On 11/18/2015 10:05 PM, pedro wrote:
On Wed, 18 Nov 2015 11:59:37 -0800, Jeff Liebermann
wrote:

Yep, and there's a problem. If you charge to 100% and leave it there,
the LiIon battery slowly deteriorates. See Fig 5 at:
http://batteryuniversity.com/learn/article/how_to_prolong_lithium_based_batteries

Yup. The two classical Li-Ion deterioration factors are temp and SOC
- in each case, the higher it is the higher the deterioration.

That's one reason you don't see any LiIon UPS/SPS battery backup power
system. There are plenty of retrofit kits and replacement batteries,
but nothing from the OEM's. Same with cordless phones, electric
shavers, and other devices that run permanently on chargers. About
the only way to do it is to use a battery that's about twice the
calculated capacity ma-hr, and only charge it to 50% of capacity.

When we were doing Li-Ion capacity/cycling testing (in association
with charger design) we tried moving from the cell manufacturer's
"specified terminating voltage for full spec capacity" (4v20) to 4v10.

We found that we were achieving between 85% and 90% of the 4v20
capacity. We didn't have enough lifetime to waste doing deterioration
testing for 4v10, but I'm sure that the average always-plugged-in
laptop would get double the battery lifetime (cycles, years) at 4v10
terminating voltage with only that commensurate loss of run time.

It's a pity manufacturers of laptops, smartphones etc are obsessed
with maximising run-time as a selling spec.

Well...offer a laptop with less battery life than the competition
and see what happens to your market share.
It's not manufacturers, it's customers who are obsessed with run-time.
If you're gonna get a new smartphone every two years, do you really
want to trade run-time for a battery that lasts 10 years instead of three?

What would be interesting is a customer available parameter that
traded battery life for run-time.

So, what was the termination voltage in your final charger design?