Curious case of the dead light that might not be??

I use all of my lights, a lot, and thought I was having my first failure the other day, in a Fenix LD25. I noticed it was getting a lot warmer in use than normal, and I was having to charge a lot more into the batteries when I changed them each day compared to what I used to for the same amount of runtime use. It was happening with all 4 pairs of cells I cycle through this light (and they are also used in a Quark Turbo AA2). The LD25 and Quark get exactly the same amount of use each day as bike lights, and it used to be that the amount I charged into the 2 pairs of cells each day were within about 200mAh of each other. Then I noticed the difference was more like about 500-600mAh. The cells are Imedion 2400 with 120-140 cycles on them doing this duty, and they get beaten about a bit because they're attached to a bike. They've taken a few knocks because they're not quite as long as they used to be...

Give or take I usually need to charge about 50% back into them each day. Now I have to charge an extra 500mAh or so into whichever pair was in the LD25 vs a relatively normal about for the Quark.

A runtime test on the LD25 at home with the "bike" batteries only gave me about 45 minutes runtime in high, vs at least 1:35 if not longer normally. But a Refresh&analyse (1000mA/400mA) on the c9000 said the cell's capacity was about 97% of their original capacity. The 1A pulsed discharge on the C9000 is less than the 2 torches though, I guess. Discharging on the C9000 didn't show a particularly reduced voltage (from high resistance). So I thought it was the torch failing but on a whim tried some less abused NiMH's and voilà, was back to the expected runtime.

This is all very curious because

a) The same cells get used in the Quark Turbo and its doesn't seem to be exhibiting the same behaviour even though it has a lower runtime (hence higher current draw) in turbo mode compared to an LD25

b) The cells don't show any of the normal signs of a poor battery on the C9000 (either capacity of <85% or high internal resistance, or voltage drop under load) ... yet

So I dunno what's going on. Any ideas? Something in the way the circuits on the 2 different lights work that makes the LD25 more susceptible to batteries that could be on their way out? eg. a higher current pulsed drain of some sort? A fault in the LD25? I've still got a month or so of warranty left.

Happy to hear some theories :thinking:

Your LD25 has better regulation than the Quark.

If you look at the Quark review:
http://www.candlepowerforums.com/vb...123-QAA-Q123-2-QAA-2-RUNTIMES-COMPARISONS-etc
the Quark has worse regulation than the LD20 (should be close enough to the LD25) on L91 / Alkaline batteries. Your old NIMHs will have higher internal resistance and look more like those primary cells to the flashlight. Your Quark won't run at full brightness and consume less power...

Sounds like a battery problem. you say they have taken a few hits because they're shorter than used to be... maybe they're not making proper contact.

BringerOfLight that's a pretty interesting theory. In those graphs though, the quark had both lower output *and* lower runtime than the ld20 on the high resistence alkalines. That's not what I'm seeing with my old nimh's, I get longer runtime in the Quark Turbo with the dodgy batteries compared to the LD25. With fresher batteries, I get longer runtime with the LD25.

I did another commute on the bike last night and this is what I used and had to be charged into the cells (the first column is the battery's label):

Q1 (old) - ld20 - 787mAh (slot 1 of c9000 :shakehead)
Q2 (old) - ld20 - 844mAh
Q3 (old) - quark - 903mAh
Q4 (old) - quark - 919mAh
L1 (fresh-ish) - ld25 - 805mAh
L2 (fresh-ish) - ld25 -784mAh

But if I load any of the 4 well worn "Q" batteries into the ld25, I would need to be charging in about 1400mAh for the same amount of use. So there's something about the LD25 that is different to the ld20 and quark.

Note, the lights are the LD20 R4, Quark Turbo R5, and LD25 R4 models.

Cataract, all of the lights have a spring in the tailcap and seem to make good contact. If it was making poor contact, wouldn't it use less mAh, not more?

bbb74: yes, poor contact would mean less current going through. It could also mean that when the voltage is getting low, not enough current goes through and the flashlight thinks the battery is weak, shutting off before time.

Yours sounds like a peculiar case and I would suggest testing with new NiMh batteries. You can't really compare a Quark with a Fenix, though, as they are not using the same driver, but you should get runtimes very close to those stated by the manufacturer on good cells.

Nah it won't be that then, because when I go to charge the cells they need way more charge (they have had capacity drained out of them), or if I discharge them there is less remaining then there "should" be. I'm just having to run with fresher batteries in the LD25 now. The quark turbo and ld20 don't seem to care about the more heavily used cells, and the c9000 doesn't mind them either.

Still interested if anybody can explain what the ld25 is doing electronically to get like this...