Nimh vs Alkaline - A real world test (running in burst vs continuous)

Batteries and Chargers Rechargeable Batteries
1

You’ve seen those runtime graphs where alkaline battery sputters almost immediately and
rechargeable nimh gives nice long regulation and runtime? Me too, but I wondered whether
this was the whole truth.

I decided to do a real world test, where I would run a light in small, 3-minute increments,
with an hour or so between them. This is more in line with actual usage. Even better
test would be to run one burst per day instead of per hour, but I thought I’d do this quick
test first and see what the result is. It took 3-5 days to complete the test for each battery.

Conventional wisdom with alkalines is that their capacity depends on how much current you’re
pulling from them. The higher the current, the lower the capacity. So continuous vs burst
running them shouldn’t mean big difference if the current required stays the same.
But then again, they seem to regain some strenght when left unused, and I wanted to
test whether this affected the overall runtime.

The thing I wanted to know in this real world test was how much usable light a flashlight
would provide when run in increments vs continuously. For this test I defined “usable light”
as 6 lumens. The light I was using was Singfire sf-348. Batteries were a 900mah eneloops and
some generic supermarket alkalines.

Here’s the first graph, alkaline vs nimh in continuous run. Alkaline loses.

Now, what if you use alkaline in short bursts? Each peak in the graph is the start of
a 3 minute run. Does that change anything? It sure does after 30 min mark.

Ok, now it’s nimh’s turn. Is there any performance change in real world testing? Yep.

And in the final graph there are both burst results:

I have to say this is a result I didn’t expect. While the graph doesn’t look all that
amazing, alkaline does provide THREE TIMES as much usable light when used in burst manner vs
continuous, whereas nimh goes the other way and actually degrades a bit (and I tested 2
batteries to be sure). Nimh does hold longer in regulation, but after 60 minutes you’re going
to wish you had gone with alkalines. That being said, I’m not giving up my nimhs anytime soon.

What this test shows is that continous runtime testing does not correspond 100% to real world,
and that alkalines aren’t as poor as I previously thought.

1 Thank
2

Very interesting!

What I would like to know is how a lithium primary AAA (for example Energizer L92) performs under the same circumstances?

3

Good thinking and testing, nice!

4

Really interesting, I use an aaa light pretty much in short burst most days, always wondered the real difference between alkaline and nimh.
Thanks for taking the time to test and post the results

5

I never thought alkaline has more/longer power in short burst. Thanks for the test.

6

Excellent data, got me thinking about the mecanism, i wonder if letting the alkaline run continuously allow it to reach detrimental temperatures that degrades the chemistry.

7

Great test! The difference is such bigger that I expected. I wonder if a regular eneloop would bee different than the pros. I have a L92 in my keychain light, and I use CR2 primary in one light and CR123A in some lights. I get a lot of time out of these lights because they are infrequent use. If I need my keychain light for more than 3 minutes (and that’s long) I’m likely to take out my 18650 EDC. I get about 6months on my L92.

8

What I don’t understand, is why using NiMH in burst mode gives you 20% less running time than if you run continuously. That doesn’t make sense. I thought it would make very little different in run time.

Is the battery getting a boost surface-charge when you’re doing a continuous test? To eliminate that possibility, charge the battery about a week before you use it for the tests.

9

Alkaline cells are good at recovering lost energy when rested. The other issue is the load is very low. Still within the okay spectrum of an alkaline. But if this was a 110 lumen light I think the results would be closer.

10

Thank you for excellent testing and reporting tek-0. Really good useful info. That will actually change the way I use cells.

That being said, I tend to recharge my Ni-Mh when I start detecting decreased lumens. That is probably before the lines meet below. With Alkaline, I get depressed and keep using the light until there is so little light that I really, really can't use the light any longer. Just hoping for the cell or cells to die a quick death. Probably some OCD of mine, but I just can't trash them before I get all the juice out of them. I do pull them to put in remote controls, but that stack of cells is way too big already.

11

I thought so too. I tested two 900mah eneloops, they measured 864 & 837mah in an analyzing charger and seem to work fine. At first I tested the 864mah one, but it started dropping fast at 35 min mark, and I switched to the 837mah one, which lasted longer and produced the graph I showed. Don’t know if there’s something wrong with the 864mah one… these are just a year old batteries.

I charged the batteries a day before starting the test.

12

Energy wasted firing up the driver, I’d guess.

I still use old Arc AAAs (a very early good LED flashlight, still available) and that design falls back to a low-like-moonlight brightness when the battery is very low.

Gransee who built them originally pointed out that, in that condition, you’d get more light by leaving the flashlight on than by turning it off and on when needed.

Point being that just getting the driver running took a surge of power from the battery, more than keeping the LED lit for quite a while at a time.

13

Well, even if that’s true (a 20% drop in run-time still sounds excessive even if a driver sucks up a whole lot of energy when turning on), why does the alkaline test not suffer the same way?

14

> even if that’s true

How the specific driver behaves would be the truth, for that particular driver. Gransee was describing the ones he built. Yours may differ.

> why does the alkaline ….

Different battery chemistry behaves differently — different rate and amount of voltage sag, ability to push current — under load.
You’d have to test the cell in the specific driver to see how long the driver pulls power from the cell before it gives up, and then how long til the voltage recovers.

I think you’d need specifics about the hardware to get answers to these questions.

15

you are right:

This light does not have much demand amps.
With 110 lumens AAA or AA with more than 200 lumens, will win in any case the NiMH. Alkaline are rubbish (and plus, little ecological)

16

nice work!
Your test does a very good job of showing that Alkalines will run longer than NiMh, at very dim levels (below 50% brightness)… :slight_smile:

I want my light to get a fresh battery any time its lumen output falls to 50%. (or sooner)
Which battery will give the most brightness up to that point?: NiMh

I personally have no use for an Alkaline that is producing long runtimes at less than 50% of full charge brightness. So to me, the test confirms that the NiMh will be brighter than the Alkaline for the runtime until brightness has fallen to 50%.

Because I will tend to “milk” an alkaline, to spare the expense of replacing it, I will be using the light at very dim levels. Otoh, because I carry a spare NiMh, any time my light gets dimmer than 50%, I just put in another battery, without fear of the expense, and no need to “milk” it at low levels.

Besides, NiMh cells dont leak and corrode the way Alkalines do…

NiMh for the Win!

17

Mmm, that's 3 days worth of self-discharge. I've seen Ni-MHs going from full to nearly empty in 2 weeks. Self-discharge could also be more pronounced the fuller is the battery… :???:

Cheers ^:)

18

Nice tests! Thanks for your findings.

Honestly I can’t remember the last time I used alkalines in a flashlight, and I’m slowly retiring my NiMh for li-ions.

19

Well, we don’t know the actual condition of the OP’s cells, but they are Eneloop Pro Black, the best Low Self Discharge NiMH cells around! As such, the 20% loss of capacity over three days is very suspect to me. Perhaps they need to be run through a refresh cycle or two?

20

Certainly much finer than jomertab's NCR18650BOOM…

Sry for this off-topic cheap joke.

will34, Ni-MHs may not have the energy density of li-ions, but there's still a lot of equipment designed to be used with them, and they certainly are waaay more fool-tolerant than li-ions. As with li-ions and their UltraTurd/GTFO range of representatives, main problem for the philistine masses is accessing enough advice and information to avoid buying BooTY crap and such for their devices (quote from a friend: “They're lighter than the old brand-name ones because they're made from light-alloy!”). LoL.

Cheers ^:)