Eneloop AAA Losing Major Capacity - Any Ideas Why?

[Ziemas]

I've recently given all of my batteries run through the 'break-in' mode of my Maha C-9000 to freshen them up and get an idea of their performance. Oddly three of my Eneloop AAA cells, all the same date code (06 09 TE), have lost major capacity.

I thought this might be a fluke and ran them through the 'break-in' mode three times with little change in measured capacity.

All of these cells have been used in low drain devices (a rear blinking bike light and small powered speakers for an MP3 player) and have only been charged on the C-9000. I don't know exactly how long I've had them, but it can't be more than two years.

Below are the measured capacity from a run through the C-9000 'break-in' mode when new and now.

800 - 674

799 - 670

790 - 656

Has anyone else experienced this? Any idea why these cells which have been well cared for should have such a large loss of capacity?

Thanks.

Z.

[45/70]

I can't really comment on AAA eneloops because I've only recently started using them (AAA). I can tell you, that NiMH AAA's in general, don't seem to last very long. I suppose it has to do with the "miniaturization factor", the smaller a device is made (all other ingredients remaining the same), the less durable it is.

I'm disappointed to hear your results. I was hoping that the eneloop AAA's would do better.

Dave

[Mr Happy]

They might have developed a resistance problem from the low drain loads. The C9000 is not very good at measuring cells with increased internal resistance. So the capacity might still be there but the C9000 can't measure it.

If you can, you might try running a long slow discharge on one of them outside the C9000 before doing another break-in charge. You want to do the discharge at a steady 50 mA or less, for example by connnecting a cell to a 20 ohm resistor and letting it discharge down to about 0.9 V. The idea of the slow discharge is to break up any large crystals that might have formed over time. I don't know if this works or not but it is what other people have recommended.

[45/70]

What a brilliant idea Mr H! It could be too, that the C-9000 is the actual problem. I've often wondered if the the C-9000 doesn't have a less than ideal algorithm for AAA cells. Hummm. It's worth trying a slow discharge, for sure.

Since Ziemas is using his cells in a low drain manner, this would actually discourage crystalline formation, and would tend to break it up during use.

I've run my under-performing non-LSD AAA's down at low rates on the CBA II, and while some crystalline formation was noted, determined that the cells were just used up. If the separator in AAA's is proportionately thinner than that in AA's, therein may lie the problem.

Dave

[TorchBoy]

... while some crystalline formation was noted ...

How exactly? Voltage recovery, I presume, but how much?

[45/70]

How exactly? Voltage recovery, I presume, but how much?

On the really bad ones, I didn't really see much voltage recovery to speak of. What happens though, is you can see the voltage spike up when the energy is released from the formations breaking up. They'll do that, then level off at the higher voltage for a while, then slowly creep downward, and then spike up again. This can go on for a considerable amount of time when you have a cell that only has the crystal problem. As I said though, with the cells I was referring to in my last post, there really wasn't much activity.

Dave

[TorchBoy]

How absolutely fascinating!

[Justin Case]

I've run my under-performing non-LSD AAA's down at low rates on the CBA II, and while some crystalline formation was noted, determined that the cells were just used up. If the separator in AAA's is proportionately thinner than that in AA's, therein may lie the problem.

Dave

Can you explain what you mean by "crystalline formation"? I thought that both the positive and negative electrode materials were already crystalline -- a nickel oxyhydroxide and an intermetallic, hydrogen storage, metal alloy. Do you mean the physical changes in the electrode material that can occur from repeated, shallow discharges, e.g., beta to gamma NiOOH transformation, and MH crystal growth?

If the OP's "low drain devices (a rear blinking bike light and small powered speakers for an MP3 player" were not designed for NiMH, they could have run the cells only to shallow discharges due to an inappropriately high cutoff voltage, which will produce an apparent loss of capacity due to the physical changes in the active electrode materials. If this is the case, you need to run the cells through a few full discharge-charge cycles to restore the cells to full capacity again.

Another possibility is excessive trickle charging, which also can transform the beta-NiOOH to gamma-NiOOH. Multiple full discharge-charge cycles can address this as well.

In NiMH cells, the negative electrode is typically designed to have excess capacity compared to the positive electrode to handle both overcharge and overdischarge. Thus, the positive electrode generally dictates a cell's useful capacity for NiMH.

[45/70]

Can you explain what you mean by "crystalline formation"?

I'm referring to crystalline formations in the electrolyte itself. Here is a .pdf that is a good read for anybody interested in NiMH cells (or lead acid as well), written by an undergrad at OSU. Relevant to your question, is article page 31 (.pdf page 40) Crystalline Formation. There are some electron microscope pictures of the actual formations as well.

My own research doesn't necessarily agree with everything he says in this article. My only pedigree is an AAS in Aviation Maintenance Technology. But, in my experience, it's usually the AASes that tell the BS'ers what they did wrong.

Dave

[Justin Case]

I'm referring to crystalline formations in the electrolyte itself. Here is a .pdf that is a good read for anybody interested in NiMH cells (or lead acid as well), written by an undergrad at OSU. Relevant to your question, is article page 31 (.pdf page 40) Crystalline Formation. There are some electron microscope pictures of the actual formations as well.

My own research doesn't necessarily agree with everything he says in this article. My only pedigree is an AAS in Aviation Maintenance Technology. But, in my experience, it's usually the AASes that tell the BS'ers what they did wrong.

Dave

Figure 18 on page 31 of the thesis shows crystal grain growth of the negative electrode material (cadmium hydroxide) in a NiCd cell. It's not really "crystalline formation", which is the term used in the linked thesis. It is crystal grain growth. In fact, the thesis indirectly states this:

"The crystals have grown from around 1 micron to 50-100 microns. After a pulsed charge, the crystals were reduced to 3-5 microns."

Note the thesis text I highlighted in bold. The Cd(OH)2 is already crystalline, with starting size of about 1 micron. The crystals grow to about 50-100 microns. These crystals are "formations in the electrolyte itself" only in the sense that the negative electrode material absorbs the alkaline electrolyte within the electrode pores. What is happening is overly large crystal growth of the negative electrode material itself, reducing the electrode's available active surface for electrochemical reaction. This was known at least as far back as the 1970s.

[45/70]

OK Justin, apparently you're nitpickier than I am. Perhaps I should have said "large crystalline formation". Yes, the growth is on the electrode and, as you said, the building material comes from the electrolyte. I did not realize that the problem was the "large" formations reducing the electrodes surface area.

Whether in NiCd, NiMH, or their chemical composition, the "Large" crystals cause the energy to be withheld (except at low discharge rates). These were what I was concerned with, in trying to restore under-performing cells, that in some cases did, or in other cases apparently did not have them.

I learned something today. Thanks.

Dave

[Justin Case]

OK Justin, apparently you're nitpickier than I am. Perhaps I should have said "large crystalline formation". Yes, the growth is on the electrode and, as you said, the building material comes from the electrolyte. I did not realize that the problem was the "large" formations reducing the electrodes surface area.

Whether in NiCd, NiMH, or their chemical composition, the "Large" crystals cause the energy to be withheld (except at low discharge rates). These were what I was concerned with, in trying to restore under-performing cells, that in some cases did, or in other cases apparently did not have them.

I learned something today. Thanks.

Dave

Has nothing to do with being nitpicky. Words have meaning, and they especially have meaning in the specific discipline of materials science which deals with things like batteries.

"Formation" suggests either nucleating a brand-new phase or else growing it from an existing phase. If it's the latter, then you probably should have called it grain growth in the first place. If it's the former, then nucleation has completely different kinetics and mechanisms than growth.

So the terms "crystalline formation" vs "grain growth" have totally different meanings.

[45/70]

Has nothing to do with being nitpicky. Words have meaning, and they especially have meaning in the specific discipline of materials science which deals with things like batteries.

"Formation" suggests either nucleating a brand-new phase or else growing it from an existing phase. If it's the latter, then you probably should have called it grain growth in the first place. If it's the former, then nucleation has completely different kinetics and mechanisms than growth.

So the terms "crystalline formation" vs "grain growth" have totally different meanings.

Ah yes, I think I see what you mean. "grain growth", now that has an intellectually pleasing sound to it. It even resembles some form of conceptual continuity as to what is actually happening within the cell. I like it!

The biggest problem I see, is trying to convince the various Forums, blogs, battery manufacturers, battery charger manufacturers, battery device manufacturers, as I recall, the U.S. D.O.E. among others, and last, but not least, me, who all call it "crystalline formation", that this phenomenon should more accurately be called "grain growth". It seems to me like you've got your work cut out for you.

Seriously, at this time, I'm going to return this thread back to Ziemas and his question about his eneloop AAA's. This concerns me, as I've had good success with my AA eneloops, and I hope an answer can be found as to why his cells seem to have crapped out.

For now, I'm going to nucleate some hot dogs, as I forgot to get charcoal on the way home, in a materialistic science mechanism, called a microwave oven, and have my dinner (does that mean I'll get puppies, or just one great big dawg? ).

Mr. Case, please, by all means, if you do decide to pursue your, in my personal opinion, hopeless endeavor, do let us know how it turns out. Good luck!

Dave

[Ziemas]

Can you explain what you mean by "crystalline formation"? I thought that both the positive and negative electrode materials were already crystalline -- a nickel oxyhydroxide and an intermetallic, hydrogen storage, metal alloy. Do you mean the physical changes in the electrode material that can occur from repeated, shallow discharges, e.g., beta to gamma NiOOH transformation, and MH crystal growth?

If the OP's "low drain devices (a rear blinking bike light and small powered speakers for an MP3 player" were not designed for NiMH, they could have run the cells only to shallow discharges due to an inappropriately high cutoff voltage, which will produce an apparent loss of capacity due to the physical changes in the active electrode materials. If this is the case, you need to run the cells through a few full discharge-charge cycles to restore the cells to full capacity again.

Another possibility is excessive trickle charging, which also can transform the beta-NiOOH to gamma-NiOOH. Multiple full discharge-charge cycles can address this as well.

In NiMH cells, the negative electrode is typically designed to have excess capacity compared to the positive electrode to handle both overcharge and overdischarge. Thus, the positive electrode generally dictates a cell's useful capacity for NiMH.

They've been through three break-in and then discharge cycles on the C-9000 with no change in capacity. They've not been excessively trickle charged either.

[clintb]

I'm not 100% convinced that all cells will respond to a break-in, or even to repeated cycles, at a low rate of charge / discharge. I have some questionable cells; once's I'd not cry over if they need to be recycled. I took those cells through two break-in cycles on the C9000; no gain in capacity. I then put them through a "Refresh & Analyze" at just over .5C charge and the same for discharge; no gain, again. Next time, I decided to hammer them, so I did a 1C charge and discharge. Guess what? They came back to almost rated capacity.

[Ziemas]

I'm not 100% convinced that all cells will respond to a break-in, or even to repeated cycles, at a low rate of charge / discharge. I have some questionable cells; once's I'd not cry over if they need to be recycled. I took those cells through two break-in cycles on the C9000; no gain in capacity. I then put them through a "Refresh & Analyze" at just over .5C charge and the same for discharge; no gain, again. Next time, I decided to hammer them, so I did a 1C charge and discharge. Guess what? They came back to almost rated capacity.

Thanks for the info. I'll give it a try and report back.

[Justin Case]

Ah yes, I think I see what you mean. "grain growth", now that has an intellectually pleasing sound to it. It even resembles some form of conceptual continuity as to what is actually happening within the cell. I like it!

The biggest problem I see, is trying to convince the various Forums, blogs, battery manufacturers, battery charger manufacturers, battery device manufacturers, as I recall, the U.S. D.O.E. among others, and last, but not least, me, who all call it "crystalline formation", that this phenomenon should more accurately be called "grain growth". It seems to me like you've got your work cut out for you.

Seriously, at this time, I'm going to return this thread back to Ziemas and his question about his eneloop AAA's. This concerns me, as I've had good success with my AA eneloops, and I hope an answer can be found as to why his cells seem to have crapped out.

For now, I'm going to nucleate some hot dogs, as I forgot to get charcoal on the way home, in a materialistic science mechanism, called a microwave oven, and have my dinner (does that mean I'll get puppies, or just one great big dawg? ).

Mr. Case, please, by all means, if you do decide to pursue your, in my personal opinion, hopeless endeavor, do let us know how it turns out. Good luck!

Dave

Your snide remarks demonstrate a remarkable ignorance, fairly typical of UI people like yourself. I note that you seem incapable of discussing the factual aspects of the definitions of the terms you are using. The fact is that formation <> growth, and it is clearly grain growth that is the operative mechanism. Your &quot;proof&quot; that &quot;crystalline formation&quot; is correct terminology is an appeal to authority argument, which is fallacious. I can find plenty of references that refer to grain growth, not &quot;crystalline formation&quot;, from DOD and academia. So what.

Not sure what endeavor you refer to. If you mean enlightening you, you are finally correct in something -- it is hopeless.

[Bones]

I'm not 100% convinced that all cells will respond to a break-in, or even to repeated cycles, at a low rate of charge / discharge. I have some questionable cells; once's I'd not cry over if they need to be recycled. I took those cells through two break-in cycles on the C9000; no gain in capacity. I then put them through a "Refresh & Analyze" at just over .5C charge and the same for discharge; no gain, again. Next time, I decided to hammer them, so I did a 1C charge and discharge. Guess what? They came back to almost rated capacity.

There does appear to be a consensus that when a cell fails to respond to a break-in charge, the next step is to ramp-up the charge rate.

I have four eneloop AAA cells that have always been used in a computer mouse and only subjected to break-in charges; usually after the mouse stopped working. While the capacity remained okay, the spread in the time it was taking them to fully discharge after a full charge on my original edition MH-C9000 was gradually increasing:

http://www.candlepowerforums.com...post3044873

In an attempt to offset this, I'm in the process of subjecting the cells to three .5C charges followed by three 1C charges, each preceded by a very slow 100mA discharge.

Although I've only completed two of the six cycles thus far, the spread between the time it takes the first and last cell to fully discharge has already decreased.

The cells were put into service in December of 2006.

[Black Rose]

I'm not 100% convinced that all cells will respond to a break-in, or even to repeated cycles, at a low rate of charge / discharge.

I've got some LSD AA & AAA cells that needed some work when I got them.

Break-in didn't do much for them, but repeated Cycle modes (12 - 15) at .5C charge and .25C discharge did more for them than the break-in did.
I also tried doing a 1C/.5C charge/discharge on them (AA), but they didn't respond to that at all.

They still fall in Silverfox's "crap" category, but they are fine for non-critical use.

[clintb]

I now remember the cells! I *thought* they were some of my crappier ones, but turns out I was wrong, dead wrong.

4 x Sanyo / GE 900mAh AAA
4 x Sanyo 1000mAh AAA

I guess I slapped "crap" on them because they were coming in at such low capacities after break-in and R&A. Still though, coming back, close to rated capacity, after being subjected to 1C charge and discharge is mighty impressive for a AAA.

[45/70]

From what you guys are saying, it's clearer to me now, how the Sanyo Harmolattice cells would be a viable alternative to regular LSD cells. I don't run my newer cells in low drain devices. I just use older cells that they have replaced, or "crap cells". In these devices, they've always worked quite well.

Most devices that I use my newer cells in all draw enough current that, I suppose anyway, is why I've not run into the problems Ziemas and others here have experienced. Thinking more about how I use AAA cells, I run proportionately higher current from them than my AA's, eg. 2C+ at times, and I'm sure that has a lot to do with their seemingly short cycle life.

Justin Case, I see my attempt at a humorous post failed. Sorry about that. I'm sure you and many others here have more knowledge than I about cell chemistry, behavior and so on. I was just messin' wid ya!

Your snide remarks demonstrate a remarkable ignorance, fairly typical of UI people like yourself. I note that you seem incapable of discussing the factual aspects of the definitions of the terms you are using. The fact is that formation <> growth, and it is clearly grain growth that is the operative mechanism. Your &quot;proof&quot; that &quot;crystalline formation&quot; is correct terminology is an appeal to authority argument, which is fallacious. I can find plenty of references that refer to grain growth, not &quot;crystalline formation&quot;, from DOD and academia. So what.

Not sure what endeavor you refer to. If you mean enlightening you, you are finally correct in something -- it is hopeless.

So anyway, thanks for the kind words, and lighten up man!

Dave

[Ziemas]

UPDATE:

After five rounds of a 800mah charge and discharge cycle there is no discernible difference in capacity. I've now got them on a 'break-in' cycle on the C-9000 to see what the rated capacity will be to compare apples to apples, but I have little hope of seeing any improvement.

I must say I'm rather disappointed that such expensive cells, which have been properly taken care of, should have such a short life.

[45/70]

Ziemas, I'm curious, do you have any idea what the voltage was of your cells, after discharge in your light and speakers? I ask because I'm wondering if it's possible that some were over discharged and thus, damaged a bit.

I know when using cells in series in low drain devices, my IR mouse for example, it's easy for one of the cells to go way to low, before you're aware of it. If this happened repeatedly, even just a little, this may explain why some of your cells are coming up short.

Dave

[Ziemas]

Ziemas, I'm curious, do you have any idea what the voltage was of your cells, after discharge in your light and speakers? I ask because I'm wondering if it's possible that some were over discharged and thus, damaged a bit.

I know when using cells in series in low drain devices, my IR mouse for example, it's easy for one of the cells to go way to low, before you're aware of it. If this happened repeatedly, even just a little, this may explain why some of your cells are coming up short.

Dave

Honestly I have no idea, but I charged them long before they ran out of juice, so I never ran them all the way down.

[Bones]

Honestly I have no idea, but I charged them long before they ran out of juice, so I never ran them all the way down.

Considering the discussion in this and other threads respecting the discharge cut-off point of the MH-C9000, perhaps you should consider some form of deeper discharge of the cells as suggested by Mr Happy in post 2.

If you've never let them fully discharge, and charger can't discharge the cell deep enough considering it's 0.9 volt cut-off, it may make a difference.

As well, it's my understanding that optimum crystal break-up is achieved through fast charging and slow, deep discharging, as set out in this post by SilverFox:

http://www.candlepowerforums.com ... post2391945

One way I've tried to achieve a deeper discharge is by repeatedly discharging a cell three or four times in row at the MH-C9000's lowest discharge rate. If the cells are allowed a rest period in between discharges, they seem to recover enough for another measurable discharge.

[alpg88]

i thought crystallization only happen in nicd.
in nicd i fix it quick and easy, flash circuit from disposable camera, and 2 nails, i zap each cell 2-3 times, than it becomes almost as good as new, never tried to fix nimh, or liion like that.

[45/70]

i thought crystallization only happen in nicd.
in nicd i fix it quick and easy, flash circuit from disposable camera, and 2 nails, i zap each cell 2-3 times, than it becomes almost as good as new, never tried to fix nimh, or liion like that.

Hey alpg88, you are confusing dendrites with crystallization, as I once did (checkout the thread in Bones link). Crystallization does however, occur in both NiCd and NiMH.

Dave

[45/70]

Honestly I have no idea, but I charged them long before they ran out of juice, so I never ran them all the way down.

Well, if you had run them all the way down, depending on the circuitry involved (I'm guessing simple DD in both instances), that certainly would have caused damage. I just thought maybe there may have been some instances where a "borderline" situation had occurred. Sounds like you have made sure that didn't happen though.

Dave

[Ziemas]

Well, I've tried all of the suggestions here and there is no change at all. I guess the batteries are ruined. Quite disappointing, really.

[TorchBoy]

I guess the batteries are ruined. Quite disappointing, really.

Yes, it's very sad but with only 674 mAh they're just no good. You'll have to send them to me.

No, seriously, thanks for this thread. I've got a plan of attack for some friends' cells.