Samsung INR-18650-29E optimal charge/discharge protocol

[mrbill]

I recently purchased some battery packs using these cells from cell_man. As some of you may know cell_man offers chargers for his Lithium-Ion battery packs that give the user the option to charge to 3.8v, 4.1v or 4.2v per cell, selectable by a switch on the charger.

Charging to 3.8v per cell prior to storage or sitting idle is said to increase the calendar life of the cells. 3.8v/cell is approximately 50% of full charge. Charging to 4.1v/cell is recommended for normal use, while charging to the full 4.2v/cell provides an additional 10% capacity at the expense of shortened cycle life.

Here's my question:

I'm trying to come up with a charging/discharging protocol that maximizes the calendar life of the batteries. I'm primarily concerned about calendar life as I see myself putting no more than about 60 cycles annually on any one of these batteries.

Which protocol would result in maximizing calendar life given that my discharges will use 90% of the battery's capacity? One that has me using more frequently the top 10% (90-100% SOC) of a battery's capacity, or one that has me using more frequently the bottom 10% (0-10% SOC)? Which SOC region should I most avoid?

Thanks.

 

[parabellum]

Best to avoid both of them, but I personally am more afraid of 0-10%, it is where the battery gets hot and out of balance. I would charge to 90% and top to 100% just before leaving.

 

[riba2233]

Store batteries at cold and dry place at 40% SOC?

 

[dogman dan]

Store for longer than overnight using the 50% charged setting for sure. More importantly, never store in a hot place, like a summer garage. Worst of all for calendar life is hot and fully charged. Inside at room temp at least, till cooler weather comes. That will be the single most important thing for calendar life. Even doing your best, you will not have enough battery by year 3. You only gave yourself 10% of headroom. GET SOME MORE BATTERIES, SO YOU HAVE MORE CAPACITY NOW. If you get more capacity now, you can charge to 4.1v for now, and only need to charge to 4.2v when the pack gets older.

Others will disagree, and want charts and references to prove it. But the consensus seems to be that the very best lifespan is had by never charging past about 80%, nor discharging below 20%. But that leaves only 60% left.

You should buy some more battery, so you can charge to 4.1v and still have enough left. You will lose about 15% if you charge to 4.1v. So you'd be better off to charge to 4.2v, and NEVER discharge below about 3.5v. IMO, I'd rather charge to 4.2v then discharge immediately, than try to suck that last 2% from the bottom. Once below about 3.6v, don't hit that battery with a high discharge rate. Below 3.6v, limp home very slow. Once over the cliff, the internal resistance will be very high, and they won't like a 2c discharge then.

If you need an early start, use the 4.1 switch to charge the night before, then top up in the am.

 

[riba2233]

You are talking about lipo, nca cells can be used down to 2.5 V without problem.

 

[okashira]

From what I have gleaned, the cells should be quite happy when stored at low charge, like 5-10%.

MFR's use 40% only to ensure the cell can be stored very long term (years) taking into account self discharge, and discharge of commonly used battery protection circuits.
What I mean is, storage at 10% will be even less stressful then storage at 40%. You just need to watch them so they don't go below 0%. but the difference is probably pretty small.
If you will have the cells completely disconnected from a load (no BMS) you could safely store them at a lower charge.

 

[okashira]

Store for longer than overnight using the 50% charged setting for sure. More importantly, never store in a hot place, like a summer garage. Worst of all for calendar life is hot and fully charged. Inside at room temp at least, till cooler weather comes. That will be the single most important thing for calendar life. Even doing your best, you will not have enough battery by year 3. You only gave yourself 10% of headroom. GET SOME MORE BATTERIES, SO YOU HAVE MORE CAPACITY NOW. If you get more capacity now, you can charge to 4.1v for now, and only need to charge to 4.2v when the pack gets older.

Others will disagree, and want charts and references to prove it. But the consensus seems to be that the very best lifespan is had by never charging past about 80%, nor discharging below 20%. But that leaves only 60% left.

You should buy some more battery, so you can charge to 4.1v and still have enough left. You will lose about 15% if you charge to 4.1v. So you'd be better off to charge to 4.2v, and NEVER discharge below about 3.5v. IMO, I'd rather charge to 4.2v then discharge immediately, than try to suck that last 2% from the bottom. Once below about 3.6v, don't hit that battery with a high discharge rate. Below 3.6v, limp home very slow. Once over the cliff, the internal resistance will be very high, and they won't like a 2c discharge then.

If you need an early start, use the 4.1 switch to charge the night before, then top up in the am.

Lipo discharge voltages are VERY different from a 29E.

 

[dogman dan]

Yep, I screwed that up. But don't they still go over the discharge cliff in more or less the same place? So nothing much left below 3.5v ish?

I am not backing down one inch on my statement that his pack is too small for the longest possible lifespan, if he needs 90% of it. In a year or so, he will start losing some capacity, and any windy day he will not have enough with 100%.

 

[riba2233]

They have much below 3.5V. They have useful capacity all the way down to 3 or 2.5V, depends of the discharge rate. Just take a look at some tests of nca cells.

 

[okashira]

Yep, I screwed that up. But don't they still go over the discharge cliff in more or less the same place? So nothing much left below 3.5v ish?

I am not backing down one inch on my statement that his pack is too small for the longest possible lifespan, if he needs 90% of it. In a year or so, he will start losing some capacity, and any windy day he will not have enough with 100%.

Depends on current. Not much left below 3.5 at low load, (less then 0.3C)
But at higher current there could be a lot of mah below 3.5

 

[dogman dan]

Sorry to spew more of my ignorance on this thread, but explain that one for me a bit more?

Wouldn't a higher discharge rate result in less capacity, not more? Or at least, a sag to lvc before you suck much out? Don't all cells sag under load more as they reach the last bit of capacity, and resistance increases?

With lithium cobalt, I find that you can discharge that last 5% of safely usable capacity, (down to 3v or so) without unbalancing the cell so bad if you keep the discharge rate low enough. In such a situation, I'm just trying to limp home on 2 amps or so, keeping the discharge rate at .2c or something tiny like that. All I'm getting is a few watts of assist, so I don't have to push a cogging dd motor home.

If I just hammer the throttle right to the end, I tend to have an unbalanced pack to deal with, and some packs I did that to died years ahead of ones I never did that to.

Back to the original discussion, do you all agree that if he charges to 4.1v, he'd lose too much capacity off the top and have to take the pack to 0% or too close to it for optimal lifespan? Better to charge to 4.2v, but not until he'll ride immediately. That way, only a few minutes at full charge, and less bouncing off the packs bms lvc. Doesn't unbalancing the pack, and having to leave it sitting there with half the cells at 4.2v while the bms works overnight cause damage too?

Occasionally, he'll still have to balance the pack no matter what. But won't it be better if he has to do it less often? That's why I keep repeating he needs to increase his pack size one more cell, so he only needs 70-80% of it each ride.

Lastly, if you have a graph for these cells handy, can you put it up here? I googled, and just got lots of places selling them, but as usual, no real info about them.

 

[riba2233]

Here you go:

I think it's better not to charge to 4.2 V ever.

Look at this:

 

[okashira]

Here you go:

I think it's better not to charge to 4.2 V ever.

Look at this:

Looks like 65% of the cell capacity is below 3.5V at 5A.
PF/BE performs a bit better, I think.

 

[riba2233]

Of course they do, B is high impedance cell, I've put that for reference.

 

[parabellum]

Looks like 65% of the cell capacity is below 3.5V at 5A.
PF/BE performs a bit better, I think.

Nope, it does not.
We have no resting voltage at specific SOC. We need to stop discharge at specific point and give little rest to the cell to make this conclusion.

 

[riba2233]

Looks like 65% of the cell capacity is below 3.5V at 5A.
PF/BE performs a bit better, I think.

Nope, it does not.
We have no resting voltage at specific SOC. We need to stop discharge at specific point and give little rest to the cell to make this conclusion.

http://lygte-info.dk/info/BatteryChargePercent%20UK.html

 

[parabellum]

http://lygte-info.dk/info/BatteryChargePercent%20UK.html

Cool link, thanks. So 18% on 3.5V at 3A discharge for NCR18650B.

 

[okashira]

Looks like 65% of the cell capacity is below 3.5V at 5A.
PF/BE performs a bit better, I think.

Nope, it does not.
We have no resting voltage at specific SOC. We need to stop discharge at specific point and give little rest to the cell to make this conclusion.

I have news for you. You don't always have time to stop and let the cell rest for 15 minutes so you can measure resting voltage. :wink:

 

[dogman dan]

Thanks for that chart. I know I could have dug for it, but now it's there in this thread for everybody who looks at it in the future.


I had no idea the knee, or cliff in the discharge curve was that much lower than lico. looks like 3.2v would be equivalent to 3.5v for lico. Quite a lot of difference. Stopping the discharge of this cell at 3.3v would be quite conservative then.

I do agree also, don't charge to 4.2v. But this guy has his pack sized so charging to 4.1 will mean he never has enough, ever. Size your pack bigger, then you can have the luxury of never charging to 4.2v.

But if you must charge to 4.2v, the best compromise is to wait till just before your ride to do it.

 

[parabellum]

I have news for you. You don't always have time to stop and let the cell rest for 15 minutes so you can measure resting voltage. :wink:

I use 3 techniques to judge the SOC of the packs by resting V. None of them is in load condition.
1) At the moment you start your commute after rest. (is the most precise one)
2) Double check on red lights. (almost as good as first one)
3) Let throttle down to 0A position for few seconds during commute(still good enough)

 

[dogman dan]

Even if it takes 60 seconds to stop see what approximate resting voltage is,, It's worth it to take the time to see if you are in the last 10% of your capacity.

For lipo, I want to see how it is when I see 3.5v per cell under load,,, and for these 3.2v under load.

 

[opperpanter]

Here's the chart of 29e:

http://lygte-info.dk/review/batteries2012/Samsung%20INR18650-29E%202900mAh%20%28Blue%29%20UK.html

My pack runs at low discharge rates, so not much left after 3.3V.
I might even have to stop 3.4V because I'm mostly running around 0.5C.

 

[opperpanter]

What to when you have charged a battery to go for a ride, but you have to cancel the ride.

And the next ride will be a couple of days or a week away.

Will I better discharge the battery to ~50% again and recharge before the next ride?

Or just leave it charged to 90% for the week?

I know storing it at 90% (or even 100%) may cause some wear, but discharing and recharing will cause wear as well...

If it's charged to 100%, might be best to discharge to at least 90%?

 

[Nobuo]

What to when you have charged a battery to go for a ride, but you have to cancel the ride.

And the next ride will be a couple of days or a week away.

Will I better discharge the battery to ~50% again and recharge before the next ride?

Or just leave it charged to 90% for the week?

I know storing it at 90% (or even 100%) may cause some wear, but discharing and recharing will cause wear as well...

If it's charged to 100%, might be best to discharge to at least 90%?

I you charged to 100% (4.2V per cell) and you need to cancel the ride to the next week, in that case I would drain a 10% at least to reach 4.1V or 4.05V (some 48V headlight installed would help to do itself).

Then leave it till the next week in a cool place, although storing at 4.1V-4.05V is not the best, it would be completely insane to drain a whole half cycle just for the next week the need to recharge again to the 100%, that cycle will harm more and it is a waste of energy.

I charge always to the 100% (4.2V), I leave a pair of extra hours on the charger for balancing purposes, I estimate the time, so just after I disconnect the charger I can start to use. Then If I finish and eg. I have a 80% left, I just let it rest for the next time. Again I will use from the 80% if I won't take so much.

So when I reach a 20% - 30% I left again stored, and the next time I make again a 100% charge. That is a reliable and right way to use, also to recharge when reaching 40% would be right.

If you use a battery cycle daily, in that case charging to 4.1V have sense, but take into acount that your BMS needs to be configured to start the balancing on 4.1V and also the charger must be configured to start a CV phase on 4.1V

 

[riba2233]

Well a few days or a week wont do anything, problem is if you'd left it for longer time.