Hi all,
Here is an interesting question (or a few) regarding cell retirement.
This should be useful to both experienced and new Li-Ion and other “unsafe” chemistry cell users.
What exactly are the telltale signs that a cell should be retired?
Low terminating voltage? It won’t drop below 3.4V even when loaded with direct drive XM-L? Slow charging? Fast charging? Heating of the cell?
Any reliable tests that can be done to determine if a cell is ready to be retired?
Thanks,
Viktor
I will toss them if they get discharged below 2 volts, less than 2/3 of original capacity, or heat up while charging.
For example, I have cells that won’t DISCHARGE below 3V or even higher.
I’ve been running this one cell on FF HD2011 on high, direct drive for hours. It’s low light output (which means the cell cannot supply any decent amount of current). And no matter how long I discharge it, it shows 3.4V by the time I pull it out of the flashlight.
Am I discharging it wrong, or is it ready to be retired? Is there a way to conduct a better test?
Viktor
For me, there are 4 main factors (obviously #4 applies to Li-Ion only):
Any of those occur - I toss the battery, at this point it’s pretty much worn out.
Any advice on how to measure capacity and internal resistance without a hobby charger?
I wanted to know this as well, great thread!
i paid 2.50$+ per LiIon cell (14500 or 10440).
i waited for 1.5-2.0 months for it to arrive (DX, DD).
i cherish my little round liions.
i wont toss them.
eva.
:D
Have fun burning your house down. :P
Well, for capacity you could hypothetically put it in a light with a known current draw. For simplicity say 1 amp. If you put a battery in and it runs for 2 hours 30 minutes until 3.6 volts then that means it has 2500 mah capacity.
Internal resistance can be done I think, but I'm not sure how.
Ok, I think I found it. This does not seem like a particularly good idea with lions though.
Note: I have not done this and the idea of short circuiting a battery does not seem particularly safe. I take no responsibility for anything that happens while trying this.
1. Measure voltage.
2. Measure resistance of another object. (Preferably something with high resistance)
3. Measure how much current the battery can put out through that other object.
5. Then do the math. Internal Resistance=(VBatt/Current)-Resistance of the other object.
I’d have to toss all my (new) 3100 Panasonics - my 4Sevens charger only charges them to 4.14v!
It’s a bit problematic to keep the current constant throughout whole range of voltage 4.2V -> 3.0V. Some electronics must be involved, but I don’t know enough to design such a circuit. If any of the more experienced users know, please chime in.
With most budget flashlights 2A measured with full cell at say 4.15V will reduce to 0.8A at 3.1V. These are made up numbers, but tell you the rough idea of current shifts throughout cell charge. It requires a bit of complex electronics to keep the current constant, hence the big $$$ for the real constant current drivers and flashlight that have them. Using a simple resistor would have similar unreliable effect.
Usual cell voltage range is 4.2V to 3.0V, some cells go lower than that, some new cells have higher full charge voltage. (Empty cell is not 3.6V as Scaru suggested.)
Come to think of measuring cell internal resistance… it should not be too hard, someone just needs to write down some actual numbers and examples. I know basic Ohm’s law, which is enough for this, but it’s too late at the moment, and I’m kinda in the middle of another project. 
Perhaps tomorrow.
Charger not charging any cells above 4.14V is a charger issue. I am assuming healthy chargers, measuring equipment and common sense are used. At least I try to.
Come on, HKJ, PilotPTK, texaspyro, and other guys with in depth knowledge of cells and electronics…. your voice would be of great value here.
Viktor
It is true that 3.6 is not empty but it is what I and alot of other people consider empty in multicell lights. I do that purely for safety even though I do get lower capacity.
I agree, I do the same. Once it’s below 3.6-3.7V I consider the cell to be less than optimal, and would rather charge it than continue using it.
For calculating capacity it’s not the correct value to use in the calculation, even though we rarely use those last few 100mAh.
The simplest thing is a variable resistor (a big one).
For the money, these things are not too shabby (they used to be cheaper, and there are several sellers… search for “internal resistance meter”). Just make sure not to hook them to the battery backwards. They don’t like that at all. Also, for best accuracy, first measure the IR across the negative battery terminal (or short the probes together). Subtract that reading from the reading across the battery.
When you are not satisfied with the cell’s performance.
I use 12V 35W halogen bulb. At full battery I have 1,65A current, at 3V 1,44A current.
4Sevens charger terminates CV stage at current way higher than datasheet-specified. That charger is not a good reference point, it’s voltage precision isn’t very good either.
Measuring cell internal resistance without purpose-built test equipment is a bit problematic. You can do it with a good, stable power resistor and a good DC volt meter but the results tend to not be very accurate due to things like battery surface charge, cell heating, and resistor heating.
But here goes (caveat emptor, I’m doing the math on the fly while drinking good beer):
1) Get a GOOD low ohm, stable power resistor that will load the cell down to a reasonable current for the cell. Poor resistors have inaccurate values that change as the resistor heats up. For a lithium cell, a 4 ohm resistor would be around a 1 amp load and draw 4 watts from the cell. Call the combined resistance of the resistor and wires and connections Rload.
2) Connect the resistor across the cell. Make sure you have good, low resistance connections and wires (easier said than done, those spring loaded black plastic cell holders suck). Wait around 30 seconds and measure the voltage AT THE CELL, not at the end of the wires at the resistor. Call this reading Vload (Note that the voltage reading will be changing as the cell/resistor/life heats up and the cell is depleted… one of the reasons that this method is not all that accurate)
3) Disconnect the resistor and QUICKLY measure the voltage at the cell. You don’t want to give time for the cell voltage to recover due to surface charge effects. Call this reading Vcell.
4) Calculate Vdrop = (Vcell - Vload)
5) Calculate Internal Resistance = (Vdrop * Rload) / (Vcell - Vdrop)