Interesting - Tesco Alkalines vs. Energizer Alkalines

[tom66]

In my internship my task this week is to build a battery tester. This draws a fixed current from a single cell (AA or AAA) and accumulate mAh until the cell is dead (0.8V or less under load.)

I finished it today, so tested two batteries, new in packets. One was a Tesco Alkaline (30p each). The other was Energizer Ultra Alkaline (81p each.)

Final mAh from the batteries? Tesco=1181 mAh; Energizer=1249mAh. At 1 amp constant current discharge.

ESR of the Tesco cells also tested at a low 60mOhm new from packet; Energizer was 75mOhm. End ESR was about the same around 110mOhm.

The Energizer only just beat the Tesco one - but it costs almost 3x as much! This has changed my opinion of cheaper batteries. Full results will be published soon.

I also tested a zinc chloride, also Tesco, and as expected it only had around 110mAh. It was flat within 6.5 minutes at 1A discharge!

I am also going to test Poundland special ones (Alkalines) - 16.7p each - I'm not expecting much! I'm also planning to see how much of a difference it makes to test at 100mA instead of 1000mA, and also different load characteristics, such as simulated constant power (similar to a switching converter) and constant resistance (similar to a light bulb) characteristics.

[IanB]

That's pretty interesting, and I'm surprised. 1 amp is a pretty high discharge rate for an AA size alkaline so that was a severe test.

Some background: the ultimate capacity of a battery is broadly related to the weight of chemicals inside it. So if the Tesco battery weighed about the same as the Energizer you would expect them to have similar capacities. Weighing your batteries as part of the test would be informative. (Note that this only applies to batteries with the same chemistry: you can't compare zinc chloride or lithium batteries with alkalines this way.)

What you are mostly paying for with higher price "ultra" batteries is better high power performance. In other words, if you put a heavy load on them and try to draw more current, the ultra batteries should give a higher voltage and last longer. This is why the Tesco batteries have a surprising performance--you were putting quite a high load on them.

One thing to note about your testing is that capacity is not the only measurement of interest. When you discharge the batteries using your constant current load you should measure the voltage during the discharge and plot that against time. Higher voltages during the discharge denote better performance. (If you normalize the time by multiplying the elapsed time by the discharge current you will get an interesting family of graphs that you can plot on the same axes.)

[IanB]

This post on CPF illustrates the kind of discharge graph you can produce with the family of curves:

http://www.candlepowerforums.com/vb/showthread.php?263630-Investigating-the-Westinghouse-3-2-V-AA-LiFePO4-cells

In that case the discharge was done with a constant resistance load rather than a constant current load but the resulting graphs end up looking much the same either way.

[tom66]

The current version just spits out data via UART. I'm definitely planning to generate CSVs from it and my company will be publishing the full results on their website when complete. The current system uses an STM32 discovery board, and an ugly veroboard mashup constant-current load. The final version will probably be a pocket sized module with a 16x2 LCD and onboard EEPROM so it can log data (multiple units will be used - planning to test many different batteries to find the best value battery), all on a single PCB.

Anecdotally I noticed that both batteries stayed around 1.1V for most of their discharge life, before falling fairly rapidly to 0.8V when the test terminated.

[tom66]

More interesting results. Marks and Spencer own brand alkalines, made in Europe (Tesco ones are made in China) tested at 1180mAh, very good, basically the same as the Tesco ones. Even the Poundland ones managed 1050mAh, which for 1/5th the price of the Energizer is bloody good!

An Energizer is currently on a slow 150mA discharge overnight. Will find out how it did on Monday.

[IanB]

The slow discharge will be interesting. The capacity of an AA size alkaline cell when operated at maximum efficiency should be somewhere close to 3000 mAh (say 2800 or so). To get this number you would probably have to discharge at 25 mA, so expect to allow a week for the test.

If you measure the ultimate capacity of the cells in this manner, I wonder if more differences will show up between the cheaper and more expensive cells? Or, as I surmised above, is it all about the weight? (Have you weighed them, by the way? What is the weight?)

[tom66]

The slow discharge will be interesting. The capacity of an AA size alkaline cell when operated at maximum efficiency should be somewhere close to 3000 mAh (say 2800 or so). To get this number you would probably have to discharge at 25 mA, so expect to allow a week for the test.

Interesting - are alkalines really 3000mAh? I say this because I see most batteries rated 1500-2000mAh from reading up on this.

Approximately five of these testers will be built. They will be designed to fit onto an STM32 discovery board and contain the SO8 power MOSFETs, opamps, shunt and a bit of logic to cope with constant current discharge. I don't think they will be sold outside of the company though; I may get one for myself to play around with. The idea of having five is that multiple batteries can be tested at the same time. No LCD on these, just UART and computer interface.

If you measure the ultimate capacity of the cells in this manner, I wonder if more differences will show up between the cheaper and more expensive cells? Or, as I surmised above, is it all about the weight? (Have you weighed them, by the way? What is the weight?)

The weight of them was all approximately the same - about 25g. The Poundland ones weighed closer to 20g.

[IanB]

Interesting - are alkalines really 3000mAh? I say this because I see most batteries rated 1500-2000mAh from reading up on this.

Yes; see here: http://data.energizer.com/PDFs/E91.pdf

The weight of them was all approximately the same - about 25g. The Poundland ones weighed closer to 20g.

Then I suspect the Poundland ones to have significantly less capacity than the others (20% less "stuff" inside them).

[tom66]

Interesting - are alkalines really 3000mAh? I say this because I see most batteries rated 1500-2000mAh from reading up on this.

Yes; see here: http://data.energizer.com/PDFs/E91.pdf

Thanks for the link. I'll see on Monday how much it's accumulated. My discharge of 1A would seem to fit in with the graph giving 1250mAh. At 150mA I should expect close to 2500mAh.

Then I suspect the Poundland ones to have significantly less capacity than the others (20% less "stuff" inside them).

This matches up with being about 10-15% lower in capacity than the others.

[Nermash]

And now for something completely different:)

Accidentally I tested some Duracell and low cost Maxell alkaline just a few days ago, and then I saw this thread.

Tested on MAHA C9000, discharge current set @ 200 mA.
Duracell (came with Xbox360 controller) has expiry date 03/2017
Maxell has expiry date 10/2016
Both batteries weighted at cca. 24 g.

[IanB]

Tested on MAHA C9000, discharge current set @ 200 mA.
Duracell (came with Xbox360 controller) has expiry date 03/2017
Maxell has expiry date 10/2016
Both batteries weighted at cca. 24 g.

Wow. That's quite a difference, and a tough test too since the C9000 discharge cycle uses 1 A pulses which is harder on batteries.

Where was the Maxell made? Japan by any chance? I'm forming a hypothesis that Japan made alkalines are the best value on the market.

[IanB]

Accidentally I tested...

Um, how can you "accidentally" put batteries in a C9000 and program the discharge mode? That's quite a bit of sleep-walking you did there 

[billclay]

In my internship my task this week is to build a battery tester.

Can it test to see how long the battery lasts until it leaks all over and destroys your equipment?   

[tom66]

In my internship my task this week is to build a battery tester.

Can it test to see how long the battery lasts until it leaks all over and destroys your equipment?   

No but it can test how long until it destroys the tester . It can also enter suicide mode and attempt to draw 4A from the battery, which will fry the FETs quickly. Though there is a poly fuse...

[Bored@Work]

Approximately five of these testers will be built. They will be designed to fit onto an STM32 discovery board and contain the SO8 power MOSFETs, opamps, shunt and a bit of logic to cope with constant current discharge. I don't think they will be sold outside of the company though; I may get one for myself to play around with.

Well, well, well. The discovery stuff comes with this funny license http://www.st.com/internet/com/LEGAL_RESOURCES/LEGAL_AGREEMENT/LICENSE_AGREEMENT/EvaluationProductLicenseAgreement.pdf which tries to restricts what you can do with the board.

[tom66]

Approximately five of these testers will be built. They will be designed to fit onto an STM32 discovery board and contain the SO8 power MOSFETs, opamps, shunt and a bit of logic to cope with constant current discharge. I don't think they will be sold outside of the company though; I may get one for myself to play around with.

Well, well, well. The discovery stuff comes with this funny license http://www.st.com/internet/com/LEGAL_RESOURCES/LEGAL_AGREEMENT/LICENSE_AGREEMENT/EvaluationProductLicenseAgreement.pdf which tries to restricts what you can do with the board.

This is probably why they will not be sold, that and there are plenty of chargers with dischargers built in which would cost less than this if it were put into production. My RC battery charger/discharger for example cost £18 and it can discharge at varying power levels up to 5W. After all it's a UK manufacturing/design company I'm working for, not one in China, so it's quite expensive to build things.

[Nermash]

Accidentally I tested...

Um, how can you "accidentally" put batteries in a C9000 and program the discharge mode? That's quite a bit of sleep-walking you did there 

I have put them to test on purpose, what I meant was that my test coincided with this thread. Sorry, english in not my first language   

[IanB]

I have put them to test on purpose, what I meant was that my test coincided with this thread. Sorry, english in not my first language   

I know, I was just joking with you. "Incidentally" is probably the word you were looking for