Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
thanks again Intell,

so it seems that there is nothing I can do about the gpu as it is soldered on.

Is it worth getting another logic board in the hope that the gpu will work better than this one?
Or is the model simply doomed to this kind of temperamental failure?

Do you think it's worth changing the ram out still, might it limit the amount of shut downs?

thanks again for all your comments
 
thanks again Intell,

so it seems that there is nothing I can do about the gpu as it is soldered on.

Is it worth getting another logic board in the hope that the gpu will work better than this one?
Or is the model simply doomed to this kind of temperamental failure?

Do you think it's worth changing the ram out still, might it limit the amount of shut downs?

thanks again for all your comments

Being a 1.25 GHz model it may be worth it if you can get the replacement parts at the right price. Don't forget to consider a replacement upgraded PowerBook as that may be the better option if you have been eyeing one recently.
 
Being a 1.25 GHz model it may be worth it if you can get the replacement parts at the right price. Don't forget to consider a replacement upgraded PowerBook as that may be the better option if you have been eyeing one recently.


thanks for the advice Altemose. Do you say that because the 1.25 was a good model?
The spares aren't that expensive as far as i can see on ebay, even here in the uk. And definitely come in less expensive than a replacement which might end up with many of these similar problems and need servicing anyway.

I have added the issue
"your computer's clock is set to a date before march 24th, 2001"
to the list - so I guess a new pram battery is also necessary.
 
thanks for the advice Altemose. Do you say that because the 1.25 was a good model?
The spares aren't that expensive as far as i can see on ebay, even here in the uk. And definitely come in less expensive than a replacement which might end up with many of these similar problems and need servicing anyway.

I have added the issue
"your computer's clock is set to a date before march 24th, 2001"
to the list - so I guess a new pram battery is also necessary.

12" PowerBook G4s and iBooks store the date and time data on a capacitor on the logic board. There is no PRAM battery at all. If the main battery dies or the machine loses the power source then the time data is lost. You could get a new logic board or upgrade to a faster model machine for not much more.
 
I just replaced the logic board on a 12" PB G4 on Sunday. It was the last revision (A1104) and a logic board for that ran to £20 delivered. Better than faffing about with shims for the dodgy gpu.

It took HOURS, though and though I treble checked every stage I still ended up with two mystery spare screws with no home to call their own.

Good thing it all worked first time. I might have killed myself if it didn't and I had to open everything up again. Make a big pot of coffee before you start.
 
12" PowerBook G4s and iBooks store the date and time data on a capacitor on the logic board. There is no PRAM battery at all. If the main battery dies or the machine loses the power source then the time data is lost. You could get a new logic board or upgrade to a faster model machine for not much more.

That makes sense as that issue only cropped up i think when I started to test out the system without the battery plugged in. I think I came across some stuff abut dodgy g4 batteries and recalls. So I guess the battery never had a chance to recover.

However this model is actually a 15" not a 12". Is that the same?

----------

I just replaced the logic board on a 12" PB G4 on Sunday. It was the last revision (A1104) and a logic board for that ran to £20 delivered. Better than faffing about with shims for the dodgy gpu.

It took HOURS, though and though I treble checked every stage I still ended up with two mystery spare screws with no home to call their own.

Good thing it all worked first time. I might have killed myself if it didn't and I had to open everything up again. Make a big pot of coffee before you start.

Im glad to hear it. It turns out the 15" 1.25 logic board isnt as easy to find as i thought.

Having got stuck into this repair stuff, i know how frustrating it is, satisfying when it works out. Thats why i've been unsure yet of whether to start splashing on new bits and pieces.

Also because ive stumbled on something that could be quite interesting i think......

----------

The kernel panics were becoming more frequent and I kept searching for issues relating to the gpu etc... one site recommended a fresh install which I thought was probably futile given everything else that I have read and all the advice ive been given.

However I did a fresh install of 10.4.1 and so far it is working perfectly.

Ive tested media files, large video files etc and opened up every stock program and it all works like lightning and of course, no panics!

When I originally managed the install I immediately, on prompt, began installing all of the updates (mostly security, java and OS) and it was during and right after these that the kernel panics began.

Is it possible that any gpu failures are brought on by the software?
Im sure that updates to the os were made for stability reasons and solved many issues but is it possible that the gpu (and maybe even the ram issues) are brought on by the changes to the software?
 
That makes sense as that issue only cropped up i think when I started to test out the system without the battery plugged in. I think I came across some stuff abut dodgy g4 batteries and recalls. So I guess the battery never had a chance to recover.

However this model is actually a 15" not a 12". Is that the same?

My mistake! The 15" and 17" models have a PRAM battery located above the optical drive. This requires the removal of the top case.
 
That makes sense as that issue only cropped up i think when I started to test out the system without the battery plugged in. I think I came across some stuff abut dodgy g4 batteries and recalls. So I guess the battery never had a chance to recover.

However this model is actually a 15" not a 12". Is that the same?

----------



Im glad to hear it. It turns out the 15" 1.25 logic board isnt as easy to find as i thought.

Having got stuck into this repair stuff, i know how frustrating it is, satisfying when it works out. Thats why i've been unsure yet of whether to start splashing on new bits and pieces.

Also because ive stumbled on something that could be quite interesting i think......

----------

The kernel panics were becoming more frequent and I kept searching for issues relating to the gpu etc... one site recommended a fresh install which I thought was probably futile given everything else that I have read and all the advice ive been given.

However I did a fresh install of 10.4.1 and so far it is working perfectly.

Ive tested media files, large video files etc and opened up every stock program and it all works like lightning and of course, no panics!

When I originally managed the install I immediately, on prompt, began installing all of the updates (mostly security, java and OS) and it was during and right after these that the kernel panics began.

Is it possible that any gpu failures are brought on by the software?
Im sure that updates to the os were made for stability reasons and solved many issues but is it possible that the gpu (and maybe even the ram issues) are brought on by the changes to the software?

Skywalker chiming in,

GPU failures and RAM failures aren't usually caused by software. Software can cause kernel panics, though. I say just go with it, and hope for the best. May the force be with you.
 
GPU failures and glitches can be caused by software. One way to directly cause GPU problems is to overclock the GPU to high. The other is to stress an already failing or overheated GPU by asking it to render something complex.
 
After a few hours of experimenting it did crash again.

Here is the log, fyi not all the crash logs in the previous round mentioned the gpu

Thu Oct 2 02:46:43 2014
panic(cpu 0 caller 0x000A8D00): Uncorrectable machine check: pc = 0000000000082380, msr = 0000000040009030, dsisr = 42000000, dar = 000000002E892000
AsyncSrc = 0000000000000000, CoreFIR = 0000000000000000
L2FIR = 0000000000000000, BusFir = 0000000000000000

Latest stack backtrace for cpu 0:
Backtrace:
0x00095698 0x00095BB0 0x0002683C 0x000A8D00 0x000A7F90 0x000ABC80
Proceeding back via exception chain:
Exception state (sv=0x24F95000)
PC=0x00082380; MSR=0x40009030; DAR=0x2E892000; DSISR=0x42000000; LR=0x0008221C; R1=0x0CE3B3D0; XCP=0x00000008 (0x200 - Machine check)
Backtrace:
0x0008221C 0x0005C8DC 0x00646F60 0x006443E8 0x006451E8 0x00108A38
0x000D5DCC 0x000FE254 0x000D6BF8 0x000D6D5C 0x0021ADA8 0x002411F4 0x00242334 0x0023E6A8
0x0021EB5C 0x0023A988 0x000FD460 0x000F1400 0x002A9BF4 0x000ABE30 0x743B266C
Kernel loadable modules in backtrace (with dependencies):
com.apple.BootCache(25.1)@0x643000
Exception state (sv=0x2E08B780)
PC=0x90019BCC; MSR=0x0200F030; DAR=0x00719080; DSISR=0x40000000; LR=0x9029A7FC; R1=0xBFFF7020; XCP=0x00000030 (0xC00 - System call)

Kernel version:
Darwin Kernel Version 8.3.0: Mon Oct 3 20:04:04 PDT 2005; root:xnu-792.6.22.obj~2/RELEASE_PPC
*********
 
Time to hunt down Apple Service Diagnostic 2.1.5. I think that is the correct one for your PB and run the VRAM and the RAM tests. If there is a hardware problem it should be detected by that. Macrumors rules means no one can link to that, you will just have to do your own searching but ASDs are not that hard to track down.

If it passes all tests then your hard drive might be starting to throw up bad blocks here and there. Under the circumstances, I would hope for that.
 
Done the ASD 2.1.5, 1 then 3 then 3 more loops on a full test. Nothing.

I'm surprised. Is that 100%?

----------

what are these bad blocks ?? Ive never heard of that....

----------

the hard drive is brand new, if that makes a difference
 
A new hard drive can still be faulty.

Sure, the only odd behaviour I can see re the hard drive is that it doesn't show up in the system profiler under serial-ATA...

Do you guys have any advice on software that tests for bad sectors or even ones that fix them? Or is this just a case of buy a new hard drive??

thanks again for all your help

----------

this install is far more stable than the last one with only two crashes occurring within the space of a couple of minutes i think. And that was 24 hours ago.

I dont know whether this is because of bad blocks being overwritten (?if that makes sense?) or because i haven't updated the software from the original on the disk.
The amount of variables to troubleshoot is crazy. But at least we appear to have got it reduced to the hard drive

----------

Do I need to spend on Disk Tools Pro/Drive Genius or the like or is there a free app that will do a better job then disk utility? As for the price I could buy another two hdd's.
 
Last edited:
Sure, the only odd behaviour I can see re the hard drive is that it doesn't show up in the system profiler under serial-ATA...

Do you guys have any advice on software that tests for bad sectors or even ones that fix them? Or is this just a case of buy a new hard drive??

thanks again for all your help

PowerBooks are equipped with an ATA controller and therefore the hard drive will be listed in "ATA" and not "SATA".
 
oops, i see, nice, that solved that one then! Just, out of habit, i always look at s-ata.

I think I'll just do another clean install and see what happens.
Should I do a high-pass erase/ partition or does that make no difference?

If anyone has any views about software to identify and repair these bad blocks that would be sweet.

Before I reinstalled I thought Id update the os update combined 10.4.11 <and leave Java> just to test it out. It has crashed for the first time in a day and a night during the install/ download I'm not sure which.
Then restarted itself mid-restart

It is still 10.4.3 (although apparently airport was updated and wanted access to the keychain)

Here is the log;....no, no panic log
The console reads:

ntupdate<188>: getnetnum: "time.apple.com" invalid host number, line ignored
:no servers can be used, exiting
ntdp_initres<197>: couldnt resolve "time.apple.com", giving up on it
 
Last edited:
If your hard drive has bad blocks then zeroing the drive will map them out, so it is in theory not a bad idea. However, opinions differ as to whether it is worth the bother since bad blocks eventually multiply leading to a crash.

As for identifying those, commercial products such as Drive Genius, Tech Tool Pro and Drive 10 will carry out a detailed surface scan. There is a freeware app that runs in Terminal you can google for, if you are happy with using the Terminal.

You can also get a quick check by opening Disk Utility, right clicking on the volume you want to check and selecting Get Info. Under the S.M.A.R.T info are a couple of checks including bad blocks. Checking this over a period of time will tell you if your hard drive is heading for trouble.
 
I have received a new hard drive but immediately the crashes persist on install.
Im sure that if i give it a few tries it will install as before but obviously im in for the same issues.

Do the ASD tests coming up clean mean that the trouble is definitely not the ram?

Is there any chance that the OS disc is corrupted in some way or does the fact that it has installed before rule that out?
I'd like to test the disk and OS on a different system to rule this out but it seems that it is powerpc specific and all my other machines are intel.
Is there a workaround for this?

It does seem a little futile now but ive spent so much time on it im almost unwilling to give up on it
 
new idea

I just posted on 68kmla about my 1.25 powerbook and my kernel panic issue.

I took the hard drive out and the board under it (not the hard drive I think its the modem) was burnt, both sides. Capacitors and resistors heated up enough to flow the solder and cross mount them on the board. The tantalum caps on one side were "legged" for lack of a better term to the pad rather than flat on it.

This mac that kept locking up seems to have been totally tied to this board. No idea how that much heat happened there but just a suggestion to look at other places than the motherboard. With that board removed the powerbook has been stable almost 24 hours now without issue.
 

Attachments

  • 15395598859_79b9dbb6df_z.jpg
    15395598859_79b9dbb6df_z.jpg
    59.1 KB · Views: 85
  • 15579548741_44f5f2e879_z.jpg
    15579548741_44f5f2e879_z.jpg
    59 KB · Views: 77
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.