Why does voltage drop at the end of the charge?

[SilverFox]

Re: Surplus Ammo Boxes as a safe storage solutions for batteries?

Why does voltage drop at the end of the charge?

We have been discussing this topic in another thread and it seems like it should be given a thread of its own.

The original discussion was in this thread. I have moved posts over to this thread in an effort to split this topic out.

The reason for this interest is that some people use a fan to make sure their cells stay cool during charging. Does the use of forced cooling interfere with the charge termination?

Tom

 

[NiOOH]

I think maybe you mean decrease of internal resistance on increasing temperature?

The internal resistance increases with temperature, but I stand corrected-this cannot cause the voltage drop (Ohm's law). It is actually the charge acceptance that causes it. When the charge acceptance is close to or equals zero, i.e. when the cell enters overcharge the cell behaves as it is not being charged (apart from heating of course) and the voltage drops.

 

[Muse]

Re: Initial charge on Eneloops (AAA)

The direct cause of the voltage drop is the increase of the internal resistance of the cell, which in turn, is caused by the increase of temperature.
That is why, I usually do not recommend adding active cooling to chargers (fan). Although cells may be cooler at the end of charge, overcharge may occur.

I will discontinue the use of the fan over my BC-900. It's a 12v fan but is run at a much lower voltage, so the cooling action is not intense. The fan is barely audible. However, based on what I'm seeing in this thread the potential benefit is evidently more than counterbalanced by the potential detriment. I did a fair amount of research (mostly in these forums) about the BC-900 (and other chargers) and rechargable batteries but now I wish I'd done more. Well, as it turns out I think I have done the right thing with my 8 Powerex 2400 mah AA's. They are doing nicely in my digicam, a rather fussy Samsung Digimax V3. It's my 8 Rayovac Hybrid AAA's that likely suffered from overcharging. At least 4 of them seem to do alright in my universal remote, and I plan to relegate all 8 to that task (keeping 4 in reserve for swapping out), saving my new Eneloops for my MP3 players. Hopefully, the single overcharge I subjected 4 of them to has not damaged them too severely.

 

[NiOOH]

Re: Initial charge on Eneloops, charging strategies. (AAA)

To help cooling of charger electronics I use heatsinks under the charger. I've found that this effectively lowers the end of charge temperature of my cells by couple of degrees Celsius.

 

[Hoggy]

Re: Initial charge on Eneloops, charging strategies. (AAA)

I think maybe you mean decrease of internal resistance on increasing temperature?

The internal resistance increases with temperature, but I stand corrected-this cannot cause the voltage drop (Ohm's law). It is actually the charge acceptance that causes it. When the charge acceptance is close to or equals zero, i.e. when the cell enters overcharge the cell behaves as it is not being charged (apart from heating of course) and the voltage drops.

I've more often read that internal resistance decreases with higher cell temperature(hence, why people say to heat up the cell a bit to get it accepted in the C9000), but every once in a while I come across some people saying the opposite...

So, which is it????? :thinking:

 

[Russel]

Re: Initial charge on Eneloops, charging strategies. (AAA)

I recently ordered some Accupower low self discharge D cells and discovered this chart for their AA cells:

http://www.accupower-usa.com/UpLoads/Spec AA.pdf

What caught my eye on this chart is the difference in delta v (on the charge chart) between .1C, .5C, and 1C charge rates. That goes a long way toward explaining why chargers that end charge on Delta V may have a problem sensing it at lower charge rates.

Russ

 

[PeAK]

Re: Initial charge on Eneloops, charging strategies. (AAA)

.
.
...What caught my eye on this chart is the difference in delta v (on the charge chart) between .1C, .5C, and 1C charge rates. That goes a long way toward explaining why chargers that end charge on Delta V may have a problem sensing it at lower charge rates.

Russ

Our esteemed colleagure "SilverFox" posted similar information about 16 months earlier in his post on Slow Charging. But it's great to have a manufacturer's data as well. Looking at the graphs, it begs the question as to why they do not end on "zero slope" (or near zero) and live with a lesser charge ?

​

Even my "poor" batteries can exhibit a "zero slope" region...more battery sales, perhaps ?

PeaK

 

[Mr Happy]

Re: Initial charge on Eneloops, charging strategies. (AAA)

On examining that graph, it seems that zero slope is not reached at a 0.1C charge rate. Therefore a charger would have to terminate on "small positive slope". But how small is small enough? Looking at that chart, there is a small positive slope at 80% capacity, and a similar small positive slope at 120% capacity.

I think the reason chargers look for a negative slope is that it provides a very positive yes/no end of charge test. Without a negative slope test there would be a gray area with more premature terminations and more missed terminations. Overall there would be more complaints about chargers doing the wrong thing.

It shows I think that tradition often has wisdom and experience behind it and we break with tradition at our peril.

 

[Mr Happy]

Re: Initial charge on Eneloops, charging strategies. (AAA)

I've more often read that internal resistance decreases with higher cell temperature(hence, why people say to heat up the cell a bit to get it accepted in the C9000), but every once in a while I come across some people saying the opposite...

So, which is it????? :thinking:

On looking at the eneloop data sheet, the Charge graph seems to show uniformly lower charging voltages at higher temperatures, given a constant charging current of 2000 mA. This says to me that higher temperatures lead to lower resistance, and lower temperatures to significantly higher resistance.

 

[PeAK]

Re: Initial charge on Eneloops, charging strategies. (AAA)

On examining that graph, it seems that zero slope is not reached at a 0.1C charge rate. Therefore a charger would have to terminate on "small positive slope"...

O.K. 0.1C might be an unreasonable case to apply the recommendation but more "real world" conditions like 0.5C and above...it might work. For instance a charger with fixed higher rates of charge (like 1C) in combination with "zero slope scheme" enabled above some voltage like 1.4 Volts.

PeAK

 

[Turbo DV8]

Re: Initial charge on Eneloops (AAA)

Should I stop providing ventilation to the cells? :sigh: :thinking: The BC-900 actually senses the temperature of the cells in determining -delta V termination? How does it do that? Separate temperature sensors under each cell?

You don't hear much about this around here. I've made a couple posts arguing against active cooling on the BC-900, based on my experiences with the BC-900. I mounted small cooling fans to hover directly over my three BC-900's, thinking cooler is better. I charge AA cells at either 500 mA or 700 mA. I quickly noted that with great regularity most of my cells were missing termination with the fans running. The voltage would climb, the current would accrue well past the time it should have terminated. But within five minutes of turning off the active cooling, all the cells would then terminate. I could repeat this over and over and over. Keeping the fans off, the cells will terminate at exactly the expected moment. I don't think it likely this is attibuted to the temperature sensor in the La Crosse being too cool, but more that prohibiting the cells temperature from rising thwarts, or "dampens", the small voltage change necessary for conventional termination. That's all a theory, of course. I haven't heard anyone else offer similar observations, but I myself am utterly convinced using active cooling on the LaCrosse will cause many failed terminations. Maybe if you charged at 1 Amp or higher (if that's your thing) it would terminate properly even with the fan going, I don't know. But, if anyone has problems with the LaCrosse terminating, the first thing I ask is, "Do you have a fan blowing on the charger?" When I use refresh mode, I turn the fans on, to carry away the dissipated heat, and might even leave the fan running for the first half of the charge cycle. But after that, it's "fans off" for me! Try it, you might be surprised.

 

[Turbo DV8]

Re: Initial charge on Eneloops, charging strategies. (AAA)

On looking at the eneloop data sheet, the Charge graph seems to show uniformly lower charging voltages at higher temperatures, given a constant charging current of 2000 mA. This says to me that higher temperatures lead to lower resistance, and lower temperatures to significantly higher resistance.

I thought it was generally understood within the universe that higher temperatures yield higher resistance within a conductor, and lower temperatures yield lower resistance. Hence, superconductors are not super-heated, but rather, super-cooled, to achieve their superconducting property. This is also why a light bulb does not draw a gazzilion watts of power after it is turned on. If you measure the resistance of a tungsten bulb filament, then try to apply ohms law to get the power, you will be shocked at how much power your calculator says the "100 watt" bulb draws! But the instant the filament heats up, it's resistance increases dramatically, lowering the current draw, and the power consumption to 100 watts. These two examples should fairly well demonstrate that resistance within a conductor increases with temperature. But I could be wrong...

 

[Mr Happy]

Re: Initial charge on Eneloops (AAA)

I haven't heard anyone else offer similar observations, but I myself am utterly convinced using active cooling on the LaCrosse will cause many failed terminations.

I certainly recall discussion of this in other threads, maybe not only in regard to the BC-900. I generally concur with the observation.

When you are charging NiMH batteries on a charger that uses −∆V end-of-charge detection, the batteries must get warm at the end of the charging period. This is a required part of the charge termination process. If you provide active cooling that is not designed into the charger you may interfere with its correct operation and you may do more harm than good.

If the batteries get more than warm during charging, for instance hot cup of coffee hot or even hotter, then there is a problem, either with the charger or with the batteries. If the charger is not a known recommended good one you may want to replace it. If the batteries get hot on a known good charger you should use a refresh or conditioning cycle on the batteries, and if that does not work you should replace the batteries.

 

[Mr Happy]

Re: Initial charge on Eneloops, charging strategies. (AAA)

I thought it was generally understood within the universe that higher temperatures yield higher resistance within a conductor, and lower temperatures yield lower resistance. Hence, superconductors are not super-heated, but rather, super-cooled, to achieve their superconducting property. This is also why a light bulb does not draw a gazzilion watts of power after it is turned on. If you measure the resistance of a tungsten bulb filament, then try to apply ohms law to get the power, you will be shocked at how much power your calculator says the "100 watt" bulb draws! But the instant the filament heats up, it's resistance increases dramatically, lowering the current draw, and the power consumption to 100 watts. These two examples should fairly well demonstrate that resistance within a conductor increases with temperature. But I could be wrong...

Your observation is correct in certain cases, but it is not generally true. If you replace the word "conductor" with "metal" then resistance does increase with increasing temperature.

But the same is not always true of non-metallic conductors. For instance graphite decreases its resistance with increasing temperature, as do LED junctions. Another common material that decreases resistance with increasing temperature is the material in NTC (negative temperature coefficient) soft start devices sometimes used in hotwires.

In a battery where conduction is by moving charge carriers in the electrolyte, then increasing temperature can increase the mobility of those charge carriers and thus decrease the resistance to flow.

 

[Turak]

Intersting thread....a few observations.

I have 2 of the BC-900's and 2 of the MH-C9000's.

A few things to note about each of them regarding their charging characteristics....

Both units are supposed to terminate using the following strategies;

MH-C9000 - Independent Negative Delta V, temperature, and timer.
BC-900 - Independent Negative Delta V, Temperature, and timer.

Because of various revisions over time, the newer models actually terminate using the following strategies;

MH-C9000 - Preset MaxV, Independent Negative Delta V, temperature, and timer.
BC-900 - Independent Negative Delta V, temperature, and timer.

Using the MH-C9000, with MOST batteries you will almost never hit the -detaV because the Preset Max V point is set to 1.47V plus or minus .01V. Take the Eneloops for example, a cell in good condition typically hits -deltaV at about 1.54 to 1.58V. A problem though is that the -deltaV point actually changes (unfortunately it gets higher with less of a dip) as the cell ages and the internal resistance builds over time.

Regarding the cells heating at the end of the charge. The heat build up observed when a cell reaches 'full saturation' is simply a by product of the fact that the cells reactive material can not absorb any more of the charge current and starts converting the excess to heat, since the cell is fully charged and there is now more excess current...you get more heat. It really does not have anything to do with the actual production of the negative V dip that you see in the voltage.

People talk quite a bit about missing terminations, which is one of the worst things that can happen to a cell.

A few things that I have noticed regarding missing terminations;

New cells - DEFINITELY keep your eye on them when running them through their first 2 or 3 cycles. ALL cells can miss terminations at this point no matter what charge rates you use. Until a cell has been ran through a couple of charge/discharge cycles the -deltaV slope is very small at first. The LSD cells are almost immune to this problem because they come precharged, although I have seen it happen with them. Using an MH-C9000, this should not be much of a problem because of the cells typically terminating on PresetMaxV instead of -deltaV. With a BC-900 this is definitely a problem because of the smaller dip on the -deltaV slope the first couple of cycles.

Aging cells; The characteristics of each charger are somewhat different in this respect. As the cells age their internal resistance tends to increase. This definitely causes the cells to heat a bit more at the higher charge rates. A bit more in the BC-900 than the MH-C9000 because the BC-900 does truly terminate on -deltaV.

BC-900 - This charger tends to terminate properly on all charge rates with good cells. Most of the missed terminations I have encountered or helped people investigate seem to revolve mostly around brand new regular NiMH cells and any older cells using the lower rates. Unfortuneately, with this charger as the internal resistance increases, the cells tend to get hotter at the end of the charge cycle (because it is truly terminating on -deltaV slope).

MH-C9000 - For this charger, you BETTER stay within the recommended .5C to 1C charge rate. I have went as low as .3C fairly reliably, but any lower and you WILL start getting many missed terminations, especially with aged cells. THe charger is definitely gentler on the older cells because of the fixed max charge point. The cells typically never hit -deltaV, so they don't heat up near as much. BUt a side affect is that they tend to only be about 85-95% charged....even after their topoff and trickle charges finish. Unfortunately, how 'full' the battery is or how completely it was charged actually gets worse as the cells internal resistance increases with this charger.

For both chargers, I have noticed that if you want RELIABLE terminations, you will need to increase the charge rates as the cells age due to the internal resistance build-up. The -deltaV slope definitely becomes less pronounced as the internal resistance of the cell climbs. A side affect of the increased charge rates with the BC-900 would be the batteries getting hotter at the end of the charge cycle. With the MH-C9000, warmer cells at first but at some point it will simply start rejecting the cells, because of the overly aggressive high impedence test.

 

[shadowjk]

Re: Initial charge on Eneloops, charging strategies.

So has anyone tried blocking the top vents on the BC900, sawing a hole in the sides, and having a fan (with ducts to direct the air) blow over the circuit board only, not over the cells?

My own measurements with IR thermometer seems to suggest that most heat is generated inside the charger. The hottest spot seems to be on the right side of the charger, roughly "under" the fourth slot button.

 

[PeAK]

Re: Is TEMP the STORK in the picture ?

.
.
Aging cells; The characteristics of each charger are somewhat different in this respect. As the cells age their internal resistance tends to increase. This definitely causes the cells to heat a bit more at the higher charge rates...
.
.

.
.

...The -deltaV slope definitely becomes less pronounced as the internal resistance of the cell climbs. A side affect of the increased charge rates with the BC-900 would be the batteries getting hotter at the end of the charge cycle...

I've seen statements to the effect that cause of the "drop in battery voltage" is directly due to the increase temperature. Is temperature the "stork" that happens to fly by when a baby is born ?

More bluntly, can anyone verify this statement or is it (the temp increase) a side effect that just comes along for the ride. I have some batteries that show a rising voltage over a twenty second interval just prior termination and on "count 20" the charger just suddenly terminates. No heating...just very cool batteries.

Has someone done a charge profile where they keep the external temperature constant and compare it to one where it is allowed to heat. More importantly is there a realiable paper reference that points to temperature as being the direct cause of the negative delta-V drop ?

To play the devil's advocate, I'll say that the drop is due to state of the battery (charge state) when a change in the reaction process (???) somehow drops the battery voltage as charge is pumped into a battery.

PeAK

 

[Mr Happy]

Re: Is TEMP the STORK in the picture ?

I've seen statements to the effect that cause of the "drop in battery voltage" is directly due to the increase temperature. Is temperature the "stork" that happens to fly by when a baby is born ?

More bluntly, can anyone verify this statement or is it (the temp increase) a side effect that just comes along for the ride. I have some batteries that show a rising voltage over a twenty second interval just prior termination and on "count 20" the charger just suddenly terminates. No heating...just very cool batteries.

Has someone done a charge profile where they keep the external temperature constant and compare it to one where it is allowed to heat. More importantly is there a realiable paper reference that points to temperature as being the direct cause of the negative delta-V drop ?

To play the devil's advocate, I'll say that the drop is due to state of the battery (charge state) when a change in the reaction process (???) somehow drops the battery voltage as charge is pumped into a battery.

You ask good questions. I would like to see the constant temperature charging experiment, but I don't think I myself have a good way to set it up.

I did a quick search last night for a reference pointing to the underlying reason for the voltage drop but I could not find one.

What we have at present is circumstantial evidence. Firstly, if you charge at lower rates the voltage drop is less pronounced and at low enough rates may not happen at all. This is consistent with the ability of the batteries to dissipate small amounts of heat better than large amounts of heat and thus show a much less pronounced temperature rise at the end of charge.

Secondly, with older high resistance batteries the voltage drop is also less pronounced. This is consistent with greater temperature rise throughout the charging process and a much less pronounced "temperature spike" at the end of charging.

Thirdly, we have the observation that if people apply active cooling during charging it appears to delay or prevent termination. If the voltage dip were unrelated to heat, why should this happen?

So what I believe happens is this. If you charged a cell while externally holding the temperature constant its voltage would rise until all the chemical reactions in the cell had been activated to their maximum potential and then the voltage would plateau. Further application of constant charge current would cause recycle reactions to occur and generate heat, but no further rise in voltage. At this point the cell voltage would consist of two parts added together: (i) the chemical EMF of the cell, and (ii) the voltage difference due to internal resistance acting on the applied current.

If we suddenly removed the active cooling in this plateau state the cell temperature would rise sharply, since all applied current is being dissipated as heat. This would cause the internal resistance component of the voltage (ii) to decrease and the overall cell voltage would drop a little.

Putting all this together, what we hope to happen in charging is a sharp and pronounced drop in internal resistance at full charge so that the fall in voltage due to this overtakes the rise in voltage due to increasing chemical potential and a net decrease in voltage is seen. Two things work against this: (i) a low charging current produces a slow and limited rise in temperature, giving it less chance to overtake the still rising chemical potential; (ii) old high resistance batteries exhibit an early and continuous rise in temperature so that there is a less sudden change in temperature at the end of charging, producing the same outcome as (i).

 

[Turbo DV8]

Re: Is TEMP the STORK in the picture ?

To play the devil's advocate, I'll say that the drop is due to state of the battery (charge state) when a change in the reaction process (???) somehow drops the battery voltage as charge is pumped into a battery.

That's roughly the way my guessing goes, but then, I'm not a chemist.

Regarding the cells heating at the end of the charge... It really does not have anything to do with the actual production of the negative V dip that you see in the voltage.

Even assuming this to be 100% accurate, there is definitely some sort of distinct correlation between cell temperature and termination. I have observed way too many times the cells failing to terminate, just hanging and cool as a cucumber, and within 5-8 minutes of turning off the fan and allowing the cells to warm a bit, all four cells terminate within several seconds of each other. It's just to repeatable to be a coincidence.

 

[NiOOH]

Re: Initial charge on Eneloops, charging strategies.

To me, the voltage drop is due to charge acceptance.
Switching on the current the charging reactions begin to take place in the cell. For a NiMH cell, these reactions are exothermic, i.e. produce heat, so even without the heat generated by charger electronics we may see some increase of cell temperature. However, the charge acceptance in early stages is close to 100% This has been verified by measuring charge and discharge capacities on the c9000. If you pull the cell off when "done" appears and discharge it you will get almost the same capacity as the current put to the cell. Nearing 100 % state of charge (SOC) the charge acceptance drops sharply. Parallel to this more and more currennt is dissipated as heat. At exactly 100 % SOC charging efficiency reaches zero. The voltage is at plateau. From now on, the cell cannot accept any charge, in other words it is behaving as it is not being charged and a voltsge drop occurs. Still, we have the temperature factor that keeps the overvoltage reactions going.
If one would have a way to measure charge acceptance it would have been possible to sense exactly 100 % SOC. Unfortunately, charge acceptance cannot be measured directly. Second best is temeperature, since its rate of increase is directly proportional to the rate of decrease of charge acceptance. Voltage is more complex and not always 100 % reliable.