7 years of EDC and still at 82% of rated capacity
"I bought three of these in early 2019 for a Fenix PD35 V2. 0 that I carry every day. I labeled them 1, 2, and 3 so I could keep them in a proper rotation, and they've been going round that rotation ever since — swapped roughly once a month, usually before the light got dim rather than running them flat. I'd guess I typically used somewhere between half and three-quarters of a charge before swapping, though I wasn't measuring at the time. One or two got run down hard over the years. None of them has ever swelled, split a wrapper, or corroded. This week I finally ran the numbers on them. All three measured between 2832 and 2888 mAh, against the 3500 mAh rating — about 82%. They were within 2% of each other, which surprised me more than the number itself. After seven years of independent use, I expected at least one to have fallen behind. I can't tell you they've lost 18%, because I never measured them when they were new. What I can tell you is that after seven years of real use, they still deliver 82% of what's printed on the wrapper. For what that means day to day: these have never let me down and have never quit prematurely. If I were buying today, I'd want to know a cell would still be useful in year seven, and on this evidence these are. Do I wish there was a USB-C version? Yes — micro-USB is showing its age. But I'd buy more of these without hesitation. I just don't need to. Maybe at ten years they'll need replacing — I'll test them again and find out. HOW I TESTED, AND THE FULL NUMBERS All three cells were charged to full on a SkyRC MC5000 analyser — not the built-in USB port — then rested four hours before testing. Ambient was 78F. Each was discharged at 1750 mA, which is 0. 5C for a 3500 mAh cell, down to a 3. 20 V cutoff. All three ran simultaneously, one per slot. Data was logged every two seconds by software I wrote for the MC5000 to handle this kind of testing and analysis. Capacity is calculated two independent ways: integrated from the logged current, and read from the charger's own counter. Both are shown. They agreed to within 0. 15% on every cell, which is the main reason I trust the numbers. Cell 1 — 2853 mAh, 10. 05 Wh, mean 3. 523 V, 67 mOhm, ran 2. 02 h (integrated 2852. 9 mAh vs charger 2853 mAh, 100. 00% agreement) Cell 2 — 2888 mAh, 10. 17 Wh, mean 3. 525 V, resistance not measured, ran 2. 21 h (integrated 2884. 6 mAh vs charger 2888 mAh, 99. 88% agreement) Cell 3 — 2832 mAh, 9. 98 Wh, mean 3. 531 V, 53 mOhm, ran 2. 03 h (integrated 2827. 8 mAh vs charger 2832 mAh, 99. 85% agreement) Each cell delivered about 10 Wh at a mean voltage of 3. 52 V. That's a more useful comparison figure than mAh alone when cells have different voltage curves — two cells with identical mAh can deliver noticeably different energy if one sags harder under load. One anomaly worth flagging. About 53 minutes into the run, cell 2's measured voltage dropped 56 mV in a single step and stayed there, while cells 1 and 3 showed no such shift. The discharge slope either side of the step was unchanged, so this looks like a change in contact resistance in that slot rather than anything happening inside the cell — roughly 32 mOhm appearing in the measurement path at 1750 mA. Cell 2 also returned an impossible internal resistance reading at the start of the run, which points the same way. That matters because it probably explains why cell 2 measured highest. Extra resistance in the path means the charger reaches the 3. 20 V cutoff while the cell still holds charge, then recovers it slowly during the taper. Cell 2 had the shortest full-current phase of the three and by far the longest taper — 46 minutes against 34 and 36. I don't think it's genuinely the strongest cell; I think its slot had a contact problem. On the taper, and why these numbers may not match anyone else's. About 11% of the total came out after the cell had already reached 3. 20 V. The charger holds that voltage and winds current down — from 1750 mA to under 100 mA — rather than stopping dead. A tester that cuts off hard at 3. 20 V would report roughly 2500 mAh for these same cells. Neither approach is wrong, but they're not comparable. That taper is longer here than on unprotected cells I've tested, and I think the protection circuit explains it. Voltage dropped 0. 127 V the instant load was applied, which means real resistance in the path — some of it the cell, some of it the protection board. That makes the cell hit 3. 20 V while it still holds usable charge, so more of the capacity comes out slowly at the end. Bare cells I've tested sag far less and taper for about a third as long. Measured at the point where current first started dropping, the three cells read 2545, 2501, and 2500 mAh — closer together than the totals suggest, and the fairer comparison between them. What I can't tell you. The charger measures internal resistance once, a few seconds into a run, then reports that figure unchanged. Cell 1 read 67 mOhm and cell 3 read 53 mOhm, both plausible for a protected cell. Cell 2's single measurement came back at 1 mOhm, which is impossible, so I've discarded it rather than publish it. I also have no way to verify the charger's absolute accuracy — I've checked that its numbers are internally consistent, which is not the same as checking they're correct. And I never measured these cells when they were new, so I can't tell you how much capacity they've lost. Only that they now deliver 82% of the rated figure. The chart has a light median filter applied to smooth single-sample noise. All figures above are calculated from the raw data."