Li-Ion cells. Determining SOC?

[45/70]

As my Li-Ion cell collection increases in size, so do the different types. What is the best way to determine the approximate state of charge of:


1. 3.0 Volt LiCo Li-Ion cells.

These are normal LiCo cells with diodes incorporated that in effect, reduce the voltage from 4.2 to about 3.2 Volts under load.

When I measure the OC voltage of this type cell, I often can get 3 different voltages. I'm assuming they are, the cell w/o diodes, the cell with 1 diode, and the cell with both diodes. Is there an easier way to get a more accurate/dependable voltage reading of this type cell?


2. LiFe/A123 Li-Ion cells.

Detemining the SOC of these cells by reading the OC voltage seems likely not to work, as the voltage stays relatively flat throughout the discharge, as compared to LiCo cells. Can it be done, or is there no relationship between voltage and SOC in LiFe cells, similar to CR-123A primaries?


3. LiMn/IMR Li-Ion cells.

I don't have any of these......yet. Is the SOC determined in the same manner as for LiCo's, or?


4. What's the best flashlight? What's the best torch? Which came first, the chicken, or the egg? I'm the 999,999,999th visitor, did I win?


:oops: Sorry, scratch #4, wrong forum. My apologies, DM51. :D

Dave

 

[VidPro]

put a small load on, check the voltage. compare it to knowns

discharge it all the freaking way (within spec), read the capacity or runtime

just charge it to much, what the heck, as long as it is voltage termination charging there is no "overcharge" at the end of charge. then it will be at that capacity it was last measured.
to store it, lower it back down a bit by running it, or for long term storage lower it halfway by using it to about 1/2 its runtime.

carry a spare, or better yet use a device with LOTS of Battery in it.

they both came at the same time, without the egg no chicken without the chicken no egg, so they had to both exist, dont make it harder than it is :-)

 

[45/70]

Yeah Vid, I'm familiar with all the "hope and pray" and rigging up a load methods. It's just that with a LiCo cell, it's so easy to tell the present capacity, as well as when it's time to recycle. I was just hoping that there was some easy way to check the others.

My biggest concern is 6 AW 3.0 Volt RCR-123's I have. They are getting on 4 years old now and I'd like a more accurate way to determine if they're at or under 4.00 Volts after charging/resting, as pertains to EOL cycle and whether it's time for the recycle bin. I have 8 Powerizer LiCo 3.0 Volt 123's the same age, but if you apply the test leads just right, they aren't so hard to beat the diodes. They are all still around 4.10 Volts rested.

I'm pretty much replacing the 3.0 Volt LiCo's with LiFePO4s. I'm taking a capacity hit for sure, but I have better piece of mind. Still, it'd be nice to know how to gauge them, as well.

As for the chicken and the egg? My conclusion, neither one ever existed. "There is no spoon". :)

Dave

 

[VidPro]

My biggest concern is 6 AW 3.0 Volt RCR-123's I have. They are getting on 4 years old now and I'd like a more accurate way to determine if they're at or under 4.00 Volts after charging/resting, as pertains to EOL cycle and whether it's time for the recycle bin. I have 8 Powerizer LiCo 3.0 Volt 123's the same age, but if you apply the test leads just right, they aren't so hard to beat the diodes. They are all still around 4.10 Volts rested.

so they are regged down simplely 3.7s anyways right? you could still use the voltage method, either way, either measure with the dmm with no load, and assume your reading close to the cell voltage, or add a load and read the low then add in the voltage drop. it leans a little this way and a little that way wheather your charging or discharging. depending on the meter it will see right through mostly.
there is just a voltage drop across other than that its the same as li-co? if you dont apply the load, then you guess that the meter your using this week wont lean it when testing, probably not good to guess.


charge them up, on the proper stuff, read the voltage, leave it sit for a few days, read again, if it is going down by its lonesome then it is not doing well too.

but that still leaves the "best" way: have some kinda charge control. charge - discharge test the capacity, return to safe storage, store, wait 6 months repeat, then recycle when doomed. if they are just going to be parked forever, they are way old, recycle so your not doomed :-)

Always apply a load, put festoon bulb on leads, measure across 123 with festoon bulb attached, no more guessing, the load will guarentee the voltage drop as it will be with a load, the load will make the cell work a bit instead of just sitting there, and that voltage would be more usefull, if it starts going down fast, if it cant maintain the load, then it is getting suckey.

a 4 year old AW cell could still be completly good for 2-3 more years, depends on the cell item that is under there.

There didnt have to be a spoon, when your hungry enough you will see.

 

[45/70]

The problem is when you put the leads to one of these cells (the LiCo 3.0 Volt) the reading changes. I don't have one here at the moment, but it'll show something like 4.12, then in a second 3.21, then 3.84. Those are numbers I just pulled out of the air because I don't have one here, but the point being, the readings aren't allways the same, consistent, but not the same. Sometimes, particularly with the AW versions, you never get the actual cell voltage (in my example, the 4.12 Volt), so I'm never sure about what I"m getting. It seems to all depend on whether the diode(s) is/are recognized. I just wondered if there was a way to consistently get a reading. It wouldn't matter which one, as long as you knew which it was, you could correct the reading. As I said, the readings are consistent, you just don't know which one your getting because of the diodes.

Dave

 

[VidPro]

ahh yes, hence the load must be placed on, because your meter is putting just a wee bit of load on and throwing off any usefull results.
use the load, then add in the voltage drop.

i have 3 meters, and i never know what time it is :) nor would i remember which one did what, hence the load, draw some current through the diode or whatever, and it will stabalise at a close enough voltage drop to be usefull to know what is under there.

 

[45/70]

ahh yes, hence the load must be placed on, because your meter is putting just a wee bit of load on and throwing off any usefull results.

Yeah, that's the problem, I figured that. Maybe I can build a test lead with a resistor built in to solve the problem. Any idea what value I'd need?

I'd still like to know if there's a way to check the capacity of LiFePO4 cells also. I suspect these will need a similar setup, but I have no idea what voltages would correspond to what capacity. The big problem I see with LiiFePO4 chemistry, is that the OC voltage and even under load on a CBA II, is so flat during discharge that it isn't going to be easy.

Dave

 

[VidPro]

yes, the li-fe would have to be the toughest of all. but so to would ni-mh, and they can be read with a load too, its just knowing about where it would be FOR that item, with your load with your meter.
so its not like i could hand you numbers, you will see your numbers and get a fair idea, with your rig.

charge them up let them settle for a while, then test, discharge them half way let them settle then test, discharge them down all the way (within spec) let it settle test. then for the time that the battery is acting the same, those numbers will be the same.
then post them here , cause i love to see much and much data and info.

a 10ohm resister is adequite for most of the stuff, i use a fat 10W one, so it doesnt cause problems with much of anything. then i use bulbs sometimes too.

 

[45/70]

It sounds like the only way to figure all this out, it to do it the hard way. :sigh: I was hoping somebody, somewhere, had figured all this out already.

I have 8 new LiFePO4 123's. Once I've sorta broken them in, I'll see what I can do. It seems the best way to do it would be with the CBA II. I'll just run them down to the start of the "knee" and call that 0%, then divide the discharge up into sections, like maybe 75, 50, and 25 percent and check rested voltages.

Don't hold your breath. This may be a while. :)

Dave

 

[VidPro]

compare, that is how i use it, is this one charge this one not, how does it look when compared with this one that is charged. cant always remember the numbers i get, but i can see if say 4 of them would form a team or if one is strangely off.
its just like ZTS testing but with a much fuller range of digits (so to speak) vrses 5 leds lighting up one way or another, and set at a factory as best as they could determine it.