10.3.6 -- what's with the battery?

I know the battery isn't really failing
Robert B. Peirce wrote on :

"Ray Curry" raycurry@redacted.invalid wrote in message news:raycurry-F647E5.21081209112004@redacted.invalid

In article bob-A59657.19415109112004@redacted.invalid, Robert Peirce bob@redacted.invalid wrote:

In article 1gn0ein.f1fgkl1sdlio2N%simonecorelli@redacted.invalid, simonecorelli@redacted.invalid (Simone Corelli) wrote:

Robert Peirce bob@redacted.invalid wrote:

This sounds like a software issue

Can you try a restart from a different drive or partition using old Mac OS, different than 10.3.6? I can't do it simply... you can?

WE HOPE TO SOLVE!!!

Simone Corelli

Unfortunately not. I upgraded the only drive I have.

I'm having the same problem after installing 10.3.6. If my machine is unplugged for more than 3 or 4 minutes and off, the battery indicates zero charge and the system won't boot. If I plug in the AC power adapter, the machine immediately boots and indicates for 3 or 4 minutes that the battery has zero charge. Then boom it goes to 100% and I can unplug and run for 4 or 5 hours. Very consistant. I suspect the update has corrupted something in the battery manager so that it thinks there's no charge until it's run on the adapter for a while. It makes no difference whether there's 100% charge or 10% or anywhere in between. The system could even be leaving the battery electronics on when it should be shutting it off. The lithium based batteries have very sophisticated controls built in. I hadn't associated with the upgrade until I saw the post having done the upgrade Friday and the weekend was the first time the book was off for any time. I had just assumed the battery was in process of dying.

"Good" to know I am not the only one this is happening to. I filed a bug report with Apple, but I could not comment on the generality of the issue.

Ray Curry replied on :

In article 4191ff3d$1_3@redacted.invalid, "Robert B. Peirce" rbp@redacted.invalid wrote:

"Ray Curry" raycurry@redacted.invalid wrote in message news:raycurry-F647E5.21081209112004@redacted.invalid

In article bob-A59657.19415109112004@redacted.invalid, Robert Peirce bob@redacted.invalid wrote:

In article 1gn0ein.f1fgkl1sdlio2N%simonecorelli@redacted.invalid, simonecorelli@redacted.invalid (Simone Corelli) wrote:

Robert Peirce bob@redacted.invalid wrote:

This sounds like a software issue

Can you try a restart from a different drive or partition using old Mac OS, different than 10.3.6? I can't do it simply... you can?

WE HOPE TO SOLVE!!!

Simone Corelli

Unfortunately not. I upgraded the only drive I have.

I'm having the same problem after installing 10.3.6. If my machine is unplugged for more than 3 or 4 minutes and off, the battery indicates zero charge and the system won't boot. If I plug in the AC power adapter, the machine immediately boots and indicates for 3 or 4 minutes that the battery has zero charge. Then boom it goes to 100% and I can unplug and run for 4 or 5 hours. Very consistant. I suspect the update has corrupted something in the battery manager so that it thinks there's no charge until it's run on the adapter for a while. It makes no difference whether there's 100% charge or 10% or anywhere in between. The system could even be leaving the battery electronics on when it should be shutting it off. The lithium based batteries have very sophisticated controls built in. I hadn't associated with the upgrade until I saw the post having done the upgrade Friday and the weekend was the first time the book was off for any time. I had just assumed the battery was in process of dying.

"Good" to know I am not the only one this is happening to. I filed a bug report with Apple, but I could not comment on the generality of the issue.

Since I noticed your post, I went ahead and tried resetting the PRAM. That seems to have fixed the problem but I've only tried once to have the machine off and unplugged. I'll try a couple more times and post the results if necessary (or significant).

Robert B. Peirce replied on :

"Ray Curry" raycurry@redacted.invalid wrote in message news:raycurry-97298D.04125210112004@redacted.invalid

In article 4191ff3d$1_3@redacted.invalid,

Since I noticed your post, I went ahead and tried resetting the PRAM. That seems to have fixed the problem but I've only tried once to have the machine off and unplugged. I'll try a couple more times and post the results if necessary (or significant).

I've tried that. It seemed to help a little bit at first (I think I got almost an hour) but not later. I don't know why.

Ray Curry replied on :

In article 4192291c$1_2@redacted.invalid, "Robert B. Peirce" rbp@redacted.invalid wrote:

"Ray Curry" raycurry@redacted.invalid wrote in message news:raycurry-97298D.04125210112004@redacted.invalid

In article 4191ff3d$1_3@redacted.invalid,

Since I noticed your post, I went ahead and tried resetting the PRAM. That seems to have fixed the problem but I've only tried once to have the machine off and unplugged. I'll try a couple more times and post the results if necessary (or significant).

I've tried that. It seemed to help a little bit at first (I think I got almost an hour) but not later. I don't know why.

I too find it was a little help but not a fix. After a couple more hours, my Titanium wouldn't boot without a couple minutes one the power adapter. Maybe the system is actually draining the battery too much. But it does seem to hit 100% charge awfully quick from being totally dead. I may try removing the battery while it's off and see what that does.

Robert Peirce replied on :

In article raycurry-271CA4.07244510112004@redacted.invalid, Ray Curry raycurry@redacted.invalid wrote:

I too find it was a little help but not a fix. After a couple more hours, my Titanium wouldn't boot without a couple minutes one the power adapter. Maybe the system is actually draining the battery too much. But it does seem to hit 100% charge awfully quick from being totally dead. I may try removing the battery while it's off and see what that does.

Interesting that we both have Titanium PowerBooks. This problem only seems to have affected a few people. I wonder if it is somehow restricted to those with Titanium PowerBooks.

Huan the hound replied on :

On 2004-11-11, Robert Peirce bob@redacted.invalid wrote in bob-2AFDC6.19292610112004@redacted.invalid:

In article raycurry-271CA4.07244510112004@redacted.invalid, Ray Curry raycurry@redacted.invalid wrote:

I too find it was a little help but not a fix. After a couple more hours, my Titanium wouldn't boot without a couple minutes one the power adapter. Maybe the system is actually draining the battery too much. But it does seem to hit 100% charge awfully quick from being totally dead. I may try removing the battery while it's off and see what that does.

Interesting that we both have Titanium PowerBooks. This problem only seems to have affected a few people. I wonder if it is somehow restricted to those with Titanium PowerBooks.

Please keep posting about this, because I too have a Titanium PowerBook.
I'm very interested to know what's causing this battery problem for you. I haven't done the update yet.

Ray Curry replied on :

In article 2vhg1tF2mn067U1@redacted.invalid, Huan the hound huanthehound@redacted.invalid wrote:

On 2004-11-11, Robert Peirce bob@redacted.invalid wrote in bob-2AFDC6.19292610112004@redacted.invalid:

In article raycurry-271CA4.07244510112004@redacted.invalid, Ray Curry raycurry@redacted.invalid wrote:

I too find it was a little help but not a fix. After a couple more hours, my Titanium wouldn't boot without a couple minutes one the power adapter. Maybe the system is actually draining the battery too much. But it does seem to hit 100% charge awfully quick from being totally dead. I may try removing the battery while it's off and see what that does.

Interesting that we both have Titanium PowerBooks. This problem only seems to have affected a few people. I wonder if it is somehow restricted to those with Titanium PowerBooks.

Please keep posting about this, because I too have a Titanium PowerBook.
I'm very interested to know what's causing this battery problem for you. I haven't done the update yet.

My Titanium seems to be getting better. Reseting the PRAM helped a little I thought by increasing the time the Powerbook could be off before the battery "locked up" which it would still do after about 3 hours. Yesturday I pulled the battery while running on the AC and then shutdown. After about 6 hours I reinserted the battery and it was okay so I'm pretty sure the problem is something the computer does when powering down or while off with the power adapter not plugged in. Then I changed my Energy Saver preferences just in case there was some sort of corruption there. Last night I powered off normally and checked this morning and the battery was still okay, first time that has happened since the upgrade. So I'll try a couple of more times to see if the problem has really gone away or not. I know the battery isn't really failing because once it wakes up, I can run 4 to 5 hours still on the battery and it only takes about 15 minutes for a "locked" battery to go from 0 capacity to 4% and then immediately goes to 100%.

Robert Peirce replied on :

In article raycurry-B382A1.08074011112004@redacted.invalid, Ray Curry raycurry@redacted.invalid wrote:

My Titanium seems to be getting better. Reseting the PRAM helped a little I thought by increasing the time the Powerbook could be off before the battery "locked up" which it would still do after about 3 hours. Yesturday I pulled the battery while running on the AC and then shutdown. After about 6 hours I reinserted the battery and it was okay so I'm pretty sure the problem is something the computer does when powering down or while off with the power adapter not plugged in. Then I changed my Energy Saver preferences just in case there was some sort of corruption there. Last night I powered off normally and checked this morning and the battery was still okay, first time that has happened since the upgrade. So I'll try a couple of more times to see if the problem has really gone away or not. I know the battery isn't really failing because once it wakes up, I can run 4 to 5 hours still on the battery and it only takes about 15 minutes for a "locked" battery to go from 0 capacity to 4% and then immediately goes to 100%.

This is very useful info. I will have to give it a try. Thanks.

Tom Harrington replied on :

In article bob-C734D2.21062411112004@redacted.invalid, Robert Peirce bob@redacted.invalid wrote:

In article raycurry-B382A1.08074011112004@redacted.invalid, Ray Curry raycurry@redacted.invalid wrote:

My Titanium seems to be getting better. Reseting the PRAM helped a little I thought by increasing the time the Powerbook could be off before the battery "locked up" which it would still do after about 3 hours. Yesturday I pulled the battery while running on the AC and then shutdown. After about 6 hours I reinserted the battery and it was okay so I'm pretty sure the problem is something the computer does when powering down or while off with the power adapter not plugged in. Then I changed my Energy Saver preferences just in case there was some sort of corruption there. Last night I powered off normally and checked this morning and the battery was still okay, first time that has happened since the upgrade. So I'll try a couple of more times to see if the problem has really gone away or not. I know the battery isn't really failing because once it wakes up, I can run 4 to 5 hours still on the battery and it only takes about 15 minutes for a "locked" battery to go from 0 capacity to 4% and then immediately goes to 100%.

This is very useful info. I will have to give it a try. Thanks.

Did you try resetting the power management unit on the Powerbook?
Someone else suggested this but I don't think you ever responded one way or the other. I'd guess that the scheme described here might have the same effect, although it would take much longer.

Ray Curry replied on :

In article tph-38E902.22373511112004@redacted.invalid, Tom Harrington tph@redacted.invalid wrote:

In article bob-C734D2.21062411112004@redacted.invalid, Robert Peirce bob@redacted.invalid wrote:

In article raycurry-B382A1.08074011112004@redacted.invalid, Ray Curry raycurry@redacted.invalid wrote:

My Titanium seems to be getting better. Reseting the PRAM helped a little I thought by increasing the time the Powerbook could be off before the battery "locked up" which it would still do after about 3 hours. Yesturday I pulled the battery while running on the AC and then shutdown. After about 6 hours I reinserted the battery and it was okay so I'm pretty sure the problem is something the computer does when powering down or while off with the power adapter not plugged in. Then I changed my Energy Saver preferences just in case there was some sort of corruption there. Last night I powered off normally and checked this morning and the battery was still okay, first time that has happened since the upgrade. So I'll try a couple of more times to see if the problem has really gone away or not. I know the battery isn't really failing because once it wakes up, I can run 4 to 5 hours still on the battery and it only takes about 15 minutes for a "locked" battery to go from 0 capacity to 4% and then immediately goes to 100%.

This is very useful info. I will have to give it a try. Thanks.

Did you try resetting the power management unit on the Powerbook?
Someone else suggested this but I don't think you ever responded one way or the other. I'd guess that the scheme described here might have the same effect, although it would take much longer.

I did not try resetting the power management though I intended to tonight. I didn't previously because the symptoms shown in the note on reseting didn't match anything I had seen but it's worthwhile to try. I had one more battery lock last night after about 10 hours with the computer off and unplugged from the AC. Again 5 minutes of running on the AC fixed it back to 100% charge indicated. I'm in the midst of having the battery out of the system for a similar amount of time to see if the lock only occurs with the battery in the system. After that I'll try resetting the power management.

Robert Peirce replied on :

In article 4559e820.0411121459.4da5f226@redacted.invalid, sdfisher@redacted.invalid (Steven Fisher) wrote:

The wrinkle is that I have two batteries. I tried swapping them yesterday; it didn't make a difference.

That is critically useful information. I have the same problem with my PB and have been puzzling over whether it was software or hardware and no way to find out. Looks like it is definitely the 10.3.6 upgrade.
The related problem is why some and not all machines. To the best o my knowledge, mine is totally stock, yet other stock machines are not showing any problems.

Renzland replied on :

On 2004-11-13, Robert Peirce bob@redacted.invalid wrote:

In article 4559e820.0411121459.4da5f226@redacted.invalid, sdfisher@redacted.invalid (Steven Fisher) wrote:

The wrinkle is that I have two batteries. I tried swapping them yesterday; it didn't make a difference.

That is critically useful information. I have the same problem with my PB and have been puzzling over whether it was software or hardware and no way to find out. Looks like it is definitely the 10.3.6 upgrade.
The related problem is why some and not all machines. To the best o my knowledge, mine is totally stock, yet other stock machines are not showing any problems.

In all this discussion, none of those who are experiencing this problem has ever stated what the Capacity of their battery is reported as, and if/how it is affected by recalibration.

Did anyone run the program I offered? bat.sh

-- Peter

Robert Peirce replied on :

In article DTsld.64560$km5.2603452@redacted.invalid, Renzland phr040919@redacted.invalid wrote:

In all this discussion, none of those who are experiencing this problem has ever stated what the Capacity of their battery is reported as, and if/how it is affected by recalibration.

Did anyone run the program I offered? bat.sh

I missed this. If you will repost I will try it.

Renzland replied on :

On 2004-11-13, Robert Peirce bob@redacted.invalid wrote:

In article DTsld.64560$km5.2603452@redacted.invalid, Renzland phr040919@redacted.invalid wrote:

In all this discussion, none of those who are experiencing this problem has ever stated what the Capacity of their battery is reported as, and if/how it is affected by recalibration.

Did anyone run the program I offered? bat.sh

I missed this. If you will repost I will try it.

O.K. Here it is:

Subject: Battery Management Software issues Date: Wed, 10 Nov 2004 18:09:41 -0500

ioreg no longer returns a useful value for "Amperage" in all cases. (This started happening long before 10.3.6)

It may return something like 18446744073709551208 when plugged in. While it is debatable what should be returned, that is not one of the more reasonable answers. :-)

I can think of --

  • what the computer is consuming
  • what the battery is soaking up (when charging)

Right now my battery is charging, and while ioreg returns that highish value above, Xbattery displays "Amperage: 360"

That definitely is the charging current. I wonder how it gets it?

Perhaps it guesses it by integrating "charge" over time

In 1 minute, battery charge content ("Current") rises by 6mAh, (from 992mAh to 998mAh) so that amounts to 360mA.

Anyone here know?

If anyone is interested --

/usr/sbin/ioreg -p IODeviceTree -n battery -w 0 | grep IOBatteryInfo

gives you the battery status.

I think that the only reliable part is the Voltage. The rest is guesswork, based mostly on the value of "Capacity" which is what calibration is supposed to estimate.

The following shell script may be useful in monitoring what's happening with your battery during calibration:

bat.sh ## contents may settle during shipment ##

#! /bin/sh

bat.sh ## monitor battery while discharging or charging ## Peter Renzland

MINUTES=0 while : do echo -n " $MINUTES " set -- /usr/sbin/ioreg -p IODeviceTree -n battery -w 0 \ | grep IOBatteryInfo | tr '[\"\=\(\{\)\}\,]' ' '

shift 4
while :
do
echo -n "$1: $2  "
shift 2
case $# in 0) break;; esac
done
date  "+ %H:%M:%S"

sleep 60 MINUTES=expr $MINUTES + 1 done

-- Peter

Ray Curry replied on :

In article sHuld.50314$Z7.1503872@redacted.invalid, Renzland phr040919@redacted.invalid wrote:

#! /bin/sh

bat.sh ## monitor battery while discharging or charging ## Peter Renzland

MINUTES=0 while : do echo -n " $MINUTES " set -- /usr/sbin/ioreg -p IODeviceTree -n battery -w 0 \ | grep IOBatteryInfo | tr '[\"\=\(\{\)\}\,]' ' '

shift 4 while : do echo -n "$1: $2 " shift 2 case $# in 0) break;; esac done date "+ %H:%M:%S" sleep 60 MINUTES=expr $MINUTES + 1 done

The current does read abnormal values when the battery is in the falted state. In the past, I've seen these symptoms when a process reads a signed value with an unsigned operation. But why some machines and not others, I could believe that a single model like the Titaniums could be affected but only some are.

What I do know is:

  1. It only happens to a battery when in the computer when off and no AC adapter. It never has happened when the computer is running on the battery or when charging from a discharge state or running under AC with full charge. It doesn't happen when the computer is off and the AC adapter is plugged in.
  2. Reseting the power management didn't change it.
  3. It takes 8 hours or more in the off state before the battery "develops the error condition".
  4. It only takes about 10 to 15 minutes of running on AC for the battery to recover and show it's true capacity.
  5. The battery indicates a single blinking led during the lock phase.
  6. The battery has full capacity, I can go 4 to 6 hours on the battery.
  7. Powering down while the battery is at less that full charge doesn't change the outcome. The battery locked with 100% charge and with 20% charge.
  8. It happens with the original battery and with a new one.
  9. When I checked this morning, the battery in the computer was locked. I boot on AC and saw zero capacity. I swapped the batteries and the other battery that had been out of the computer was okay. I reswapped and the first battery still indicated zero. In 5 minutes, that battery had recovered and showed 100%.

I do remember the patch that was added to make sure the battery was fully charged because some users weren't getting full battery life. Maybe individual sensors can cause give readings that overflow a sign or something. Without knowing the internal electronics of the battery and where the actual sensors are, I can only guess. Maybe I'll set up a tracking for the IOBatteryInfo and see how the values vary in various states before and after the lock occurs but for now I'll just have to adjust.

Peter Renzland replied on :

On 2004-11-14, Ray Curry raycurry@redacted.invalid wrote:

In article sHuld.50314$Z7.1503872@redacted.invalid, Renzland phr040919@redacted.invalid wrote:

#! /bin/sh

I made a new improved version ... enclosed below ...

  1. It only happens to a battery when in the computer when off and no AC adapter. It never has happened when the computer is running on the battery or when charging from a discharge state or running under AC with full charge. It doesn't happen when the computer is off and the AC adapter is plugged in.

Not sure what "it" is. :-)

  1. It takes 8 hours or more in the off state before the battery "develops the error condition".
  1. It only takes about 10 to 15 minutes of running on AC for the battery to recover and show it's true capacity.

Oh? What does it show before? (This is the first time anyone with the problem has mentioned "capacity" (albeit without giving values)).

  1. The battery indicates a single blinking led during the lock phase.
  2. The battery has full capacity, I can go 4 to 6 hours on the battery.

I'm having difficulty understanding some of these words.

Keep in mind that "the battery has full capacity" is not precise enough be very useful. What is the capacity as returned by ioreg (or some other means, e.g.: X-Battery)? How does that change? (Others have said it drops to 0 after a few minutes)

  1. Powering down while the battery is at less that full charge doesn't change the outcome. The battery locked with 100% charge and with 20% charge.

Not sure what "locked" means.

  1. It happens with the original battery and with a new one.
  2. When I checked this morning, the battery in the computer was locked. I boot on AC and saw zero capacity. I swapped the batteries and the other battery that had been out of the computer was okay. I reswapped and the first battery still indicated zero. In 5 minutes, that battery had recovered and showed 100%.

Here is where you might be able to say something more precise by posting bat.sh logs. I recommend runnung the one enclosed with 5 second interval, i.e.

bat.sh 5

I do remember the patch that was added to make sure the battery was fully charged because some users weren't getting full battery life. Maybe individual sensors can cause give readings that overflow a sign or something. Without knowing the internal electronics of the battery and where the actual sensors are, I can only guess. Maybe I'll set up a tracking for the IOBatteryInfo and see how the values vary in various states before and after the lock occurs but for now I'll just have to adjust.

Excellent idea!

That is the reason I wrote the following:

#! /bin/sh

bat.sh [SECWAIT] # monitor battery while [dis]charging # Peter Renzland 041114

SECWAIT=${1-60} # Wait this long between checks SEQNUM=0 # How many measurements have we made

while : do case $SEQNUM in ?) echo -n " $SEQNUM " ;; ??) echo -n " $SEQNUM " ;; *) echo -n "$SEQNUM " ;; esac set -- /usr/sbin/ioreg -p IODeviceTree -n battery -w 0 \ | grep IOBatteryInfo | tr '[\"\=\(\{\)\}\,]' ' ' # X X X X Capacity 2405 Amperage 1108 Current 2390 Voltage 12109 Flags 4 # N.B. ioreg may return "18446744073709550464" for "Amperage" when plugged in

shift 4
echo -n "$6 / $2 mAh  "
shift 2
echo -n "$6 mV  "
echo -n "$2 mA  "
echo -n "Flags: $8 "

date "+ %H:%M:%S" sleep $SECWAIT SEQNUM=expr $SEQNUM + 1

done exit

-- Peter

Robert Peirce replied on :

In article 26Old.5564$Ho4.185987@redacted.invalid, Peter Renzland phr040919@redacted.invalid wrote:

#! /bin/sh

bat.sh [SECWAIT] # monitor battery while [dis]charging # Peter Renzland

041114

SECWAIT=${1-60} # Wait this long between checks SEQNUM=0 # How many measurements have we made

while : do case $SEQNUM in ?) echo -n " $SEQNUM " ;; ??) echo -n " $SEQNUM " ;; *) echo -n "$SEQNUM " ;; esac set -- /usr/sbin/ioreg -p IODeviceTree -n battery -w 0 \ | grep IOBatteryInfo | tr '[\"\=\(\{\)\}\,]' ' '

X X X X Capacity 2405 Amperage 1108 Current 2390 Voltage 12109 Flags 4 N.B. ioreg may return "18446744073709550464" for "Amperage" when plugged

in

shift 4 echo -n "$6 / $2 mAh " shift 2 echo -n "$6 mV " echo -n "$2 mA " echo -n "Flags: $8 "

date "+ %H:%M:%S" sleep $SECWAIT SEQNUM=expr $SEQNUM + 1 done exit

I am running this right now. I will post the results once the battery drops to 0%.