got some young 2012 modules

Do you have good data from the previous 2 discharge cycles?

 

GOOD, FAST, CHEAP . . . pick two. So I would answer the question with a question:

• Does the charger scanning provide value for time spent versus doing something else?

If the initial question about trying to get an Ebay challenge in has been resolved, doing the charger scans are interesting but perhaps not critical for meeting a 30 day complaint window. It would provide a better metric for matching the weak and strong modules but your initial, parallel charge with monitored, resistor discharge also gave a good metric . . . just not deep enough. A target of 6.5V, no lower than 6.0V would be fully adequate.

You know your schedule and the value of your time much better than any of us. It is like the old joke of the pig and chicken who decided to have a breakfast of ham and eggs to discuss their issues with the farmer. The chicken was interested but the pig was committed.

Bob Wilson

 

Well I would call them a BF hammer:

If you were to offer those modules for sale, you could ask for and probably get more than you paid. But every time they are overcharged:

• 7500/5990 ~= 125%
• 7500/6096 ~= 121%

Some capacity is 'eaten' away.

The analogy is blowing up a balloon:

A maximum inflated balloon is impressive but:
BBC News - What happens when a balloon explodes?

From this data, I would stop charging at 6.5 Ahr and finish screening all the modules.

Bob Wilson

 

That has not been my experience. Others have made that claim but this is what I observed:

Notice the module at the bottom that went through multiple charge-discharge cycles and never gained any capacity. In contrast, the other two modules were re-hydrated by an EXTREMELY EXPERIMENTAL technique of drilling a hole, adding water, and going through a large number of charge-discharge cycles.

I don't 'set my hair on fire' when someone posts what under very well controlled circumstances might work. Over charging at a low rate is nominally 'OK' but we're looking at a complex problem that I'll describe in the next post.

The dV/dt peak is something each cell in the module reaches. Even though they share a common, internal electrolyte state (more about this in the next technical post), local variations and the very small values can be missed by the charger. If six people in an audience each clears their throat independently, no one would notice. But if they all clear their throat at the same time, it would be noticed.

Overcharging is one of those cases where less-is-more. But they are not my modules and you have an excellent set. My testing was with crappy modules from failed packs (i.e., packs with one or more failed modules.) But based upon the observed 'charge-discharge' efficiency of 90-93%, I would limit the over charge to:

• 6.0 Ahr / 90% ~= 6.7 Ahr (6,700 mAhr)
• 6.7 Ahr * 90% ~= 6.0 Ahr

Bob Wilson

 

TECHNICAL CONTENT

A NiMH (Nickel Metal-Hydride) battery has two electrode in an electrolyte separated by a thin, plastic fiber:

The failure mechanism is loss of water from the electrolyte that reduces the area current flows. As the conductive area gets smaller, the resistance increases and the temperature gets hotter and hotter at the conductive point. When it gets hot enough to melt the plastic, it shorts the two electrodes and that particular cell is destroyed. It can not be recovered.

Now there are two chemical processes involved in charge and discharge. Nickel is a multivalent metal. What this means is it has two hydroxide states:

• Ni(OH){2} - 1 nickel, 2 oxygen, 2 hydrogen
• NiO(OH) - 1 nickel, 2 oxygen, 1 hydrogen

So what happened to the hydrogen?

The other electrode is mix of rare earth metals that has the ability to absorb or release hydrogen:

• M - metal hydride, a mix of rare earth metals
• MH - metal hydride, 1 hydrogen

So in effect, our NiMH battery simply transfer a hydrogen atom from one molecule, the nickel oxide electrode, to a metal-hydride matrix that hold the hydrogen atom. It does this without changing the density, the size, of the two nickel oxides or the metal-hydride matrix. This keeps the cell from falling apart like some LiON chemistries do.

Now to discuss the electrolyte, a saturated solution of potassium hydroxide, KOH:

• H{2}O + KOH

This is called an ionic fluid because the potassium has such a high affinity for hydrogen that it in effect can tear H{2}O into ions H+ and OH-. But the important aspect is if you take a hydrogen from the electrolyte and stuff it in the MH, the electrolyte will nearly instantly suck a hydrogen on the other side of the separator from the nickel oxide electrode. This happens when the cell is being charged.

When the cell is being discharged, the metal-hydride forces an hydrogen into the electrolyte. Then nearly instantly on the other side of the separator, a hydrogen is forced into the nickel oxide. The whole cell works by passing hydrogen between the different electrodes using the soup of ions that is the electrolyte.

Now this is not a formal chemical analysis of what goes on in the NiMH cell chemistry. A formal chemical analysis uses hydroxide ions, OH-. But thanks to symmetry of H{2}O, water, it doesn't matter if the analysis uses OH- or H+. The KOH is a powerful ionic compound that allows this to work. If someone wants to do a formal, chemical analysis, I'll smile and say 'thank you' but my minds eye sees the hydrogen protons doing the 'heavy lifting.'

Bob Wilson

 

MORE TECHNICAL CONTENT

NiMH cells exhibit a dV/dt effect as it passes peak charge:

Now I have never seen a good explanation for the dV/dt peak but we know on the overcharged side:

• 2H{2} and O{2} is produced - hydrogen and oxygen
• the nickel oxide+KOH solution+M converts the 2H{2} + O{2} into water and heat

This happens individually with each cell. But normal manufacturing variability makes slight differences in each cell capacity. So the dV/dt actually happens in each cell, independently and to the extent the cells are matched, the dV/dt may or may not over lap. But there is something else I just noticed.

The magnitude of the dV/dt appears to be a function of the charge rate. This suggests a higher current relative to the Ahr capacity of the cells may be critical to detecting a measurable dV/dt. But the chargers look for the decrease after peak charge. In effect, they 'tickle the tiger.'

A higher charge rate, closer to the Ahr capacity of the module is likely to result in dV/dt detection by the charger. But I would still want a lower Ahr maximum setting just incase the charger misses the dV/dt blip.

There are cheap, battery chargers that use the temperature increase to detect 'full charge.' But the temperature increase is slower than the dV/dt. Temperature based, NiMH battery chargers should be fixed with a BF hammer.

Bob Wilson

 

Exactly. You had proposed 6800 mAhr and that seems reasonable since my historical data suggests 6000 mAhr modules with a 90% charge-discharge efficiency should be good with a 6700 mAhr charge.

Bob Wilson