Hello there,

I am currently designing a battery test for CR2032 button cell batteries. I am gathering a lot of info, but trough the years a lot of methods have been developt, for many applications. As said, my application is an Lithium battery.
Now the main method is loading the battery with a resistor, and measuring current and voltage etc.
Another method is using ac signal. But i cannot find the real benefits of this.
Why would i prefer the AC method, against the DC methode? Is it more efficient? I know the results of the 2 are VERY different, because with AC i am measuring the impendance and frequency response, but what does this tell me?
The batterys im testing are from a transmitter, some of them degrade too fast, so we want to check them before we assemble them in the transmitter.

 

The batteries are not rechargeable so that limits the ways you can test them. The different methods available are explained in the attached link below. Take your pick:
https://batteryuniversity.com/learn/article/battery_test_methods

 

The batteries are not rechargeable so that limits the ways you can test them. The different methods available are explained in the attached link below. Take your pick:
https://batteryuniversity.com/learn/article/battery_test_methods

i understand that. I already checked this website from a to z, but it is most theoretical. it does not show many circuits, components etc etc.

 

The batteries are not rechargeable so that limits the ways you can test them.
https://batteryuniversity.com/learn/article/battery_test_methods

And please, tell me, what are those limits?

 

Hi,

There is no one test that can tell all, but a simple load test is usually indicative of the batteries ability to deliver power to a circuit especially when dealing with the very same type and make and model of the battery for every test.

For example, if you were to test an alkaline AA cell vs a heavy duty AA cell it would be hard to tell what was going on, but if you always tested the alkaline AA cell you would realize that there are certain factors that show how the battery can put up with a current being drawn from it. So experience with the same make and model battery goes a long way with knowing the state of a randomly selected battery (of the same make and model).

If you have a bunch of the same type cells then you can develop a test for those cells. Simply load the battery with a known resistance using the same resistance for every battery you test. Measure the voltage and note the reading you get from a new cell and maybe from a cell that has operated for a known time. You can then start to get a feel for what is good and what is not good.

An AC test is an attempt to keep the operating condition of the battery constant so that the series resistance is more apparent. When you load it with a resistor the battery starts draining so it's harder to tell. But if you apply a pulse which means you connect the resistor for a known time period then you can get information from that too, although you may have to experiment with the best time period to use for the pulse. Too short and you measure the short term equivalent capacitance and too long you start to drain the battery too much.

I have several AA alkaline cells laying around here, i can pick up any one of them and measure the voltage using a digital volt meter and be able to tell roughly how used they are or if they were new, and even if i buy a pack of AA cells i can measure the unloaded voltage and get an idea how long they have been sitting on the shelf (ie new or slightly aged) and this works between manufacturers too. But with a heavy duty cell it is harder to tell because for one i dont have as much experience with them.

In the old days it was harder to tell for most batteries using a volt meter because all we had was an analog type with its limited measurement resolution. In the age of digital meters it gets easier because we get more consistent good resolution readings.

 

When a battery degrades its internal resistance increases. This is a better indicator of the batteries ability to provide power more than the no-load voltage across the battery.

As the internal resistance increases, the output voltage will fall with increasing load on the battery.

I don't know anything about the AC method but you can determine the internal resistance just as well with a DC measurement by applying increasing loads.

PS. I have a simple test jig for testing AAA and AA batteries. It consists of a knife-switch, a simple analog meter movement with no scale markings and a flashlight bulb as the load. The meter indicates the battery voltage with the switch open and how the voltage drops under load when the switch is closed. I simply watch to see how much the meter drops along with the brightness of the bulb to give me an indication of whether to keep or toss out the battery.

 

Maybe i didnt tell enough about my project. Whe are using batteries of type CR2032. We place them in transmitters. Frequently, we get the transmitters back, with the note that the battery is dead. When we read the data from it, it didnt make de calculated lifetime. After long testing, the only think that can be wrong is the battery. We are no longer searching for the problem in the transmitter. i want to test those batteries, before we place them in the transmitter and send them to te client. Which part of the battery is faulty, we dont know yet. I want to figure it out with testing, testing and testing. The ohmic test will be one of the test. But maybe, the internal resistance (which you calculate with this test) is not the problem. Now i know that this internal resistance changes over time, depending on load, enviremont etc etc.
But during the tests i do not want to drain the batterie. Therefore i was looking at AC testing, but cannot find very much info on how to test this way, like schematics. I also want to be able to tell something about the capacitance in the battery (search: randles model)

 

And please, tell me, what are those limits?

The most useful test for a rechargeable battery is the capacity test. This is done by discharging a fully charged battery to it's cut off, measuring the current and voltage at regular intervals. This can not be done with a non-rechargeable battery if you want to use it after the test.

 

When a battery degrades its internal resistance increases. This is a better indicator of the batteries ability to provide power more than the no-load voltage across the battery.

Hi,

It used to be like that in the old days, but digital meters changed that because we get good repeated readings and good resolution. I can tell what state many of my batteries are in just by measuring the open circuit voltage and i do it all the time.

Also, cell phones use voltage measurements to determine state of charge and there are guides for this too on the web. This works really good when you have experience with the battery type.

However, he may be dealing with a group of batteries that are often just bad which can be for a number of reasons so he may have to test other things too. Most new batteries are good, but if they sometimes came in bad a test would have to be developed for them too.

 

I have several AA alkaline cells laying around here, i can pick up any one of them and measure the voltage using a digital volt meter and be able to tell roughly how used they are or if they were new, and even if i buy a pack of AA cells i can measure the unloaded voltage and get an idea how long they have been sitting on the shelf (ie new or slightly aged) and this works between manufacturers too. But with a heavy duty cell it is harder to tell because for one i dont have as much experience with them.

Only measuring the voltage is not the way to go. There are good batteries with lower voltage, bad batteries with relatively higher voltage, it doesnt tell you anything about the capacity of the battery.

An AC test is an attempt to keep the operating condition of the battery constant so that the series resistance is more apparent. When you load it with a resistor the battery starts draining so it's harder to tell. But if you apply a pulse which means you connect the resistor for a known time period then you can get information from that too, although you may have to experiment with the best time period to use for the pulse. Too short and you measure the short term equivalent capacitance and too long you start to drain the battery too much.

Thank you for this comment. So AC is more efficient then DC? Must be, cause DC testing seems like a thirsty one.

 

Maybe i didnt tell enough about my project. Whe are using batteries of type CR2032. We place them in transmitters. Frequently, we get the transmitters back, with the note that the battery is dead. When we read the data from it, it didnt make de calculated lifetime. After long testing, the only think that can be wrong is the battery. We are no longer searching for the problem in the transmitter. i want to test those batteries, before we place them in the transmitter and send them to te client. Which part of the battery is faulty, we dont know yet. I want to figure it out with testing, testing and testing. The ohmic test will be one of the test. But maybe, the internal resistance (which you calculate with this test) is not the problem. Now i know that this internal resistance changes over time, depending on load, enviremont etc etc.
But during the tests i do not want to drain the batterie. Therefore i was looking at AC testing, but cannot find very much info on how to test this way, like schematics. I also want to be able to tell something about the capacitance in the battery (search: randles model)

Hi again,

What hit me was that some batteries work and others dont, and i assume you are always using new batteries or you are just asking for problems anyway.

So it is starting to sound like there could be two reasons for this problem.
1. The device is overrated. Most batteries can not provide the projected service life so many fail.
2. You are getting bad batches of batteries.

In case 1 you have to do some redesign or modify the specifications.
In case 2 you should do a full discharge test on several batteries and compare to what is expected. If the manufacturer can not provide good batteries all the time then switch to another manu.

 

The most useful test for a rechargeable battery is the capacity test. This is done by discharging a fully charged battery to it's cut off, measuring the current and voltage at regular intervals. This can not be done with a non-rechargeable battery if you want to use it after the test.

Ah like that, i forgot about that. Your right, cant be done with mine.

 

Only measuring the voltage is not the way to go. There are good batteries with lower voltage, bad batteries with relatively higher voltage, it doesnt tell you anything about the capacity of the battery.



Thank you for this comment. So AC is more efficient then DC? Must be, cause DC testing seems like a thirsty one.

You're welcome, but measuring the voltage works very well for known battery types. As i said,, i do it all the time and can judge very well what state the battery is in. This is with mostly AA alkalines and various Li-ion cells.
I could quote some examples, but i believe your problem could be for another reason.

 

Only measuring the voltage is not the way to go. There are good batteries with lower voltage, bad batteries with relatively higher voltage, it doesnt tell you anything about the capacity of the battery.

Hi again,

What hit me was that some batteries work and others dont, and i assume you are always using new batteries or you are just asking for problems anyway.

So it is starting to sound like there could be two reasons for this problem.
1. The device is overrated. Most batteries can not provide the projected service life so many fail.
2. You are getting bad batches of batteries.

In case 1 you have to do some redesign or modify the specifications.
In case 2 you should do a full discharge test on several batteries and compare to what is expected. If the manufacturer can not provide good batteries all the time then switch to another manu.

Thank you for this comment. So AC is more efficient then DC? Must be, cause DC testing seems like a thirsty one.

Absolutely right. We expect that it is case 2. After all the tests we did it cannot be the device. Perhaps we will switch to another manu, but first we are making a battery tester.

 

Ah like that, i forgot about that. Your right, cant be done with mine.

Hi,

No it can be done with any battery. You just have to sacrifice a few to test. I did this a lot too when i buy a pack of AA cells that i never bought before from a different manu. I test at least one cell by draining it all the way down.
Once drained i can compare it with other cells of the same type to see which is better or the same.

 

So which circuit would you recommend, for AC testing? And what is there to expect? Which readings should i get, where to measure etc.

 

Hi,

Just wondering, do you have an oscilloscope?

 

Hi,

Just wondering, do you have an oscilloscope?

offcourse i do, tektronix tds2002 and philips pm3055. Why are you wondering?

 

The most useful test for a rechargeable battery is the capacity test. This is done by discharging a fully charged battery to it's cut off, measuring the current and voltage at regular intervals. This can not be done with a non-rechargeable battery if you want to use it after the test.

But why would i do this? This is information that is given by the manufacturer in the datasheet of the battery right? I mean, i assume that you want to test how long the battery can last under certain loads? ie, the capacity test? So whats the point of this test? Testing if the specs of the battery are the same as the manufacturer says? The battery is 230 mAh with a load of 5k6. Thats 430 Hours before the battery is drained. Seems to me like a long test

 

offcourse i do, tektronix tds2002 and philips pm3055. Why are you wondering?

Hi,

Because there is a square wave test you can do if you can monitor with a scope.
I'll see if i can find the pdf or you can search yourself.

If this problem is subtle you may have to resort to testing and logging, then shipping, then noting which ones came back and which ones didnt, then compare waveforms so in the future you can start to predict which cones would come back and simply dont use those. As time goes on you should be able to narrow it down to which cones will fail. This is like a last resort.