1.2 volt NiMH vs 1.5 volt alkaline cells

[Curious_character]

I keep reading the statement that NiMH cells won't work in some or all devices designed for alkaline cells because they're "1.2 volts" instead of the alkaline's "1.5 volts". Instead of explaining over and over why this isn't true, I've put together an essay on the topic. Anyone interested can find it here. I hope you find it useful. Comments are welcome.

c_c

 

[MorePower]

There are certain devices which will not function properly on Nimh cells, but they are few and far between.

Generally, the problems arise when the device uses multiple cells in series and has a fairly high cut-off voltage. Electronic hotel door locks, for instance, often contain a 4 cell battery consisting of AA cells. The circuitry requires a minimum of 6 volts into a 2 ohm load at 20ms. Unless the Nimh cells were hot of the charger, there's no way they'd work for this application. Some older digital cameras also were known to have higher than 0.9V-per-cell cutoffs (I remember a Kodak that died at 1.27V-per-cell) and these could also have issues with Nimh cells.

Another problem can occur when the device relies on the internal resistance of alkaline cells to help reduce the flow of current. Due to the lower internal resistance of Nimh cells, it's pretty easy to let out the "magic smoke" in these poorly designed products.

 

[FlashCrazy]

Excellent essay, Curious Character! That's why I recommend NiMH batteries be used in the lights I mod and sell...alkalines just can't maintain the same voltage under high loads. I've tested AAA alkaline batteries at a 1 amp discharge...they fall to 0.9 volts in just 4 minutes. The same test on a typical NiMH AAA yields 50 minutes!

 

[ElectronGuru]

+1 - Awesome!

I've been reading and reading but none of it compares to your line graph showing the Duracells dropping while the Eneloops carry on. Time to call up BatteryBob

 

[Christexan]

I agree, excellent essay, and also excellent response in the following post pointing out where it is that alkalines can sometimes be required, it's not a "single-cell" issue, but a multiple-cell issue where the basic voltage difference can become too great for the design. A 6-AA alkaline (9.0V nominal) solution might require a minimum of 7.0V to operate properly, but will only have a nominal 7.2V under NiMH, and if it is cutoff based on voltage, might cutoff at only a 0.1V drop below nominal on the NiMH.

I have an older Kodak digital 4-AA camera that almost immediately shows "low-battery" when using NiMH, but continues to work for a long time after that warning appears (it seems that it appears around 1.2V nominal, so fresh off the charger, around 20 shots it pops up, but can get about 220 shots out of the camera on NiMH, maybe 100-120 on alkalines).

Hopefully anything "modern" takes rechargeables into account, and I'm a huge proponent of rechargeables considering the terrible load characteristics of alkalines, for CPF usage especially they make no sense at all. For "long-life light duty usage" E2 Lithiums seem to me a better choice, otherwise rechargeable all the way.

 

[Curious_character]

I agree, excellent essay, and also excellent response in the following post pointing out where it is that alkalines can sometimes be required, it's not a "single-cell" issue, but a multiple-cell issue where the basic voltage difference can become too great for the design. A 6-AA alkaline (9.0V nominal) solution might require a minimum of 7.0V to operate properly, but will only have a nominal 7.2V under NiMH, and if it is cutoff based on voltage, might cutoff at only a 0.1V drop below nominal on the NiMH.

Thanks very much for the comment. I see that I have to add a bit to my essay. Any device designed for multiple alkaline cells that shuts down at a voltage of about 0.9 volt PER CELL will result in your throwing away cells that have considerable remaining energy. Let's take your example of a 6 cell device which shuts down at 7.0 volts. This is 1.17 volts per cell. Take another look at the third graph in the essay, which shows the cell voltage and percentage of energy extracted with a 500 mA load. If the device is drawing 500 mA, it'll shut down when the battery has 60% of its energy remaining -- it'll get only 40% of the available energy from the battery. If the drain is 1 amp, the next graph shows that it'll get only 25% of the available energy out. This is just as poor a design as a single cell device that shuts down at 1.17 volts. I'm not claiming that there aren't such poorly designed products out there. But thankfully they're not the norm.

Ironically, a NiMH cell would do very much better in that device at 500 mA, as the fifth graph shows. Its voltage doesn't drop to 1.17 volts until it's nearly completely discharged. The device would run more than twice as long with a NiMH cell than alkaline. Even at one amp (next graph), you'll get the lion's share of energy out of the NiMH before hitting 1.17 volts, while the alkaline would quit after delivering only 25% of its energy -- and its total energy is a lot less than the NiMH to begin with at this current.

I have an older Kodak digital 4-AA camera that almost immediately shows "low-battery" when using NiMH, but continues to work for a long time after that warning appears (it seems that it appears around 1.2V nominal, so fresh off the charger, around 20 shots it pops up, but can get about 220 shots out of the camera on NiMH, maybe 100-120 on alkalines).

Battery gauges designed for use with alkalines are badly inaccurate when used with NiMH. And as you can see from the essay graphs, the voltage of a NiMH cell drops relatively little during the discharge. It can vary more from cell to cell or with temperature than it does over a discharge period. So it's nearly impossible to make an accurate battery gauge for NiMH cells based on voltage. I have a Garmin GPS receiver which does a moderately good job, but its algorithm is more complex than just looking at the cell voltage -- it appears to also look at the rate of drop, and perhaps temperature and other factors. About the only way to do it accurately is to measure the charge or energy extracted.

Hopefully anything "modern" takes rechargeables into account, and I'm a huge proponent of rechargeables considering the terrible load characteristics of alkalines, for CPF usage especially they make no sense at all. For "long-life light duty usage" E2 Lithiums seem to me a better choice, otherwise rechargeable all the way.

Hopefully anything "old" or "modern" is designed to extract most of the energy from alkalines, in which case it'll work fine with rechargeables.

I agree that alkalines are a poor choice for even moderate drain applications, and very poor for high drain.

c_c

 

[TorchBoy]

Electronic hotel door locks, for instance, often contain a 4 cell battery consisting of AA cells. The circuitry requires a minimum of 6 volts into a 2 ohm load at 20ms.

That's very interesting. How many AAs can provide a minimum of 1.5V @ 0.75A for any amount of time? If that's what's required I presume they tend to use lithium or NiZn or NiOOH to ensure the voltage. And even then they'd have lots of partly used cells. And lots of people getting locked out of their rooms because it would be quite unreliable and battery changes would be required so often.

Ironically, a NiMH cell would do very much better in that device at 500 mA, as the fifth graph shows. Its voltage doesn't drop to 1.17 volts until it's nearly completely discharged.

Not just any NiMH cell though. Eneloops are... oh never mind, you've heard it all before.

In SilverFox's test of Energizer 2300 cells at 500mA they give about 0.5Ah before dropping under 1.17V, which is only a quarter of their capacity at that current. To get an Eneloop's voltage to sag that much you have to draw at least 3A from it.

Not all NiMHs are the same.

 

[Curious_character]

. . .Not just any NiMH cell though. Eneloops are... oh never mind, you've heard it all before.

In SilverFox's test of Energizer 2300 cells at 500mA they give about 0.5Ah before dropping under 1.17V, which is only a quarter of their capacity at that current. To get an Eneloop's voltage to sag that much you have to draw at least 3A from it.

Not all NiMHs are the same.

Absolutely true. In fact, I just posted a direct comparison between an Eneloop and an Energizer 2300 here. But I think the Energizer is more of an exception than the Eneloop -- that is, I think the Energizer is exceptionally bad rather than the Eneloop being exceptionally good. I'll have to study SilverFox's excellent and complete analysis to see if that is indeed the case. If not, I'll have to qualify what I've said to note that it applies only to superior NiMH cells.

c_c

 

[Curious_character]

I've added a brief section to the essay dealing with multiple cell devices, and clarified the part about the amount of energy delivered at different currents. Here's the link again.

c_c

 

[SilverFox]

Hello Curious_character,

Your comments are correct if your primary goal is to remove all the energy you can from a cell. However, you are ignoring upset conditions and safety factors. Proper design looks at these issues also.

If the device you are designing the circuit for is expensive, you have to consider what happens when the device is returned under warranty because the batteries leaked, or disassembled, or ruined the cells within the battery. In a perfect world, every one would use a matched and balanced set of cells and these would be "non-issues." However, we don't live in a perfect world and the design has to allow for these "issues."

Let's look at the 6 cell battery that the circuit shuts off at 7 volts. The extreme case would be if the 6 cells were at 1.4, 1.4, 1.4, 1.4, 1.4, and 0.0 volts. At this limit of imbalance we are set up to reverse charge the weakest cell, which will ruin the cell, and could possibly ruin the device powered by the battery.

As a design engineer, your question now becomes "How much imbalance should I allow for in my design?"

Tom

 

[MorePower]

That's very interesting. How many AAs can provide a minimum of 1.5V @ 0.75A for any amount of time? If that's what's required I presume they tend to use lithium or NiZn or NiOOH to ensure the voltage. And even then they'd have lots of partly used cells. And lots of people getting locked out of their rooms because it would be quite unreliable and battery changes would be required so often.

Nope, they use alkaline cells almost exclusively. The actual discharge time is so short (a pulse length < 300ms) that the required voltage can be maintained for a large number of unlock cycles. At end of life the cells do have a large percentage of their capacity remaining, but they're inexpensive so that's not an issue.

 

[Christexan]

As in the other thread, I'm very well versed in the current versus capacity issues of alkalines and other chemistries, it doesn't change the fact that a multiple-cell device (or maybe even a single-cell) at a LOW current draw, may not like NiMH instead of alkalines, due to the low nominal voltage that it "sees". Again, the essay you made is EXCELLENT and I'm not disagreeing at all with your observations, in fact they are exactly the ones I would make, but they are going the wrong way to where an issue can occur, which is that under LOW current conditions (100mA or less for instance), alkalines stay above 1.25V for a long time compared to NiMH which nominally operate around 1.2-1.25 for a long time. If a device is looking for any multiple above 1.2V at low currents, it may think (incorrectly) that a 1.25V NiMH is "nearly dead"... if it thinks at all, if it simply requires a multiple of voltage at 1.25V or higher to function properly, it will just not work on NiMH or may be erratic.

I'm in 100% agreement with all your observations and have made them myself many times in the past, that doesn't mean however that there aren't rare exceptions out there.

Especially with ICs, quite a few devices are designed to operate specifically on a range of 2.5-3.5V (3.3V nominal)... such a device might work fine on 2 1.5V alkalines until they drop to 1.25V (as in the other thread, potentially wasteful, but functional)... however these same devices would be pretty erratic after the early discharge portion of the NiMH curve, they would be operating at the margin of the 1.25V (typical) nominal portion. If the batteries aren't at 1.25V in their nominal portion the device won't function properly or at all for very long.
Still, it would be a rare device that has this issue, good design would plan for lower voltages (to increase alkaline run-times if nothing else), however there are many more bad engineers than their are good ones out there (true in any profession IMO, most people just go through the motions to get a paycheck), so bad designs ARE going to happen including non-regulated DC power supply digital devices needing 2.5V-3.5V input to operated.
So you are right, but you haven't covered all the possibilities, with your essay, the biggest assumption remains (and you say it in there not saying you haven't acknowledged it) "well designed devices"... I pretty much assume railroad signals won't work and look both ways as I approach a train track, because one day someone is going to make a mistake on a well-designed signal system and it won't know a train is coming... it has happened before and will again, so I don't assume the last guy who worked on it got it right.
A final thought, is that some devices may incorporate a conservative low-voltage cutoff deliberately, to avoid over-discharging alkalines and the resultant leakage that often occurs as a result. Anything over 1.0V is clearly TOO conservative for that concern, but who knows... it could even be a marketing ploy... "3 hours on alkaline, or 15 hours on our special lithium "long-life" batteries".... even though they could run the alkalines for 10 hours down to 0.9V, it'd be a way to "upsell" an expensive lithium pack upgrade. Or even a rechargeable pack (instead of 4AA alkalines, uses 5 smaller NiMH cells in a sealed pack to up the nominal voltage, even though the "capacity" is lower in that case, the runtime is longer since the NiMH don't slope through the "cutoff" voltage until they are nearly dead unlike the alkalines. Conspiracy theory, I know, but you never know...

 

[Christexan]

Wow, 2 other posts since I started typing my reply, LOL, hot topic! And I see Tom beat me to my final thought as well, LOL... oh well a fun and lively discussion for sure.
Bottom line for me is that if it's anything expensive, or critical for max duration, or just want something "light and portable with long life" lithium primaries are the answer to all the issues presented (well, except the cell-imbalance issue as Tom pointed out, but I don't think someone is going to toss a $2.00 battery in a drawer and mix it in with others of unknown charges, at the current price, we are likely to buy them purposefully and put them into the device/set they are going to live and die in. Otherwise rechargeable if at all possible. I will be happy if/when lithium primaries and rechargeables replace alkaline primaries as the main shelf holder in stores. Much lighter, longer shelf life, much more "usable" capacity, much flatter discharge curve, better temperature range performance, hold their "nominal" voltage in REAL usage, less toxic, better color scheme (I like the E2 blue and silver)... only thing not to like is the price tag currently.

 

[Codeman]

Heat also has to be taken into account. If the device doesn't manage the higher sustained current that NiMH cells can provide, additional heat could be present which the device probably wasn't designed for.

 

[lamperich]

here. I hope you find it useful. Comments are welcome.

c_c

yes very useful


You should also add a "heavy duty" cell in the comparison.
especially @1A load

P.S.: http://www.ife.ee.ethz.ch/~zinniker/batak/ideal/index.html

 

[Curious_character]

yes very useful


You should also add a "heavy duty" cell in the comparison.
especially @1A load

P.S.: http://www.ife.ee.ethz.ch/~zinniker/batak/ideal/index.html

That would be good for a laugh (or at least a

), but little else.

I'm currently discharging a NiMH cell at 100 mA, to be followed by an alkaline, to see how valid my conclusions are at lower currents. I'll post the results and/or add them to my essay when they're done. 100 mA is about typical of a modern GPS receiver, but much less than most flashlights.

c_c

 

[Curious_character]

Hello Curious_character,

Your comments are correct if your primary goal is to remove all the energy you can from a cell. However, you are ignoring upset conditions and safety factors. Proper design looks at these issues also.

If the device you are designing the circuit for is expensive, you have to consider what happens when the device is returned under warranty because the batteries leaked, or disassembled, or ruined the cells within the battery. In a perfect world, every one would use a matched and balanced set of cells and these would be "non-issues." However, we don't live in a perfect world and the design has to allow for these "issues."

Let's look at the 6 cell battery that the circuit shuts off at 7 volts. The extreme case would be if the 6 cells were at 1.4, 1.4, 1.4, 1.4, 1.4, and 0.0 volts. At this limit of imbalance we are set up to reverse charge the weakest cell, which will ruin the cell, and could possibly ruin the device powered by the battery.

As a design engineer, your question now becomes "How much imbalance should I allow for in my design?"

Tom

That's an excellent point, and one I hadn't considered.

If I were designing a device to run from NiMH cells, I'd certainly consider the imbalance you mention. In fact, in a battery pack I've used for portable radio operation for years, I've put taps every 4 cells so I can safely discharge to 1 volt per cell for testing without a chance of cell reversal. Without the taps, you're unavoidably stuck with a higher cutoff to avoid reverse charging, as you say. This will, of course, reduce the amount of charge you extract from the cells.

But what precautions are needed in a device designed for multiple alkaline cells? I can't see worrying about "ruining" a dead cell, except for the potential for cell leakage -- something which others have also mentioned. I wouldn't be surprised if reverse charging would precipitate leakage -- if so, it obviously should be avoided if possible. But is leakage also more likely when the cell is discharged to a low voltage than a higher one? The old carbon-zinc cells would leak when discharged because the outer jacket was the zinc of the negative electrode. It was consumed during discharge, frequently resulting in leakage even though the manufacturers apparently made the jackets thicker than necessary for supplying material for the discharge. Jackets are now steel, so what's the mechanism causing leakage of modern alkaline cells? Is it more likely to occur when the cell is more fully discharged? If so, then it is indeed quite possible that device manufacturers would intentionally try to prevent the cells from getting fully discharged.

Does anyone have any firm data showing the likelihood of alkaline cell leakage as a function of discharge depth?

c_c

 

[TorchBoy]

OK, I have to say it - the missing quote marks in the title of the PDF are beginning to bug me. And the voltages are the other way around from the filename.

 

[csshih]

:bump: for an excellent read.

 

[Fallingwater]

Some older digital cameras also were known to have higher than 0.9V-per-cell cutoffs (I remember a Kodak that died at 1.27V-per-cell) and these could also have issues with Nimh cells.

True that. It's what prompted me to do this. Ironically, that digicam came with two NiMH cells and a charger to suit...

As for the problem of running stuff from NiMH: it used to be that some devices (like portable CRT TVs) would blow if running off NiMH because they were relying on the high internal resistance of alkalines to limit current, but I think it's fairly safe to say that it's a problem of the past. Nowadays there's no electronic device that doesn't have a DC-DC regulator of some sort inbetween battery and electronics. I haven't actually seen a recommendation to run stuff exclusively on alkalines for a long time...