g. for 20sec time intervals. for the sake of this detailed review with tests and measurements, i am venturing continuous discharge tests on Hi mode (and on the other 2 modes).

actual runtime measurements.

Protected 10440 runtimes on Hi, until the protection trips:

cell1 (marked): 20m27s (60sec intervals) / 28m12s (full run) / 25m24s (full run) / 18m36s (60sec intervals)

cell2 (unmarked): 17m37s (90sec intervals) / 20m15s (full run) / 20m37s (full run) / 20m13s (full run)

Protected 10440 runtimes on Med, until the protection trips:

cell1 (marked): 33m26s / 32m50s / 31m38s / 30m16s

cell2 (unmarked): 26m25s / 25m48s / 25m8s

Protected 10440 runtimes on Lo, until the protection trips:

cell1 (marked): 2h19m31s / 2h18m46s

cell2 (unmarked): 2h7m4s / 2h8m36s

some details of the runtime measurements.

Now let's put some efforts in measuring tailcap current and cell voltage in detail during the above testruns, especially on Hi-mode. See the following discharge graphs of continuous runtimes (marked cell on Med, unmarked cell on Hi) to get an idea of how fast the current decreases within the first 2mins (="120,0" on the x-axis) both on Hi- and Med-mode. The above 2 underscored numbers are taken directly from this measurement graph:

measurements and graph by kreisler, UT61E and Excel2010, thx to HKJ for helping with it

you can clearly see that the light is not regulated on 10440's. if it were, then the current would increase(!) with progressing runtime, to stabilize the brightness ( wattage = amperage x voltage ). the LD01 R4 on 10440's, when time progresses voltage decreases (because the 10440 cell is being depleted) and amperage decreases! therefore the light dims. no constant brightness on 10440's! what else can you see from the above graph? you can see that shortly before reaching 7.0mins continuous runtime on Hi-mode, the light drops in amperage from Hi1 (~0.86A) to Hi2 (~0.76A) .. in a similar way as we have been observing with Eneloops. this means that the light uses a built-in timer (capacitor?) every time when you switch the light to Hi-mode, and this is independent from the cells (Eneloops or 10440's). Anything else we can see from the above graph? Well, the actual "average continuous" current draw on Hi-mode is ~0.8A (2.5C), and ~0.55A (1.5C) on Med-mode, which is way below the initial 4.4C. You ask me: So kreisler is 2.5C finally a safe discharge rate yes or not or what?? And i answer: Get off my back thx haha and simply enjoy the next visualized 3 testruns, all on Hi-mode, but this time conducted in short intervals.

Each interval has a length of 90sec runtime:

unmarked 10440 cell accumulated runtime offline voltage at rest
interval 0 0'0'' 4.200V
after interval 1 1'30'' 4.06V
after interval 2 3'00'' 4.024V
after interval 3 4'30'' 3.952V
after interval 4 6'00'' 3.898V
after interval 5 7'30'' 3.856V
after interval 6 9'00'' 3.807V
after interval 7 10'30'' 3.788V
after interval 8 12'00'' 3.744V
after interval 9 13'30'' 3.703V
after interval 10 15'00'' 3.676V
after interval 11 16'30'' 3.535V
after interval 12 17'37'' tripped at 2.750V

Each interval has a length of 60sec runtime:

marked 10440 cell accumulated runtime offline voltage at rest
interval 0 0'00'' 4.200V
after interval 1 1'00'' 4.09V
after interval 2 2'00'' 4.03V
after interval 3 3'00'' 3.97V
after interval 4 4'00'' 3.90V
after interval 5 5'00'' 3.83V
after interval 6 6'00'' 3.81V
after interval 7 7'00'' 3.79V
after interval 8 8'00'' 3.77V
after interval 9 9'00'' 3.75V
after interval 10 10'00'' 3.74V
after interval 11 11'00'' 3.73V
after interval 12 12'00'' 3.72V
after interval 13 13'00'' 3.71V
after interval 14 14'00'' 3.69V
after interval 15 15'00'' 3.68V
after interval 16 16'00'' 3.66V
after interval 17 17'00'' 3.63V
after interval 18 18'00'' 3.54V
after interval 19 19'00'' 3.42V
after interval 20 20'00'' 3.29V
after interval 21 20'27'' tripped at 2.750V

Same cell, same 60sec interval length, just another testrun, this time with stricter pauses of exact 9.0min between each interval:

marked 10440 cell accumulated runtime offline voltage (after 9min rest)
interval 0 0'00'' 4.20V (4.200V)
after interval 1 1'00'' 3.93V (4.116V)
after interval 2 2'00'' 3.88V (4.063V)
after interval 3 3'00'' 3.84V (3.996V)
after interval 4 4'00'' 3.82V (3.949V)
after interval 5 5'00'' 3.79V (3.904V)
after interval 6 6'00'' 3.76V (3.867V)
after interval 7 7'00'' 3.74V (3.839V)
after interval 8 8'00'' 3.72V (3.816V)
after interval 9 9'00'' 3.71V (3.797V)
after interval 10 10'00'' 3.69V (3.780V)
after interval 11 11'00'' 3.67V (3.766V)
after interval 12 12'00'' 3.66V (3.750V)
after interval 13 13'00'' 3.64V (3.731V)
after interval 14 14'00'' 3.62V (3.708V)
after interval 15 15'00'' 3.60V (3.691V)
after interval 16 16'00'' 3.56V (3.676V)
after interval 17 17'00'' 3.45V (3.578V)
after interval 18 18'00'' 3.22V (3.376V)
after interval 19 18'36'' tripped at 2.750V (3.212V)

The above tables are more detailed than graphs and they can be helpful to determine how much runtime you got left if you check your (unprotected, for that matter) 10440 cell with a digital voltmeter. For example, when you remove the cell and it has been resting and the DMM measures 3.80V, then it means that you got only ~10mins on Hi-mode left, if you use the light intermittently for 60-90sec intervals. Likewise, if it measures the magical 3.60V limit, then you got 2-3min left (= 2x 60sec, or 2x 90sec) before the voltage under load reaches the (another magical) 2.75V limit. I mean, on the LD01 it's easy to notice when you just have reached the 2.75V limit: the light isnt bright anymore!! haha. No, seriously. As soon as you notice a considerable drop in brightness on Hi-mode (or Med-mode), it means that the light has passed the 2.75V limit (e.g. down to 2.10V) and you must stop using the light by all means. Then let the unprotected cell recover (e.g. up to 2.50V) and finally feed it to your charger and hope that the charger accepts the overdischarged cell. As a general rule, if the cell's resting voltage has reached 3.5999V (=2-3mins on Hi-mode left), you should stop using the cell and go recharge it. The above tables are also helpful/valuable for you for reproduction purposes: if you can reproduce the above measurements (tailcap readings with freshly charged cells, voltage readings after 60secs or 90secs of usage on Hi-mode), then it means that your copy performs exactly as my sample. Fenix has high production quality standards and we should assume that all produced LD01 R4 units perform equally well and safe (e.g. with 10440's) but users all over the world (amazon, forums, google, wiki, ebay, haha) do report about stupid tints, malcentered emitters, and dying specimen.

The above 3 tables (containing 4 sets of voltage measurements) generate these line graphs with interpolated points. The 3 blue lines (=resting voltage of cell1 or cell2) share an interesting common characteristic: after 16mins of accumulated Hi-mode runtime, the lines would part from that intersection on! The resting voltage at that time is ~3.67V which is therefore a critical voltage in the LD01 R4 so to speak:

measurements and graph by kreisler, UT61E and Excel2010, thx to HKJ for helping with it

comment: as you can see from my published multiple runtime tests on 10440's, continuous runtimes heavily depend on the state and quality of the particular cell. there seems to be a considerable quality difference (capacity difference) between cell1 and cell2, both of which were bought together in the identical 2-pack from DX. also note that the runtimes got shorter with each test run! indication of cell degradation and cell aging? You're asking 'hey kreisler i've seen such shaped line graphs before! is it coincidence that the Fenix LD01 R4 10440 discharge curve (amperage vs. runtime) looks so similar to battery discharge curves (voltage vs. capacity)?' and my answer is 'well observed, well well! maybe it has something to do with DD'. haha.

duh.

and last but not least note that continuous runtimes are longer than using the light in intervals. the interval method, e.g. max 60-90 secs on Hi1 with 10440's, may be healthy for the Cree XP-G R4 LED's life but dont wonder about the short accumulated runtime. the high initial discharge currents (e.g. 1.316A with a freshly overcharged 10440 cell) when you switch on (or to) Hi1-mode do explain why accumulated runtimes from the interval method have to be shorter in sum.

IMPORTANT NOTE (DISCLAIMER):

since the LD01 R4 is not certified for use with 10440's i do not recommend or encourage feeding 10440's. during the many runtime tests conducted for the sake of this review (and for the sake of giving complete information, which was btw never done before! so this is a world-premiere for the LD01 R4 --- a full-fledged review!!!) there was no harm done to my LD01 R4 copy and other users on cpf report the same: their LD01 R4 copies do withstand the high current/high voltage of protected or unprotected 10440's without problems. however, in all fairness, this is no guarantee that your copy will not fry at some point. my strong advice is: DO NOT USE 10440's IN YOUR LD01 R4! the point is: the light is incredibly bright on Eneloops (conservatively rated 72lm) and it looks even brighter because of the tight beam (tight spill + very tight hotspot, rated 50m throw), the LD01 R4 produces higher actual brightness and runtimes than the Klarus MiX6 (rated 85lm lol), you also got the most perfect beam beauty and the best possible pure white tint and in both it beats the legendary premium MiX6 which comes itself with immense beam beauty and pure white tint. in other words, it's already the #1 top best performing 1xAAA keychain light (and many people who "know" the long-existing Fenix LD01 product are not aware of this sensational fact), so why would you desire or risk driving the light beyond design specifications? for what good reason? out of boredom? or maybe simply because you are extremely curious, possess tons of dollars and dont mind possibly ruining a perfectly working #1 top best performing 1xAAA keychain light? well, if you're curious, which is a valid good reason imho, then this detailed review (and up to this date, it's the only existing review of the R4 version of the LD01!!!) should have satisfied your §$%&! curiosity. 10440's in the LD01 R4: I've proven that it works, i've proven that it rivals the iTP A3 XP-G R5 on Hi, i've shown that you get 3 distinct (even if bad-spaced) modes, i've clocked continuous runtimes, i've measured accumulated runtimes, i've measured amperages both on fresh and on depleted cells, i've been ruining (=wasting) my cells with the multiple and repeated test runs, i've published full amperage discharge graphs on Hi and Med, and i've created several animated beam slice GIFs which visualize the absolute and comparative brightness. so what else could you possibly be curious about?? all you need to know about 10440's in the LD01 R4 is presented in this kicka$$ review so your curiosity (about this sort of info) is herewith officially satisfied, and it cant be named anymore as reason why you would want to risk it. in any case or which other personal reason you might have .. the LD01 R4 is really(!) not designed to run on 10440's: you lose a real Lo! With 10440's on Lo, the LD01 R4 is as bright as on Hi with Eneloops. that's a real bummer and kills many possible uses of a versatile keychain light (preserving night vision, reading a book at total darkness). a keychain light without a real Lo is inacceptable, you'll realize this in practice in the long run believe me. there are lights which you can run with confidence on 10440's so use them if you need to wow your fellas: the iTP A3 XP-G R5, the Tank E09, and others. Here a suggestion: if you are beyond the 30 months Fenix warranty period and dont care much anymore for yours (e.g. because you got better ladies at home), then ..etc..

brightness levels, given freshly charged 10440's:

lo (eneloop) < med (eneloop) < hi2 (eneloop) < lo (10440) < hi1 (eneloop) < med (10440) < hi (10440):

ANIMATED GIF

jojo

3

So that’s a definite no then, or perhaps a yes with reservations, or even a moot maybe :~

Thanks for the info :slight_smile:

I’m going to have to stick my head in a bucket of water after reading all that , my head is overheating. I’m better with dogs than LEDs :slight_smile:

4

kreisler

I read .

I appreciate your exhaustive reviews and humorous insights and I have decided that you are the BLF AAA flashlight info go-to .

5

thanks for the poem Jack ;)

( naah. i am only E09 and Eagletac freak, that's all. but i have 1 sunwayman titanium incoming!! bling bling :love: )

6

When it comes to the LD01
The Fenix is truly the one
It handles the power
And produces a tower
Of light that will get dark undone.

When it comes to a very small light
Kreisler lends the greatest insight
He exhausts every view
And brings the story to you
With a smirk and a grin worth the fight

Yeah yeah, I know, but Jack didn’t deliver.

I’ve run my LD01 on an UltraFire protected grey cell (UltraFlame?) for quite some time, but as it’s a stainless body it’s been pretty easy to see it getting hot and shut it down in short intervals of use. When I stepped up to better lights for my keychain I put the LD01 on my wife’s keyring with an Imedion NiMH LSD cell for safety.
Great write up, most I’ve ever seen you say and it was all good! :slight_smile: Thanks, will definately bookmark this one for reference, in fact it should be a sticky!

SB? Care to sticky this one?

7

DBCstm, omg .. thanks so much for it! :D

8

Kreisler, have you tried an Olight i3s?

I have an LD01 and i3s and while output is similar, I prefer the shorter size, cleaner beam profile and less cool tint of the i3s, which also seems to have a brighter hotspot despite the deeper LD01 reflector.

An LD01 with XP-G2 would be very nice though.

9

Hi, sorry i havent tried the i3s yet. Might do so in future. But honestly i am thinking that i have enough good quality AAA lights including the E09 and Thrunite Ti. All these small lights perform similarly imho, so in the end it's a matter of taste and liking.

I really love the design, looks and feel of E09 above all (and now my new ES12). And then there is still my ITP A3 Titanium which looks gorgeous and robust and isnt too far from the i3s.

I am sure that the i3s is a really great light. Afaik our friend UPz is going to leview it soon, in his new integrating Fenix sphere:

http://www.forolinternas.com/viewtopic.php?f=24&amp;t=5176

He's the real master of AAA lights.

10

jeje

This reminded me how I feel when I first open some of your reviews :stuck_out_tongue:

11

I run it in my LD01…no probs so far and it is a bright little beast.

12

Maybe some additional heat sinking might help? Have you tried thermal grease?

OH… you were talking about your head, not a flashlight :(!!

Just kidding :)!

13

This is my LD01 , some metal trimmed off the barrel should help cooling but then again I’ve added some so,?

My piles will never make me famous
I’m a battery ignoramus
I can’t tell a Ni-Cad cell from a Li-on cell
Or even a good cell from a bad cell

With apologies to Ogden Nash and real poets everywhere