Li-Ion life cycle management

NO, absolutely not. That is literally throwing away useful cycle life.

The factors that cause deterioration in Li-Ion cells are heat, terminal voltage, use - not necesarily in that order..

Depending on whether your longevity is to be considered in cycles or years, and whether this is used regularly/intermittently/infrequenctly/rarely the tradeoffs may well be different. There is no free lunch in sight, only compromises that may or may not suit your objective(s).

For maximised "storage" life, I would aim to keep the cell(s) around the 40-60% state of charge, and cool without being paranoid about the latter. The higher the maintained cell voltage, the higher the rate of deterioration. Ditto higher temperature, but trying to beat room temp generally isn't worth the effort.

For a cell/battery in semi-regular use the cycle life can be improved by reducing the charge voltage if this is under your control. Dropping to 4v1 from the more typical 4v2 will probably double the cycle life while reducing available capacity by something like 20% or less. Unfortunately appliances (eg laptops) don't give you that option, but packs using discrete chargers may do so either directly or indirectly.

And in regard to temperature ... Considerable heat is generated in laptops, and manufacturers go to great pains to try and keep the battery away from it. Most do a fair job, but there isn't enough freedom in the design of ever-more-powerful-and-compact machines to avoid a temperature rise in the battery pack from adjacent hardware. So if you are a daily laptop user and operating from a wall supply, the single biggest factor you can control is the battery temperature. Once it has charged sufficiently for your next period of dependency, pull the battery out of the machine.

ltage,

, and

deoffs

s that

higher

o higher

ces (eg

do so

ains

isn't

avoid

re a

tor

ently

My application will likely be the following: equipment stored for months, then broken out and set up for a 5-day exercise. During the exercise, battery packs will be swapped out as needed, but the user prefers that the swap times be at least 12 hours apart. The batteries are combined in tactical packages whose nominal voltage is a around

voltage,

and

tradeoffs

higher

from

(eg

so

avoid

That might not be a cost effective application for Li-Ion. Isn't their self discharge alone a few percent per month?

Have you considered D-cells?

Do

e.

voltage,

ars, and

tradeoffs

ises that

e 40-60%

he higher

itto higher

ort.

to 4v1 from

iances (eg

may do so

t pains

ere isn't

to avoid

u are a

factor

iciently

They will be charged at the start of the exercise and they will be judged to be successful if the charge lasts 12 hours.

voltage,

and

tradeoffs

that

higher

higher

from

(eg

do so

pains

isn't

avoid

are a

factor

sufficiently

Ok, if the client wants it that way, the client is king :-)

But what if it is successful for a couple years and then one fine day it gets charged but conks out after 2hrs? That's what happened to me on our stove backup supply last month. I just don't think this is the right app for a rechargeable. Then again, the customer is king.

=A0Do

ife.

al voltage,

years, and

e tradeoffs

omises that

the 40-60%

=A0The higher

=A0Ditto higher

ffort.

by

g to 4v1 from

ng

pliances (eg

s may do so

eat pains

there isn't

es to avoid

you are a

t factor

fficiently

ne.

es

The last time I got involved in a portable instrument design, we used primary lithium cells, for the same reason we are using rechargeables now.

It's in the spec.

No, our measurements place it around 0.1%/month.

voltage,

and

tradeoffs

higher

from

(eg

so

avoid

I'm comfortable with the two charts at the bottom of that page, but s**te that self-discharge spec worries me. They are either being ultra-conservative or the quality of cell is not all that good. As I mentioned in another reply, our continual experience with chinese/korean 18650 cells is 0.1%/month.

Just as a check, I measured two samples from the workshop. One - a single cell, no protection appendage, was stored fully charged in 2003. Now measured 3v99. The other is a three-cell pack with protection module, stored at the same time at about 3v75 according to my notes. Currently cells read about 3v28 - 3v30.

Anyway, let's assume the packs you use are at worst case and lose 4%/mo. Annual charge-before-use would be OK, and given that figure I'd be putting them back into storage fully charged - the 40-60% storage SOC is based on self-discharge below 1%/mo.

I'm assuming that there *is* the opportunity to charge-before-use. If not, that

With the usage pattern you suggest, I'd also leave them at 4v2 charge. They'll still outlive the client. (If you were talking a laptop used in daily commute or similar, I'd certainly be back down at 4v1.)

IME Lithium cells, of most technologies, they are intolerant of overcharge, excess charge/discharge rate, or discharge below 30% of capacity. YMMV

oltage,

s, and

adeoffs

es that

higher

to higher

t.

nces (eg

y do so

pains

e isn't

o avoid

are a

ctor

iently

Does that mean we should discount the advertised capacity figures and only expect 70% of it in use?

Well, the data from Henry's link states "Self Discharge: < 4 % per Month". Quite a lot, actually. But he wrote that they'll be charged right before the event, so should not be a problem. Lack of storage capacity after a few year will be, though.

My experience with rechargeables is that there will be a reduction in capacity after many cycles (~hundreds). Is there a similar decline in capacity over time if they are not used much?

Yes, definitely. I just don't know what the numbers are for Li-Ion. I'd talk to manufacturers about it.

For example, the usual Lead-Acid gel cells often go after 3-4 years. One of our sets here rapidly dropped to about 20% of capacity after four years and when I asked other engineers about their experience most said "that's about normal".

Same for old NiCd. I found most to be practically dead after less than eight years despite having experienced less than 20-30 charge cycles.

I have here a Li-Ion pack in an Acer (AcerNote Lite 370) laptop which is date-stamped week 37, 1996. It still delivers about 2/3rds of its initial run-time. What it DOESN'T do is live in the machine when AC power is used.

Considering the low number of cycles that your application is likely to rack up in say ten years (6 cycles per exercise x N exercises per year x 10 years) the deterioration will largely come from (a) initial build quality issues and (b) temperature issues. Temperature as the one you can control to some extent.

The effect is most probably best considered as the cumulative effect of elevated temp x time at elevated temp, If your storage temperature is sane (as is the case with my OLD Acer pack) I wouldn't foresee a dramatic capacity loss.

Correct on all counts. That is why it is imperative to use proper chargers and pack mangement modules for these chemistries, and not improvise.

Decent chargers ensure a controlled current-limited constant voltage charging regime.

Protection modules integrated into the packs ensure that excess temperature or low temperature disqualify charging. They monitor individual cell voltages and disqualify the pack if these differences exceed a threshold value. They prevent discharge below an end voltage which is typically 3v0. They also (depending on selection) limit the discharge current allowed.

One of the best descriptions I have found for the proper care and feeding of Li-Ion cells is in the datasheet for the Maxim MAX1737 - a controller we used several years ago in a commercial charger design. It is a worthwhile read.

No, quoted capacity is invariably usable capacity down to that cutoff value.

ers and

ing

e or

ges and

prevent

ding on

of

sed

ad.

e.

Then saying that the last 30% is unusable is a mis-statement.

ir

al

ed.

ck up

the

(b)

nt.

evated

s the

Hide quoted text -

The customer already has a battery storage discipline - they are kept on chargers in an air-conditioned shelter. I guess some random testing of batteries that they have been keepoing in storage for a while would be a good data point.

In use, the batteries are not shut up in the equipment but eefectivley hang out in the air, so the on;ly temperature extremes they will wexperieince will be daily weather variations and self-heating dueing discharge.

On Fri, 23 May 2008 07:32:02 -0700 (PDT) in sci.electronics.design, Richard Henry wrote,

It's available, but in my experience using it will reduce battery life considerably. Better not to push the discharge limit unless you really need that capacity. .

up

elevated

Hide quoted text -

If the charger is intelligent (*) that may be acceptable for Li-Ion, but a better practice would be to remove them entirely, and only return them if the SOC had dropped materially.

(*) A charger suitable for continual interface to Li-xx cells should actually disconnect charge once the current has tapered off to Imax/20 or so, and should not reconnect until the OC voltage has dropped to a threshold value i.e. NOT a "trickle charge" arrangement. Again, the MAX1737 explains and exhibits this regime.

Indeed. But remember also that deterioration occurs as a result of sustained high cell voltage. To maintain it at (say) 4v20 will increase the deterioration. Removal and allowing to discharge down to a 50% SOC before permitting further charge will reduce this factor, while leaving the danged things at 50% is probably the best approach - unless you need to quickly deploy without warning.

which is as good as it gets without aircon.