Messages in this thread | | | From | Baolin Wang <> | Date | Mon, 11 Mar 2019 19:11:24 +0800 | Subject | Re: 5.0-rc1 Open circuit voltage tables -- 3afb50d7125bc |
| |
Hi Pavel,
On Mon, 11 Mar 2019 at 18:23, Pavel Machek <pavel@ucw.cz> wrote: > > On Mon 2019-01-07 20:05:55, Pavel Machek wrote: > > Hi! > > > > Can we do something with that code? > > > > First -- tables should not be necceessary. Formula for li-ion open > > circuit voltage is well-known, and does not really depend on the > > device -- only on the chemistry of battery. > > > > state of charge = f( voltage, current, internal resistance ) > > > > (And yes, internal resistance depends on temperature and health of the > > battery).
As we know, experiments had proved that the OCV had a steady relationship with the state of charge for a given chemical characteristic. So many companies' charging method will use OCV table to look up the initial capacity of the battery according to the current OCV value. That's why we introduce the OCV table.
> > > > Second, even if we need tables, why is it doing multiple allocations? > > > > info->ocv_temp[index] = -EINVAL; > > > > That's just wrong. Ocv_temp is temperature in celsius. Temperatures go > > below zero.
This requirement is from Qualcomm. Since they have multiple OCV tables in different temperatures.
> > Ping? That code is just bad. Does someone care?
Sorry, I missed your previous email. -- Baolin Wang Best Regards
| |