Cycle Testing Observations…

[bob_ninja]

NiOOH,

I have been having a lot of problems with Energizer 2500s. In one high draw application they kept collapsing early, like to almost 0V. Now I removed them from that use. In a music box for my kid they seem to have developed a memory effect and had to cycle them on BC900 to bring them back.

I am using Sanyo 2500s for high draw and they are doing much better. Also, my older 2K and sub-2K cells are fairly reliable. I wonder if I should avoid 2K+ cells altogether. Seems manufacturers are sacrificing a lot of other attributes (such as cell lifespan) in order to achieve 2.5K+ capacity.

I even paid too much for Eneloops just so that I can have a reliable standby batteries I don't have to top off all the time.

I use Sanyo charger for the initial charging @1amp until they batteries warm up. Then I transfer them to BC900 @0.5amp to finish at lower rate, cooler.

Anyway 100+ cycles for 2.5amp batteries is a joke. I consider any battery that handles less than 300 cycles a failure.

Oh, and looking forward to an update; thanks Tom

P.S.: I am almost tempted to test a 2K cell on BC900 at 0.5 charge, 0.25 discharge (very slow rate based on Tom's scale

 

[BentHeadTX]

More good info from SilverFox as always
My battery eater is a MillerMods UWAJ 1.7 watt L1P bicycle helmet light. Those poor AA cells get hit with 1.7 amps of current draw and I have noted the 2500's just don't have a decent capacity for this.
Figure I'll go with the Titanium 1800mAH cells for high current drain applications, if it will do 18 amps--it should not have a problem with 1.7 amps. Must keep the voltage from sagging to Miller's constant current regulator to keep the light bright.

 

[NiOOH]

Hi Tom.

My observation on the MQH01E (the European model that is) show that it does have a trickle charge. Thi is how I tested it:
As soon as the light goes out, i.e the charger has finsihed, I removed one cell and measured its OCV. It measured 1.42V. I waited for 4 hours and then measured again. The same cell showed 1.38. The remaining 3 cells were left in the charger. I took them out and measured them. All showed 1.41-1.42 V. The only conclusion I could make was that there is a maintenance charge on my MQH01E. Also, the charger continued to make these thicking noises even after the light goes off.
The easiest way to to test this would be to connect a multimeter to one of the slots. I've been thinking to do that, but couldn't find thin enough copper plates. If you have some improvised setup, handy, you may try it. Otherwize, I a'm planning ti go to the hardware store the comming weekend to do an extensive shopping and may put something together

 

[bob_ninja]

NiOOH,
I thought I read some place that it does provide trickle charge after the charge light turns off.

 

[NiOOH]

Hi Bob.
So far my measurements show that it does indeed. I'll do some more testing soon.

 

[bob_ninja]

NiOOH,
I can no longer find info on trickle charge (might have read it in an older post here). The only indication of tirckle charge presence is in the decription of the LED:
"OFF stand-by or charge complete"
From the Kodak K6000 (which appears to be the same charger)
"LED off = power off, not charging, charger finished"

Since it makes a distinction between "not charging" and "charger finished", I would assume finished mode is using tickle charge.

 

[SilverFox]

Hello NiOOH,

I conducted the same test as you did and I do not think it does a trickle charge.

I took two matched cells, I charged one of them on the Sanyo charger and the other on the BC-900. At the end of the charge, the cell on the Sanyo charger was at 1.45 volts, and the cell on the BC-900 was at 1.46 volts. 12 hours later, the cell on the Sanyo charger is reading 1.40 volts, and the cell on the BC-900 (which is trickle charging at roughly 17 mA) is reading 1.51 volts.

I then removed the cells from the chargers. Measuring them after 4 hours I show 1.40 volts for the cell that was on the Sanyo charger, and 1.42 volts for the cell that was on the BC-900 charger.

My conclusion is that the Sanyo charger (at least the one that I have) does not trickle charge.

Tom

 

[NiOOH]

Hi Tom. Thanks for the test. This weekend I'll conduct sme more testing with mine. I guess there is a difference between the chargers. Mine is MQH01E, i.e the European modification. I believe yours is the U-model that does a cell check on slot #1. Mine does not have this feature, but trickle charges instead. I'll make some flat plate electrodes and will connect a multimeter to it.

And some comment on your measurements. I assume you set the BC900 to charge at the same current as the Sanyo. Was it the slow slot on the BQH01 (ca 1 A) compared to 1 A charge on the BC900? I ask because my BC900 maintains higher trickle current after charging at 1 A. I think it was 50 mA. Also your voltage reading seems a bit high for a cell that is trickled at 17 mA. 1.51 V looks more like a peak voltage during quick charge. Did you get an agreement between your measured values (I suppose using a multimeter) and the value shown on the display of the BC900?

 

[SilverFox]

Hello NiOOH,

Yes. My charger has the state of charge tester, but it is in slot 2.

I used the 1.0 amp charge slot on the Sanyo charger and charged the other cell at 200 mA on the BC-900. All voltage measurements were taken with my Fluke meter. The cell on the BC-900 had finished charging about 10 minutes prior to me taking the voltage measurement.

12 hours on trickle charging does have an impact on the cells. That is why I don't recommend leaving cells to continually trickle charge.

Tom

 

[NiOOH]

Hi Tom.
Thanks for the info. Somehow I don't thing this test is conclusive. You were using two chargers at quite different charging currents. What I did was to use only the Sanyo with 2 matched cells. One was removed shortly after the light went off and its OCV measured within 10 sec. The other was left for sevral hours on the charger and then taken off and measured. The one remaining on the charger showed higher OCV, which led me to believe that there must have been some current flowing to it.

 

[SilverFox]

Hello NiOOH,

On the contrary...

The purpose of the test was to check for the presence of a trickle charge. These cells end up with a voltage of around 1.5 volts under the presence of a trickle charge and drop down to around 1.4 volts with no trickle charge. The cell on the Sanyo came off the charger (after 12 hours) at around 1.4 volts. The cell on the BC-900 with a very low trickle charge came off at around 1.5 volts.

Just for fun, I took the cell that came off the Sanyo charger and placed it in the BC-900. It registered as fully charged and is trickle charging at about 16 mA. The voltage has climbed to 1.49 volts.

So, if the Sanyo charger had a trickle charge, the cell should have been close to 1.5 volts. Since it didn't, I can conclude that the Sanyo charger does not trickle charge.

Tom

 

[SilverFox]

Update:

I have posted the data from the testing on the Titanium 2000 mAh cells in the first post. They handled the cycle testing better than the Sanyo 2500 mAh cells did.

Tom

 

[SilverFox]

I have been thinking about this test. I often find that you set out to find the answer to a question, but during, and after, the testing, more questions come up.

There are two influences at work during these tests. Charge rate, and discharge rate. I don't believe the 1C discharge rate has a drastic effect, so that leaves the heat generated during the charge cycle.

I recently read an article suggesting that the anode and cathode materials are not thermally balanced. The have a different rate of thermal expansion. This allows some sliding back and forth, that can wear down the separator. When the separator wears thin enough, soft shorts start to form, and the battery capacity goes down and the self discharge rate increases.

In the quest for higher capacity cells, the manufacturers are thinning down the thickness of the separator. Add to this some heat from the rapid charging and we have greatly reduced cycle life.

I find it interesting that the battery manufacturers are giving us higher capacity cells, along with 15 minute chargers. Perhaps this is a way to increase the sales of batteries...

I decided to take a look around and see if I could find some 2000 mAh cells. They are hard to find. Then I remembered the Eneloop cells.

I find it interesting that Sanyo came to market with 2000 mAh cells. They are advertising 1000 cycles along with their very low self discharge rate. This suggests, to me at least, that engineers had a lot of influence over these cells.

I also find it interesting that the other low self discharge cells are labeled with slightly more capacity, and a slightly higher self discharge rate than the Eneloop cells.

Could it be that 2000 mAh is the magic number?

I have been hearing about high rates of self discharge with high capacity cells. In the RC crowd, they are finding that the newer high capacity cells are fragile, and while they do deliver longer run times, they need additional care, or the cells die.

Perhaps our quest for higher capacity is a little misguided. We may want batteries that are more robust, at the expense of some capacity. In marketing, higher numbers seem to sell better, so I guess there would have to be a paradigm shift to get the manufacturers to go back to 2000 mAh cells.

Tom

 

[NiOOH]

Hi Tom.

I share your toughts. When Sanyo 2500 mAh cells came on the market I was excited, and got a couple of industrial-grade sets at once. Now, a year and a half later, I can no longer rely on them for critical applications. They'll be going to long term strorage, probably never to return in service. I replaced them with with a set of 2100 mAh cells and another that is 1700 mAh, both from Sanyo. They could still be found over here. The interesting point is that they are made in China. Only the 2500 and 2700 mAh cells are still made in Japan. I guess the Eneloops will be too.

 

[jtr1962]

Those are interesting observations, Tom. The quest for ever higher capacities reminds me of the MHz/GHz race in the computer industry. Anyone familiar with microprocessors knows that a faster clock speed doesn't necessarily correlate with faster day-to-day operation. There are lots of other underlying variables. Unfortunately, the average person on the street thinks a 3GHz machine is better than a 2GHz one so the computer manufacturers aim for higher clock speeds instead of more operations per clock cycle. However, a distinct disadvantage of these higher clock speeds is more heat and more complex pcb design. It's only in the last year or so when physics has begun to place an inherent limit on clock speeds that we have started to use things like dual-core processors.

Rechargeable cells seem to be following a similar trend. The average person on the street knows little about self-discharge (at least until they go for a camera which has been sitting a few weeks only to find the batteries are dead) or cycle life. However, they are easily wowed by large capacity numbers. It's only now that cell capacity is approaching the inherent physical limits of NiMh chemistry that cell manufacturers have needed something else besides capacity to distinguish their cells. Hence, the low self-discharge cells like Eneloop. Interestingly, it was probably the very high self-discharge rates of the ultra-high capacity cells which created the need for cells with a lower self-discharge in the first place. And 15 minute charging has been done I suspect largely for marketing reasons as well since the marketplace needed another thing to distinguish cells other than capacity.

My take on all this is that if 2000 mAh is really the magic number for very low self-discharge then I'd rather that the manufacturers just concentrate on making such cells cheaper and better, and forget the higher capacity ones entirely until they can produce more robust, low self-discharge versions. A 2500 mAh cells which requires constant babying and only lasts 100 cycles isn't terribly useful to me. From my perspective, 2000 mAh is plenty. I remember well the days of 450 or 500 mAh AA Nicads. A cell with over four times the capacity and which retains 85% of capacity for a year is a quantum leap compared to what I used 15 years ago. If a 2500 mAh version can be made with these charactersitics, wonderful. If not, I'm more than happy with a reliable 2000 mAh cell rather than a finicky 2500 mAh one.

 

[NiOOH]

If a 2500 mAh version can be made with these charactersitics, wonderful. If not, I'm more than happy with a reliable 2000 mAh cell rather than a finicky 2500 mAh one.

I agree 101%.

 

[NiOOH]

In order to achieve higher capacity, the engineers had to sacrifice cycle life. It has to do with the chemical and electrochemical reactions taking place, inside a NiMH cell. I'll try to explain why without involving too much chemistry.

The capacity of a NiMH cell is determined by the capacity (the amount) of the positive electrode, which is made of nickel hydroxide. However, there is an excess of a negative MH-alloy material inside the cell. During charging, the positive electrode becomes charged first and enters overcharge. This is accompanied with the emission of oxygen-gas from the positive electrode. This gas passes the separator and is recombined on the negative, MH electrode via a couple of chemical and electrochemical reactions. Mind that this is not yet an overcharge of the whole cell, and is necessary for the proper charging and sealing of the cell. If you look carefully at the internal pressure and temperature profiles during fast charging, you'll see that both parameters start to increase sharply at approximately 80 % state of charge (SOC). This is exactly the point when the positive electrode is fully charged and the process of oxygen recombination on the negative electrode begins. At 100% SOC, both electrodes are fully charged and the negative electrode's ability to recombine oxygen is decreased. The positive electrode, however, still produces oxygen at the same rate. If quick charging is not terminated, the internal pressure and temperature rise catastrophically, and the cell eventually vents.

Now, the only way to increase the capacity of a NiMH cell is to increase the amount of the positive electrode material. Since the cell volume for a given size is the same, the way to go is to decrease the amount of material at the negative electrode. Thus, the cell's ability to absorb overcharge is decreased. The cell is also running hotter even before reaching 100% SOC, i.e. during the recombination phase, and its average internal pressure is higher compared to lower capacity cells. The result is shorter cycle life. Another way to get some space, as Tom has pointed out, is to decrease the thickness of the separator between the positive and negative electrodes. Thus, it wears out faster due to thermal sliding, and is more prone to rupture from crystals formed at the electrodes surface. It is the wearing out of the separator that causes increased internal resistance towards the end of the cell life.
Running the capacity race, the engineers battle to use virtually every cubic micron inside a cell. This is particularly true for smaller cells, mainly AA and AAA size. Just look at the mass of the newer cells compared to older, lower capacity ones. Companies are well aware of the harmful influence this has on cycle life. It is not a coincidence that the major NiMH cell manufactureres still keep their lower capacity cells in production. Look at Sanyo, GP and Panasonic, which together hold over 90% of the market. Sanyo still makes 1700 mAh AA cells albeit in China. GP keeps its 1800 mAh cells in production. Panasonic has a 1600 mAh cells out.

 

[Handlobraesing]

Hello Bcwang,

Keep in mind that the results are for 100% depth of discharge cycles. I believe that if you re-charge frequently, the actual number of cycles goes up exponentially.

I saw the "hill" several times. It seems that under this load, the cell recovers slightly when the chemistry heated up. I found it very interesting...

Tom

Perhaps. The charge manager on-board Prius keeps the battery pack below fully charged and above heavy depth of discharge. Perhaps this is the key to preserving longevity?

 

[WildChild]

Just take a look at batteries used in wireless phones. All NiMH are below 1600 mAh, probably because they are more robust.

 

[NiOOH]

Just take a look at batteries used in wireless phones. All NiMH are below 1600 mAh, probably because they are more robust.

..and cheaper too