Looks like no one’s replied in a while. To start the conversation again, simply ask a new question.

G5 problems

Hi there,


I have a G5 Power mac 2.3 Ghz with 5.5 GB of RAM running Mac OS X 10.4.11.


This machine has been my workhorse since I bought it in 2005 however in the last 6 months it has begun to develop some problems.


The first symptoms were hearing it sounding like a jet engine taking off. It would always happen in the early hours of the morning so It would have been asleep for a couple of hours. This noise would wake me up and I would just power it off and go back to bed.


Then this would happen in the day. The machine or rather displays would always have just gone to sleep (I always set energy saver to put displays to sleep but never the machine itself) and the jet engines would slowly start up and would reach such a speed that you'd swear it was about to take off.


After a few weeks of this I decided to run disk utility from using my installer disk. This returned an error with one of my RAM chips.

In deciding what to do next I had stumbled upon a thread where someone mentioned gently rubbing the pins of each RAM chip with an eraser as this could remove some offending matter and put me back on track. As I thought that I had nothing to loose I did just that, replaced them and then re-ran disk utility.

It returned no errors. Happy days I thought and we were back up and running. Everything has been fine for about 3 months until last week when things got bad.


I had just opened up a large photo document when I got the dreaded grey curtains and a message saying that I needed to re-start my computer. This I did and it promptly happened again within moments of re-starting. Another message to re-start. Again grey curtains moments after re-start. I ended up in this circle for about 5 or 6 attempts before I quit trying. I somehow managed (I can't remember how) to get my installer disk in and was able to restart from the disk and run disk utility. The verification process failed and gave me the following message.


Disk Utility stopped repairing "Macintosh HD"

because the following error was encountered


The underlying task reported failure on exit


And in the progress window of Disk utility it said- and this is from the checking Extents Overflow file line


Checking Extents overview file

Checking Catalog file

Keys out of order

Rebuilding catalogue B tree

The volume Macintosh could not be repaired


Error: The underlying task reported failure on exit


1 HFS volume checked

1 volume could not be repaired because of an error

Repair attempted on 1 volume

1 volume could not be repaired



So after a good few hours trawling forums and google it seemed that the keys out of order message pointed to some kind of bad data of my hard disk and my only two options were to buy something like Disk warrior that would possibly be able to get rid of the problem or otherwise a clean install of the operating system would most likely remove the problem.


I decided against the DW option and do a clean install. Also I thought that a good spring clean would be a good idea in general as I had not done so in a while.


This I did and after zeroing out my HD I now have everything back up and running. Well at least I thought I did.


My new problem is that whenever the displays go to sleep I get a stuck cursor/pointer. This would mean that I need to force re-start the machine. Sometimes the cursor gets stuck after the re-start. and I've even had to re-start 2 or 3 times to get thing going again. I've also encountered my menu bar flickering on and off (never seen that before). Also I have no start up chime.


So with all this new weirdness going on I wondered if this might be the RAM issue again. Maybe it never got fixed before and was just missed in the subsequent hard ware tests that I have performed.


I did some google searches looking for ways to test the RAM and found two bits of software. Only one was compatible with OS X 10.4 and that is memtest

First off I ran the test in Full os mode - this means normal user mode. The results came back as being all ok.


The read me recommends running memtest, however, in single user mode as all memory will be tested as none of it (or at least a very small amount) will be in use.


This is where I am now stuck because within 20 seconds of being in SU mode the old jet engines fire up and it's 10 seconds to take off again before I hit the power off button.

I've tried 10 times now and it's the same each time.


I do normally have lots of external hard drives connected as well as 2 apple monitors and the usual firewire and usb hub. I have disconnected everything and just have my screens connected.


Anyone got any ideas what I can do next as I am totally stuck at the moment?


Thanks

PowerMac, Mac OS X (10.4.11)

Posted on Mar 11, 2012 5:15 PM

Reply
11 replies

Mar 13, 2012 1:27 PM in response to BDAqua

Hi BDAqua,


I've managed to install istat menus - istat pro needs OS X 10.5.


I've had it running for a couple of days and the only thing that stands out in the heat department is the U3 heatsink. This runs at around 72 - 75. The 2 other closest temps are CPU A and B that are running at 55 - 60.


Other than that the only thing that stands out is the fan speed of the drive bay which is at 1000rpm. CPU A int, CPU A ex, CPU B int and CPU B ex are running at 300rpm. Slot runs at 76rpm and Backside runs at 51rpm. The machine is totally idle whilst showing these results and the only process power is being used writing this e-mail and having Safari open.


I've only had to force restart once. The machine had been running without a hitch for about 12 hours which is better than it's been for a while but the blazing fans woke me at around 4 this morning. It took a couple of goes to restart successfully again and it's been running ok all day today with only light use.


The reduction in force restarts could be down to the fact that I have most of my external drives disconnected, just using one or two at a time. I'm also working with fewer applications open and so as a result using less process power.


I've also tried to run the full RAM test using single user mode and the same thing keeps happening. Takes a about a minute now for the fans to start up and before long they are at FULL speed again.


Any clues here as to what I can try next.


Thanks

Mar 13, 2012 2:45 PM in response to del72

I've also tried to run the full RAM test using single user mode and the same thing keeps happening. Takes a about a minute now for the fans to start up and before long they are at FULL speed again.

That would be considered as normal as long as the processes (memtest) continues normally..

Energy control kexts are not in effect, and the "full blast" fans are there to protect your system while in an "unprotected" mode.

Mar 13, 2012 2:55 PM in response to BDAqua

Hi BD,


So what is the problem that the fan speeds and U3 temp points too?


I did blast all the dust out during initial crash last week. So that has not solved anything.


I've read up about the hairdryer thing and to be honest that seems a bit scary and also doesn't seem to be a permanent fix.

Although I guess it would tell me what the problem is.

Is my machine a lead free solder version?


Thanks again

Mar 13, 2012 4:19 PM in response to BDAqua

So I've just run Apple Hardware Test and the logic board passed and everything else had passed up until the memory test. The test got suck with fans blazing during the memory test.


When I began the 'quick' test I noticed that it already said in the window that a fault had been detected with the memory.


Would the solder issue have been picked up during the test on the logic board?


Does the fail on the memory suggest a fault with the actual RAM or could it suggest a fault with where it is connected to the logic board. i.e technically it is a fault with the logicboard.


When I think about it the machine has mostly gone into panic mode after a big work session. Generally a few hours after I have stopped working. Would it be true to assume that the machine would have had time to cool down? Enough for whatever little bit of solder to contract and then loose contact/connection and cause panic/lockdown?


If it is faulty RAM I guess I could test using the halves technique and run AHT each time to determine which pair were the faulty chips?

Or is there a better way to do this?


To sum up.


Do you think it's a problem with the logicboard?

A problem with the RAM?

Or is it something else?


Many thanks again

Mar 13, 2012 4:48 PM in response to del72

Well, the logic board test doesn't really test the logic board physically, just that certain signals work, the Memory is the problem, but it is of course connected through the logic board, when the test tries to test memory, it can't really test whether the RAM is not listening/talking, or the logic board isn't allowing it to listen/talk.


I'd say the problem is the solder in the logic board near the Memory slots.

G5 problems

Welcome to Apple Support Community
A forum where Apple customers help each other with their products. Get started with your Apple ID.