Bbattery charging "mystery"
Hi Folks, Yesterday I observed something which surprised me (running Debian Etch in VirtualBox in Win XP on an Advent 8117 laptop). WHen I sat down at the desk, I noticed the battery-charging LED flashing. When I check the battery level monitor (in both Debian and XP) it was at 20% and charging. An hour later it was fully charged. I generally leave the laptop running continuously, days on end, and its power supply is permanently connected to the mains. It's possible that something disturbed to connection for a while (perhaps most likely where the mains cable connects to the PS brick), and that something later re-established it (I do tend to swim in a pile of clutter so stuff gets pushed around), but the event prompted me to wonder the following. Might it be that, if a laptop (at any rate for some models) is kept continuously connected to the mains for days on end, then for the sake of battery health a re-charge cycle is initiated, whereby the input from the mains-connected PS brick is internally disconnected for a while, until battery-level drops to a fairly low level, and is then re-connected? It is, of course, known that it's good for battery life to submit it to a discharge/recharge cycle every so often; but I've not heard of this being built-in to laptops' power monitoring. (But then there's a lot of stuff I haven't heard of). Ted. -------------------------------------------------------------------- E-Mail: (Ted Harding) <Ted.Harding@manchester.ac.uk> Fax-to-email: +44 (0)870 094 0861 Date: 22-Feb-09 Time: 09:44:56 ------------------------------ XFMail ------------------------------
On Sun, 2009-02-22 at 09:44 +0000, Ted.Harding@manchester.ac.uk wrote:
It is, of course, known that it's good for battery life to submit it to a discharge/recharge cycle every so often; but I've not heard of this being built-in to laptops' power monitoring. (But then there's a lot of stuff I haven't heard of).
This is only true for specific battery chemistries. Specifically NiCd and NiMH. Lithium Polymer and Lithium Iron don't need a full charge-discharge cycle because they don't suffer "memory effects". The only care these cells need is that you avoid storing them below 30% charge and avoid heavy charge/discharge cycles in freezing temperatures . However the battery monitor stuff that tells you percent discharge, runtime etc does require calibration every so often and I know that some Thinkpads at least will eventually perform an automated cycle to recalibrate if left on Mains for a log period of time.
On 22-Feb-09 10:42:09, Wayne Stallwood wrote:
On Sun, 2009-02-22 at 09:44 +0000, Ted.Harding@manchester.ac.uk wrote:
It is, of course, known that it's good for battery life to submit it to a discharge/recharge cycle every so often; but I've not heard of this being built-in to laptops' power monitoring. (But then there's a lot of stuff I haven't heard of).
This is only true for specific battery chemistries. Specifically NiCd and NiMH.
Lithium Polymer and Lithium Iron don't need a full charge-discharge cycle because they don't suffer "memory effects". The only care these cells need is that you avoid storing them below 30% charge and avoid heavy charge/discharge cycles in freezing temperatures .
However the battery monitor stuff that tells you percent discharge, runtime etc does require calibration every so often and I know that some Thinkpads at least will eventually perform an automated cycle to recalibrate if left on Mains for a log period of time.
Thanks for the clarification about battery cycling needs and recalibration, Wayne. Also for your info that Thinkpads, at least, will do this thing. It may well be, therefore, that the same applies to the Advent which I'm using -- it is many weeks since I last "ran it down", and when I inspected the connections when I noticed it doing the recharge, they seemed sound (nothing loose). Nice to know! (My other worry, of course, was that the PS box might have had something go wrong with it). Ted. -------------------------------------------------------------------- E-Mail: (Ted Harding) <Ted.Harding@manchester.ac.uk> Fax-to-email: +44 (0)870 094 0861 Date: 22-Feb-09 Time: 10:54:37 ------------------------------ XFMail ------------------------------
On Sun, Feb 22, 2009 at 10:42:09AM +0000, Wayne Stallwood wrote:
On Sun, 2009-02-22 at 09:44 +0000, Ted.Harding@manchester.ac.uk wrote:
It is, of course, known that it's good for battery life to submit it to a discharge/recharge cycle every so often; but I've not heard of this being built-in to laptops' power monitoring. (But then there's a lot of stuff I haven't heard of).
This is only true for specific battery chemistries. Specifically NiCd and NiMH.
Lithium Polymer and Lithium Iron don't need a full charge-discharge cycle because they don't suffer "memory effects". The only care these cells need is that you avoid storing them below 30% charge and avoid heavy charge/discharge cycles in freezing temperatures .
The 'memory effect' in NiCd batteries (I don't think it occurs in NiMh batteries at all) is very specialised and very specific. The chances of it occurring in normal usage are vanishingly small. It was originally diagnosed in some batteries on a satellite which had extremely regular charge discharge cycles due to the satellite going round the earth and thus in and out of sunlight at exactly the same time each day. The effect *only* occurs with these extremely regular charge/discharge cycles. All the wonderful "our charger discharges your battery first" extras are completely pointless. The real guaranteed way to make your NiCd batteries useless is to leave them charging indefinitely which is what a *lot* of older/cruder chargers do, especially the ones that come with cordless tools. Using good modern delta-V chargers (usually called 'automatic') there's no need to discharge batteries before charging. Oh, the other *big* killer of NiCd cells is charging/discharging more than one cell in series. Discharging multi-cell batteries fully is a real no-no as the weaker cells get 'reverse charged' before the stronger cells are fully discharged. Similarly fast charging the cells in series results in the weaker cells reaching full charge first and overheating before the stronger cells are fully charged. Some systems *have* to charge batteries in series, if you do this then the batteries need to be well matched to minimise weak/strong problems. I use a lot of AA and AAA cells and I *only* use chargers which charge all the cells individually, none of the chargers discharge cells first, I'm still using cells bought many years ago which must have been through a *lot* of cycles. All of which is, of course, completely irrelevant to the pressent discussion! :-) (Oh by the way a *careful* Google search for "NiCd memory effect" will find the full details of the satellite I relate above)
However the battery monitor stuff that tells you percent discharge, runtime etc does require calibration every so often and I know that some Thinkpads at least will eventually perform an automated cycle to recalibrate if left on Mains for a log period of time.
-- Chris Green
On Sun, 2009-02-22 at 11:04 +0000, Chris G wrote:
The 'memory effect' in NiCd batteries (I don't think it occurs in NiMh batteries at all) is very specialised and very specific. The chances of it occurring in normal usage are vanishingly small.
You sound knowledgeable on this subject but I have to disagree with the above. HiMH do suffer from memory effect, it is less of a problem but it is still present. Also most NiCd cells in "normal" use will show the memory effect very quickly if the cells are repeatedly not fully discharged or only partially charged. I was experiencing this in the early 90's when I was a bench technician at an electronics company and one of my first jobs was replacing less than year old (therefore in warranty) batteries in the first generation cordless phones. Their typical usage was possibly the worst application case for a NiCd because either the handset was picked up from the cradle and used briefly before being placed back down or it is left to discharge and then removed from the cradle before the charge is complete.
All the wonderful "our charger discharges your battery first" extras are completely pointless. The real guaranteed way to make your NiCd batteries useless is to leave them charging indefinitely which is what a *lot* of older/cruder chargers do, especially the ones that come with cordless tools.
Of course good cordless tools now use Li-pol batteries..For NiCd and NiMH a trickle charge at 10/C or preferably 15/C will probably be better for the batteries than a fast charge from an intelligent charger. and certainly at the latter rate you could safely leave it there for some time but as you say not indefinitely. The market desired faster charge times so we moved to smart chargers that charge at 1-3/C and then shut off (or sometimes drop to a 15+/C charge to keep the pack topped up. The auto discharge is useful at single cell level but with series packs there is a danger of cell reversal.
On Sun, Feb 22, 2009 at 10:00:02PM +0000, Wayne Stallwood wrote:
On Sun, 2009-02-22 at 11:04 +0000, Chris G wrote:
The 'memory effect' in NiCd batteries (I don't think it occurs in NiMh batteries at all) is very specialised and very specific. The chances of it occurring in normal usage are vanishingly small.
You sound knowledgeable on this subject but I have to disagree with the above. HiMH do suffer from memory effect, it is less of a problem but it is still present. Also most NiCd cells in "normal" use will show the memory effect very quickly if the cells are repeatedly not fully discharged or only partially charged.
I was experiencing this in the early 90's when I was a bench technician at an electronics company and one of my first jobs was replacing less than year old (therefore in warranty) batteries in the first generation cordless phones. Their typical usage was possibly the worst application case for a NiCd because either the handset was picked up from the cradle and used briefly before being placed back down or it is left to discharge and then removed from the cradle before the charge is complete.
Well go and search on Google, the *first* hit I got for a "NiCd memory effect" was this one:- http://www.repairfaq.org/ELE/F_NiCd_Memory.html ... and the next hit says exactly the same, as do lots of others. They're from reputable places too.
All the wonderful "our charger discharges your battery first" extras are completely pointless. The real guaranteed way to make your NiCd batteries useless is to leave them charging indefinitely which is what a *lot* of older/cruder chargers do, especially the ones that come with cordless tools.
Of course good cordless tools now use Li-pol batteries..For NiCd and NiMH a trickle charge at 10/C or preferably 15/C will probably be better for the batteries than a fast charge from an intelligent charger. and certainly at the latter rate you could safely leave it there for some time but as you say not indefinitely.
Long term on a trickle charger at the 'normal' rates will certainly kill the batteries in the long term. -- Chris Green
On Mon, 2009-02-23 at 10:09 +0000, Chris G wrote:
Well go and search on Google, the *first* hit I got for a "NiCd memory effect" was this one:-
http://www.repairfaq.org/ELE/F_NiCd_Memory.html
... and the next hit says exactly the same, as do lots of others. They're from reputable places too.
Ok so "memory effect" is bad (but common) terminology for an effect I have personally witnessed and measured where a NiCd cell's apparent capacity is reduced by a re-profiled discharge curve caused by repeated part charge/discharges. I'll agree with you for the sake of argument that it isn't a true "memory" effect but regardless you can add more curve to the discharge slope by part charges or part discharges which in turn limits the effective usable capacity of the cell in many applications. Up until a short while ago during a big clearout I even had graph stats from computerised charging equipment that would quantify this. I used to many years ago produce and race high performance chassis for the (highly competitive) Model racing clubs...we are talking £1k for a 10th scale rolling chassis here so naturally we took every step to get as much performance as we could from the 1800mAH NiCd cells we were limited to. Even the cell manufacturer data sheets backed this up and I can assure you we weren't overcharging. Ok so that doesn't quantify with "normal" use as the charge/discharge times we were working with were in the order of 20/5 minutes, but when I have measured such an effect..plotted the changes in the discharge curve and had supporting documentation from the cell manufactures. I am afraid I am going to take that over a couple of hits on Google, even if it does put me in the "There will always be people who spark off another flame-war by posting that they have seen "memory" and they will not hear otherwise." category, but I don't care :)
On Sun, Feb 22, 2009 at 10:00:02PM +0000, Wayne Stallwood wrote:
On Sun, 2009-02-22 at 11:04 +0000, Chris G wrote:
The 'memory effect' in NiCd batteries (I don't think it occurs in NiMh batteries at all) is very specialised and very specific. The chances of it occurring in normal usage are vanishingly small.
You sound knowledgeable on this subject but I have to disagree with the above. HiMH do suffer from memory effect, it is less of a problem but it is still present. Also most NiCd cells in "normal" use will show the memory effect very quickly if the cells are repeatedly not fully discharged or only partially charged.
I was experiencing this in the early 90's when I was a bench technician at an electronics company and one of my first jobs was replacing less than year old (therefore in warranty) batteries in the first generation cordless phones. Their typical usage was possibly the worst application case for a NiCd because either the handset was picked up from the cradle and used briefly before being placed back down or it is left to discharge and then removed from the cradle before the charge is complete.
All the wonderful "our charger discharges your battery first" extras are completely pointless. The real guaranteed way to make your NiCd batteries useless is to leave them charging indefinitely which is what a *lot* of older/cruder chargers do, especially the ones that come with cordless tools.
Of course good cordless tools now use Li-pol batteries..For NiCd and NiMH a trickle charge at 10/C or preferably 15/C will probably be better for the batteries than a fast charge from an intelligent charger. and certainly at the latter rate you could safely leave it there for some time but as you say not indefinitely.
The market desired faster charge times so we moved to smart chargers that charge at 1-3/C and then shut off (or sometimes drop to a 15+/C charge to keep the pack topped up. The auto discharge is useful at single cell level but with series packs there is a danger of cell reversal.
Why has this (and the previous message) resurfaced? :-) -- Chris Green
On Tue, Mar 10, 2009 at 12:48:55PM +0000, Wayne Stallwood wrote:
On Tue, 2009-03-10 at 12:36 +0000, Chris G wrote:
Why has this (and the previous message) resurfaced? :-)
Because I thought the point was worth reiterating or because my Evolution is hopelessly broken and decided to resend everything it had sent since the last restart. :)
Aha, someone else with a love/hate relationship with Evolution! :-) I'm trying to persuade Evolution to show me its dates in DD/MM/YYYY format instead of MM/DD/YYYY but to no avail at the moment. My wife's near identical xubuntu 8.10 installation shows dates in MM/DD/YYYY format but mine doesn't. Same locale, same OS, same Evolution! Grrrr. -- Chris Green
2009/3/10 Wayne Stallwood <ALUGlist@digimatic.co.uk>:
On Tue, 2009-03-10 at 13:44 +0000, Chris G wrote:
Aha, someone else with a love/hate relationship with Evolution! :-)
Hmm Love/Hate....I am only really getting one side of that at the moment if I am honest
heh. Didn't I have a conversation with you at one of the early ALUG Ipswich meets, where you tried to convince me Evolution was good? I ditched it in favour of, forgive me, Outlook. I just couldn't get it play well with our corporate Exchange server. It's probably better now, but at the time it was painful! Peter.
On Tue, 2009-03-10 at 14:59 +0000, samwise wrote:
heh. Didn't I have a conversation with you at one of the early ALUG Ipswich meets, where you tried to convince me Evolution was good? I ditched it in favour of, forgive me, Outlook. I just couldn't get it play well with our corporate Exchange server. It's probably better now, but at the time it was painful!
Oh god anything talking to Exchange other than Outlook or IE (at a push) is going to be painful. I think I probably said something along the lines of "if you need to access Exchange from Linux then Evolution is the best option" because really it is that or the basic mode OWA in mozilla. But whatever it was if it was even slightly positive towards Evolution I take it back :)
2009/3/10 Wayne Stallwood <ALUGlist@digimatic.co.uk>:
Oh god anything talking to Exchange other than Outlook or IE (at a push) is going to be painful.
I think I probably said something along the lines of "if you need to access Exchange from Linux then Evolution is the best option" because really it is that or the basic mode OWA in mozilla. But whatever it was if it was even slightly positive towards Evolution I take it back :)
Actually, we weren't talking about the Exchange connection so much - you were just bigging Evolution up in general. I disagreed given my own experiences and eventually we left it that my bad experience might well have been just down to the flakiness of the exchange plugin (tho I wasn't convinced :) ). Sounds like you've come around ... heh Peter. P.S. To be fair, the Exchange plugin I used a few years back, was the old OWA screen-scraper one that Novell open-sourced when they bought Ximian. I believe there's been work more recently in Evolution on using the native binary Exchange protocol, so things may have improved. I just don't have the patience to go through it all again
participants (4)
-
Chris G -
samwise -
Ted.Harding＠manchester.ac.uk -
Wayne Stallwood