10.2.4 killing iBook / PowerBook battery

C

cellfish

Depressed as hell
Looks like Apple screwed up big time and will have a lawsuit on their hands very soon. I am one of the people affected by this and I don't find it to be the least bit funny. Quoted from MacFixit:

Short iBook/PowerBook battery life; getting a replacement

We continue to be overwhelmed with reports of diminished PowerBook/iBook battery life after upgrading to Mac OS X 10.2.4.

Most users report that their systems have decreased by more than half of their original capacity, in some cases lasting only a few minutes before needing to be recharged.

Lucy Johnson's case, below, is typical of the dozens of reports we have received. Generally after installing Mac OS X 10.2.4, users affected by the problem see their battery life measurement slowly decline over multiple charges:

"I received my 12" PowerBook on 2/7/03 with Mac OS X 10.2.3 installed. I did not have any problems with battery life or charging at that time. I 'reconditioned' the battery as is advised. When Mac OS X 10.2.4 became available I installed it but didn't immediately see any problems. One day I noticed that the light (where power adapter plugs into computer) was glowing amber instead of green. I checked the battery life indicator on my menu bar and found that it was at 97% even though the adapter had been attached for at least 12 hours.

"I decided to run the battery completely down again and recharge. After ample time for recharging, (overnight) the battery was at 83%. I started to get worried and took a look at Apple's Discussion board, finding many others with similar experiences. Following their best advice I again ran my battery down and recharged, this time to only 68%."

If you have tried reverting to Mac OS X 10.2.3, or toggling Energy Saver settings as previously suggested, your battery may need replacement by Apple. We're not sure how Mac OS X 10.2.4 is draining the life of these batteries, but in some user cases, going back to Mac OS 9 or earlier versions of Mac OS X have no effect. The battery seems to be permanently incapacitated by the 10.2.4 updater.

For the PowerBook (12-inch), you will need to get the following Apple part replaced:

* 603-1204 Battery, LITH ION
* 46W/630-3961 PCBA
* DC to DC

As usual, if you purchased an AppleCare support membership, you will find the search, location, and wait period for your replacement battery much easier and faster.

Feedback on this issue? Drop us a line at late-breakers@macfixit.com.
 
Wow. I can't believe that no one here has had anything to say about this. I've even mentioned it in chat a couple of times, and nobody seems to have heard about it. This post is to bump this so that everyone who uses "View new threads" will see it (again), and to try to gather my findings on the subject a bit into one place...

So, there I was, on iChat with some folks, and after seeing some of the other problems discussed here, I asked if any of them had had any problems with 10.2.4, and I was very strongly encouraged to just go ahead and upgrade; they weren't having any problems. Thus began my odyssey into the realm of power management and battery technology on the iBook/TiBook. (BTW, those that told me in iChat to upgrade... you know who you are... ;) ).


What I've found:

- After upgrading to 10.2.4, users are experiencing dramatically shortened battery lives. Typical reports seem to be in the 50% less or worse. I've seen a few mentions of lives in the sub-10-minute range.

- The problem seems to very unevenly affect 'books of various vintage, all the way back to the old 'toilet seat' iBooks to the latest 12" alBook. I have a dual-usb ibook (500MHz) that is affected, but my new TiBook is fine.

- The problem seems to only happen if the incremental update was applied (10.2.3 to 10.2.4). There have been a number of reports that suggest that the combo-update to 10.2.4 is safe.

- There have been suggestions that doing the upgrade without the battery installed prevents having the problem. This is a critical point, if true, since it suggests very strongly that there is (yet again) a broken software installer in the wilds.

- The symptoms are somewhat variable, but generally follow the same sequence - discharge from 100% to 90-something percent is very rapid, occuring in less than 2 minutes, generally to 94%, for some reason. From the low 90's, consumption seems to be relatively rapid, and after a short time (between 10 and 60 minutes, usually, sometimes slightly longer), the 'book goes into hard-sleep, without warning. This is another critical point. There is no warning, suggesting that the final phase of discharge is extremely rapid. Upon connecting power, MacOSX reports the battery at 0% and begins charging. I've checked the battery's own indicator, and indeed, it has only 1 blinking light at that point.

- X-Charge (or other detailed monitoring tools) indicate a diminished capacity. This is the critical point. Reports, again vary wildly here, but in my case, a fully charged iBook battery that exhibits this problem shows a capacity of just 1.3Ah, whereas my spare battery (which seems unaffected) shows a capacity of 3.5Ah. A simple shell script was posted on macosxhints.com that allows a simple query of the battery's state, reposted here:


#!/bin/bash

[ -x /usr/sbin/ioreg ] && \
/usr/sbin/ioreg -p IODeviceTree -n "battery" -w 0 | \
sed -ne '/| *{/,/| *}/ {
s/^[ |]*//g
/^[{}]/!p
}' | \
awk '/Battery/ {
gsub("[{}()\"]","", $3)
gsub(","," ",$3)
split($3,ct," ")
# extract flag value and convert to hex
sub("Flags=","",ct[2])
str=sprintf("Flags=%d/0x%03x",ct[2],ct[2])
sub("Flags=[0-9]*",str,$3)
# get max and current charge levels
sub(".*=","",ct[4])
sub(".*=","",ct[5])
printf("%s [%.1f%%]\n",tolower($3),100*ct[5]/ct[4])
}'

# EOF


(To run this script, use your favorite editor to create a text file, paste the contents of the script into it, and save it. Be sure to make it executable (chmod 755 battery.sh for example) and run it within Terminal.)

- None of the methods I've seen posted have had a positive effect on the life of the battery, nor have I seen any "killer" solution that anyone else is having great success with - reverting to 10.2.x (pre-.4) has no effect. Nor does PMU resets, PRAM-zapping, etc.

- This problem is unrelated to the normal weakening of Lithium-Ion batteries. As noted, even very new (<6mo batteries) have been seen to have this problem.

- The problem does not seem to reoccur with a new battery or one that was not present during the update.

- 10.2.3 has/had a bug where, when the system was awakened from sleep, it would display the "running on reserve" message erroneously. This issue is noted as being fixed in the 10.2.4 release notes. Fixing this would involve an update to the PMU firmware.

- The 'books' batteries are "smart" enough to report their own capacities to the system, so that the PMU doesn't overcharge them, and so that the OS can have a reasonably good guess as to remaining life/remaining charge time.

So, in summary, here's what I think has happened. This is just conjecture based on postings all over the place, and observation of the behaviour of my own systems:

The 10.2.4 update through SU updates the PMU during the final stages of installation, where, if a battery is attached, it will have it's own internal settings corrupted, causing it to report a dramatically lower capacity. This lower capacity results in very short charge-times, and equally short discharge times. The discharge curve is so steep at the end, the system never gets a chance to warn the user, about the remaining life before power levels drop to the point that the PMU places the system in hard-sleep.

I'm pretty certain that the batteries are physically OK, but that it's just their reporting that is hosed up. That suggests that Apple may fix this through software by clearing the battery's state, forcing the PMU to reasses the actual capacity by doing a slow-charge until voltage-levels/temperatures begin to spike.

Word is that Apple, although not acknowleging the problem openly is, at least, working on the fix. I'll bump this thread in a week or so if the fix has not been released by then.

For more details (some of it unhelpful and inaccurate, unfortunately), see the threads in Macosxhints.com, Apple's discussion forums, macnn.com. I found a thread or two on Google as well..
 
I have the same thing. here - it goes to sleep at 56%. It's on a clean isntall (10 days ago) ibook 700, I wanted to get a combo update only to 10.2.3 but I didn't found so the combo update of 10.2.4 does it as well. - on the other thread there are links to AI and macnn too - it seems to be a far too common issue.. The combo update is not safe anway. And I don't want a clean install only because of that...
 
Does anybody know which chip does the battery supervision ? there should be an embedded MCU in the battery and its software has apparently been screwed up by the update.
 
I think the PMU is responsible for managing the battery's state. From what I can gather, it's exactly as you say - the upgrade hoses up the battery. But, I think it's the upgrade process itself. The second battery I had for the iBook (I'm selling the iBook) has been fine with 10.2.4 on the 'book; it was just not in the book at the time of upgrade.

... and still no word from Apple. This is getting a bit ridiculous... :rolleyes:
 
Wow, I'm so glad I'm a desktop guy...

I think it's a conspiracy on Apple's part to sell more batteries. :)
 
Thank you for this thread. I'm suffering the same problem with my 15" TiBook 1GHz bought Mid-December 2002.

I'm just trying to get a replacement battery on warranty. Does anybody of you have any news on this? Did you succeed to get your battery replaced?
 
In one Apple Store the genius suggested to install the OS X again and do the combo update only to 10.2.3 ... that did not help obviously.
Many have got the warranty battery replacement, just ask enough loudly (need to do the same) and show them you know the issue better than they do..
 
I just got an "official" info from Apple Care Europe. They say that Apple is working on the problem and they are confident that it can be solved by software. Replacement of the battery will not be required. However, it could take "a few weeks" until the fix will be available.

They said that they are aware of this problem. In the US they started to replace the batteries but not in Europe.
However, since a fix is in reach they won't continue doing that, according to what they told me.

What I will do is a) buy a 2nd battery which I intended to do anyway and b) wait for the promised fix. The 2nd battery will help me to overcome this situation in the next few weeks :)

Just to let you know.
 
Could this be why my Powerbook G3 has died again? It was doing all sorts of strange booting and rebooting nonsense, then seemed to pack up. I was thinking it was the PRAM battery, but now I've seen this, and it was just after I installed 10.2.4... Hmmm
 
Get X-Charge, then run it and let your fully-charged battery discharge until your PowerBook goes to emergency-sleep (i.e. let X-Charge monitor the discharging of your battery completely). Next, connect the power plug, wake-up the Book and let X-Charge monitor the entire charging process.

Compare your discharge/charge curves to Giaguara's shown in the other thread. If they look like this then you have the same problem as described here.
 
There's one other element of the fast-drain that I've noticed lately that's a little curious...

I use my PowerBook @ home with Airport, and occasionally @ work with ethernet. I have a second location set up to use only ethernet and a fixed address while I'm at work. It would seem that every time I come back home after having used the PB @ work, and switch to the Airport setting, X-Battery indicates my drain to be around 1400, sometimes a little higher even.

With that kind of drain, even a full charge on a healthy (4.149Ah) battery, the runtime reported by the menulet is a little over 2 hours. And this is all without doing high-drain things, like watching dvds, or burning CD's. I've gone so far as to make sure that there's not wayward process taking up inappropriate amounts of processor time, and to make sure that there aren't any processes that would be looking for the 'other' network connection hanging around. The only way I've found to correct it is a restart, after which, drain returns to normal.

Normal drain, if I haven't switched networks like that, and don't do any high-drain stuff, is around 850-900, yielding a solid 4:20-4:30 runtime (I've actually tested that).

I don't remember this kind of oddity with 10.2.3 on my iBook, which I also used in the same way between home and work. I'll continue to gather numbers and try to nail down the exact pattern that kicks it off.

Is anyone else seeing this kind of behaviour?
 
People seem to get out the batteries when they need now instead of having been told to wait for the 10.2.5 (or .6) update to possibly fix it. I will have enough impressive keynote / quiktime / whatever battery charge behaviour tables once i go to an apple store, (hopefully next week). The battery replacement seems to be mroe the policy now.
 
Yes, AppleCare meanwhile confirmed that they will send me a replacement batt for free. I called them multiple times until they offered me that. This has been just 2 days ago so I'm still awaiting it.
 
Since 10.2.5 is out now, does the battery charging / drain behaviour still occur?
I've had my battery replaced 4 months ago.
 
Once the battery woe has started, it continues. So if you weren't sooo lucky to somehow fix it with PMU resettting, it still occurs - untill you get a new battery.
 
woe is the word.

I am now down to exactly one whole existentially eternal second. Power goes out and the ibook is toast. The battery is sssooo dead that even a shutdown looses the date and time.

I've ordered a new battery, but they have to send it in from Dubai. The distributer there is all denial that it has anything to do with the software, but they did recommend a possible fix... to reset with a option-shift-power and then to start, let the battery run out and then recharge without using the computer.

Since I can't even hold a date, it obviously won't (didn't) work for me

I hope Dubai hurries up with the replacement!
 
You must log in or register to reply here.
Back
Top