How do you know how Healthy your Li-ion cells are?

[abarth_1200]

Hello all, long time user of Li-ion cells but never stopped to think about how healthy they are now and how to tell.

I have had 2 AW 18650 cells for about 2 years and they have maybe been through about 5 full cycles and about 20 half cycles, meaning they havent been discharged fully, I had them in my TK11 for the first year and they didnt get much use, one was always charged ready to when one was in the light getting some use, I would cycle them every month or so just to keep them fresh, I would run the torch for about a half hour then recharge that cell and stuck the fresh cell in for the next month of use.

Now I know cells last longer if they are stored at 75% but is 4-6 weeks too long for them to be fully charged.

Also how would I go about finding out how much charge they can hold now, ie how healthy they are. I have a multimeter but I have a feeling its not all to accurate, I have just fully charged one of my half depleted 18650 cells and tested the voltage straight off the charger 4.19 volts using a Ultrafire WF139, I then checked a brand new Energizer and a panasonic CR123 cell, both were between 3.10 and 3.20, these are meant to be 3V cells, is mutli meter off and if so is it over compensating whic means my 18650 cells are only hitting about 4 volts, which isnt a full charge in regards to li-ion chemistry.

Any help would be much appreciated and any ideas on how to get my cells up to 4.2

EDIT I have just had a thought, the fact that my MM is reading the CR123 cells at over 3 volts maybe down to the fact that they are not under load?

 

[kaptain_zero]

I've got 3 Fresh (date coded 2010) unused Rayovac CR123 batteries and they all measure to a calibrated 3.22v. A pair of slightly used 4Seven cells read at 3.06v on my calibrated meter.

There's an unspoken problem with Digital Multimeters when we use them to measure things such as LiIon voltages down to the milivolt range and assume the reading we get is accurate..... it's not. DMMs are not accurate, only relatively accurate... yes, the can read out to many decimal places but unless calibrated against a true reference source, you don't know. I have two identical meters, one older than the other but otherwise the same. One reads 0.01v lower than actual, the other reads 0.04v higher than actual. I thought I had a problem with my chargers when I first got them because I only had the older meter and it was reading a termination voltage of 4.24v on the two chargers I had.... After a brief chat with a local electronics tech friend, I ended up purchasing another meter and when I measured the same batteries off the same charger, it now read 4.19v which is great.... except, I still had no clue as to which meter if any, was reading correctly.

So, I took the two meters to my tech friend and had them compared to 4 commercial grade bench DMMs, one Fluke, one Beckman and two others I can't remember... of those 4, none of them had the same reading but all were within 0.04v of each other.... thus we went to an Analog Lab instrument that had been calibrated recently and was known to be dead on accurate in this range and thus I found that the new meter was 0.01v low and the old one was 0.04v high. Now I can add or subtract the error to my readings and know where I'm at. However, we're not done yet because cheap and even expensive meters can and will drift if components are compromised, aging or failing. That's why the tech has 4 different bench meters..... on a regular basis one gets checked against another to see if any of them are drifting. My two cheapo meters will give me a reasonable comparison if all of a sudden, I get a reading that doesn't seem right.... a quick check with the other one will show if the spread between them is still 0.05v. If it isn't, then one or both of the meters may have begun to drift, but I won't know which one unless I have a 3rd meter to compare them to. :-)

A very long winded answer to a simple question that isn't quite so simple really.

Regards

Christian aka

Kaptain "I'm working on building an accurate Voltage reference to avoid these issues in the future" Zero

 

[Black Rose]

Kaptain "I'm working on building an accurate Voltage reference to avoid these issues in the future" Zero

Such an item already exists.

 

[HKJ]

Get a hobby charger, then you can run a test cycle on your batteries an measure the actual capacity.
There are two "problems" with hobby chargers:

• They are usual designed for 12-15 volt supply, i.e. you need a external power supply (There are exceptions to this).
• They do not include any cradle for batteries, you have to make your own. Either a clamp or some magnets can be used for this.

But when you have fixed the above "problems" a hobby charger is very useful to check LiIon and NiMH batteries on. It can (of course) also be used to charge them, but a dedicated charger is much easier to use and can usual handle multiple batteries at different discharge state.

 

[abarth_1200]

Hmm I was affriad I would might have to get a more expensive charger and read a lot more about charging and li-ion than I have done already.

So is it a waste of time going out and buying a new multimeter, the one I have now only cost £10. I dont really have anything to test it against, other than plugging it into one of the 5V sockets in my PSU for my computer, how accurate are they?

 

[cottonpickers]

From what information you've given, it sounds like the rechargeable 18650 is in OK order. - I base this on the fact that you've charged it and its holding a good charge (4.19v). You could recharge it again, and see if the voltage drops after being charged.

Even if your dmm isn't very accurate this is a good way to get a quick estimate of health of the li-ion battery. A good one should hold its charge, whereas a poor one will see its voltage drop to 4.1v or less fairly quickly.

Why your primaries show different voltages is a little strange.

You could always take your current dmm into maplin and ask to check it against theirs with a view to buying a new/additional one if yours is sufficiently out.

 

[abarth_1200]

OK thanks for the replies everyone, looks like my cells are doing pretty well.

Now I have a question about CR123 cells but I should start a new thread for that

 

[Justin Case]

There's an unspoken problem with Digital Multimeters when we use them to measure things such as LiIon voltages down to the milivolt range and assume the reading we get is accurate..... it's not. DMMs are not accurate, only relatively accurate

"Relatively accurate" -- sounds like you are referring to precision.

So, I took the two meters to my tech friend and had them compared to 4 commercial grade bench DMMs, one Fluke, one Beckman and two others I can't remember... of those 4, none of them had the same reading but all were within 0.04v of each other.... thus we went to an Analog Lab instrument that had been calibrated recently and was known to be dead on accurate in this range and thus I found that the new meter was 0.01v low and the old one was 0.04v high. Now I can add or subtract the error to my readings and know where I'm at.

That's true only if you don't switch the DC voltage meter range.

Get a hobby charger, then you can run a test cycle on your batteries an measure the actual capacity.

Some hobby chargers -- the iCharger 208B for example -- also can measure cell internal resistance down to the milliohm.

 

[45/70]

The easiest way to determine the condition of LiCo Li-Ion cells is to fully charge them and take an OC voltage reading immediately after the charge terminates. Let the cell rest for a half hour or so and then take a second voltage measurement. The cell's general condition can be determined by how much the cell's voltage has dropped. This method only gives a rough estimate of the cell's condition. A super accurate voltage meter is not necessary, but it should be in reasonable working order.

No drop = 100% condition
0.05 Volt drop = 75% condition
0.10 Volt drop = 50% condition
0.15 Volt drop = 25% condition
0.20 Volt drop = time to replace

The best way to determine a LiCo cells condition, is to run it through a battery analyzer under various discharge rates, and compare the results to a cell in new condition. This is a much more accurate way to determine the condition of cells. As already mentioned, a hobby charger can also be used for this purpose, as well. A cell that retains less than 80% of it's original capacity, is considered at EOL and should be considered for replacement.

Dave

 

[kaptain_zero]

"Relatively accurate" -- sounds like you are referring to precision.



That's true only if you don't switch the DC voltage meter range.



Some hobby chargers -- the iCharger 208B for example -- also can measure cell internal resistance down to the milliohm.

DMMs read to incredible resolution.... as in it can detect a change of perhaps 0.0001v but even though the meter might read 0.0001v does not mean that that is the actual voltage, it might actually be 0.0011v and we do not know unless it's calibrated to a known source....

Yes, switching the range will require calibrating each range though auto ranging meters will tend to select the correct range for a given voltage.

I'm just concerned that many folks run out and purchase a $20 meter or even a used higher end meter off ebay and then assume it's dead nuts accurate. It isn't, unless you've verified it against a known value.... High end professional equipment do come calibrated, usually with a printed report but these are typically in a price range we would not consider for our use. And for that matter, calibration to the level of resolution we're attempting to use here, aka 4.2x volts would need to be done on a rather regular basis or at the very least have multiple meters that can be compared against each other so as to be able to detect when one starts to drift.

As for hobby chargers..... again, resolution is easy, accuracy is hard..... I'd no more *trust* a *HOBBY* charger, any more than I would trust a *HOBBY* DMM to read to high accuracy if neither had been independently verified by myself or someone I trust.... But that is my point of view as a hobbyist with a background in Amateur Radio and electronic building and design. I am by no means an expert, but I do rely on opinions of people I know who make a living working in electronics repair.

Buying a $100 meter or a $100 hobby charger will not guaranty that you are not overcharging a lithium Ion cell even if the meter reads 4.20v at the end of a charge cycle, unless you KNOW the meter is accurate because you have had it verified/calibrated.

All electronic parts used in chargers and DMMs are sold in various accuracy ranges.... 10%, 1%, 0.1% etc. and each individual component is somewhere within the specified range. Now you assemble a device using all these variable components.... each device will have a different set of variations.... all of the devices should fall withing the ultimate spec of the design and most will fall somewhere in the middle (bell curve), but some will be at either of the extremes and some may fall outside those too! The more a company spends on the components to get improved accuracy, the more accurate they can make the device. The more quality control a company does, the more they can weed out devices that are too far from the design specifications... each of these measures costs money.... lots of money sometimes. To be assured you have a very accurate measuring device, out of the box, you will have to spend a significant amount of money. Is it possible that you could get a highly accurate cheap meter? The bell curve says yes, but is it likely enough so that you could rely on it...... nope. Only way to tell is to do your own quality testing.

Sorry, I didn't mean to run on... but I felt it needed to be pointed out.

Regards

Kaptain "Blabbermouth" Zero

 

[Justin Case]

Calibration is a known issue and has been discussed previously on CPF. As Black Rose linked, Malone Electronics has a line of voltage standards that one can use to calibrate a DMM, hobby charger, or whatever.

Just because a DMM may have auto ranging is not relevant to the issue of measurement accuracy. When you switch voltage ranges, you are using a new set of resistor divider circuitry. If you want accurate measurements for all of the meter's voltage ranges, you need to calibrate all of them. Auto ranging isn't going to help you with that. Auto ranging is a substitute for the end user having to engage his brain to understand the magnitude of the expected measurements to adjust his meter range manually.

And don't forget about the effects of ambient conditions -- temp, humidity.

 

[kaptain_zero]

Agreed, I just hadn't seen anyone mention that bit and so I thought It should be mentioned. And you are so right about temperature/humidity having an effect, just hand holding a battery whilst measuring it, can cause the reading to change.

And yes, thank you Black Rose for that link.... I may just purchase one now, by the time I get all the bits together and make the board, I'll have spent at least that much in time and effort, never mind the actual components.... I was going to go with the chip mentioned on this site: http://lategahn.2log.de/index.php?calibration-standards-introduction

Christian

I forgot to mention that what I meant with the Auto Ranging meters is that if you calibrate such a meter using a 5v source or even a LiIon battery that you are reading with different meters, the auto ranging meter will tend to select the same range each time it is presented with the typical voltages provided by a single LiIon.... Naturally, if it switches to a different range, all bets are off unless you've calibrated that range as well.

 

[Justin Case]

I forgot to mention that what I meant with the Auto Ranging meters is that if you calibrate such a meter using a 5v source or even a LiIon battery that you are reading with different meters, the auto ranging meter will tend to select the same range each time it is presented with the typical voltages provided by a single LiIon.... Naturally, if it switches to a different range, all bets are off unless you've calibrated that range as well.

"Tend to"?

If you are testing Li-ions, IMO you should be manually setting the voltage range.

Suppose I have a DMM that has voltage ranges of 0V-0.4V, 0.4V-4V, 4V-40V, 40V-400V, and 400V-1000V.

Maybe the Li-ion cell being tested is partially discharged to an OCV less than 4V. If I depend on autoranging and don't pay attention, the meter could be set to the 0.4V-4V range. Unfortunately, this is a different range than what would be used for a fully charged Li-ion at 4.20V. Since you are already in non-thinking, "auto" mode, will you notice that the Li-ion reading is less than 4V? Maybe. Maybe not.

Auto ranging is not a panacea.