PPC G5 CRASHING SOON AFTER STARTUP.

Thank you for the response. I checked all the links you supplied. Sobering reliability picture of the G5 DP 2.7. Maybe I have been lucky up till now. I have both hairdryer and heat gun. Sounds like the heat gun is the better bet for the motherboard. Can I heat the motherboard in place in the G5 or do I need to do a full disassembly to get at the back of the board where the solder joints are?

With regard to the RAM. I am in a bit of a catch-22. Since my G5 is essentially "blind" I don't know how to test the RAM with Tech Tools. Let me explain:I have the original 10.4 install disks for the G5 and can force it to start with the installer disk using "C" key, but it doesn't boot past the grey screen. Therefore I can't run AHT or Tech Tools.

A possible option is: I have a G4 PowerBook 1.6Gz with 10.4.11 with it's install disks. I can connect via firewire to the G5 with the G5 in target disk mode, I am just not sure if I can access AHT on the G4 to run remote diagnostics on the G5 in target disk mode. It seems to me that AHT must be running on the host computer via it's own native instal disks in order to access hardware. Is that the case?

I have seen in many other posts how you have helped lots of people with tricky problems. thank you so much for the service you give. I try to keep my expectations realistic as I know all my macs operate somewhere between hard science and voodoo.

I am about to do the heat gun to revive the motherboard and wanted clarification. Is this done with the board in place or removed from the computer?

Looking inside, it looks like solder joints are on the backside of the motherboard, out of site.

If done inside the machine, do I just do the lower portion around the RAM or also the upper portion around the video card and battery? (ergo: remove video card, battrry, speaker)

Better safe than sorry : )

Also: should this fail and I need to get another motherboard, what version of AHT or ASD disk is needed for this G5 Dual 2.7 with OS 10.4.11?

Much thanks.

I am totally miffed...

When I first posted this problem, I was having issues with my G5 alternatingly staring with gray screen, no screen or normal screen which would then freeze up a few minutes later in kernel panic.

To try to fix this, I ran Disk Utilities for Permissions and Structure, Zapped PRAM, reset PMU button, Checked PRAM battery, Tested RAM with TechTools Pro. Everything checked out OK.

Feedback on this forum raised the prospect of bad motherboard and possibly faulty verification of RAM by TechTools.

I removed RAM and heated motherboard mostly around RAM slots to 350˚ for 5 minutes. This was as hot a my heat gun would go.

I then used MemTest86 to check RAM. IN a process of elimination, I did one pair at a time rotating them in and out of the slots I knew were OK. All RAM cards checked out OK.

These two procedures improved the ratio of good startups but didn’t eliminate the random kernel panics. Some startups would be OK for hours, others for minutes.

After hard shutdown, first restart would produce gray screen and second would produce normal startup.

First startup today produced frozen screen and text beginning with:

“System Failure: cpu=1: code= 00000007 (corrupt skip lists).

This was followed by “Latest crash info for cpu 1” and much more text .

After a hard shutdown, got a gray screen. Next startup was normal and lasted for about 30 min. before kernel panic freeze.

So... this mystery deepens. Where do I go from here? Any ideas?

Thanks

You mentioned in an earlier post that solder melts at 415˚. So is that the target temp?

In your experience, with all the symptoms mentioned in my last post, including the ominous "System Failure: cpu-1..." text, you think that the motherboard may still be the culprit? If so I will get me a hotter heat gun and go at it again.

Thanks

With another heat gun I heated the areas around the two RAM slot groups to 450˚. It had me pretty anxious when smoke began pouring out.

When it cooled, I reloaded RAM and it booted up perfectly. I used the computer for 3 hours of work before it froze up again. Following the freeze I tried rebooting 3 times in a row, each time producing only gray screen. I let it cool for a while and it rebooted perfectly. That seems to indicate the problem is heating of hardware.

Since I really cooked the motherboard with 5 minutes at 450˚ on each RAM slot group it is hard to believe that I will get any better results by doing it again. What do you think?

At this point, kernel panics or freezes seem to only occur after the computer has run for a while (heated up?). Is there a way to zero in on which component is failing without just swapping expensive parts in a process of elimination? If I had some idea, I could determine when to cut my losses rather than rolling the dice.

Thank you ChitlinsCC for your input. Very good troubleshooting article and ideas.

To BDAqua: Yes this 2.7 has LCS. I have heard comments on other threads about the thermal paste. I will have to give that serious thought.

My G5 must be inhabited by gremlins delighting in changing things just to stay one step ahead of me. After a good restart last night, I let the computer run all night. It was OK this morning with no frozen screen and no "afterburner" fan. Tempting fate, I shut it down for a little bit and restarted to a normal screen just to see if my luck holds. So far it's been a couple of trouble free hours. What I do notice is that the fan is running somewhat faster than normal for a 75˚ room and light CPU load. It is not in full "afterburner" mode but faster than normal. I checked in Prefs> Energy Saver> Processor Performance is set to Automatic.

Pondering possibilities: If the heating of the motherboard did actually "heal" it, is it possible that the sensors now see it as a "new" motherboard and and have subsequently thrown the fan speed off? If so, I will have to track down the elusive ASD 2.5.8. to readjust sensors.

I will persist through the weekend to see if things have improved and report back.

Thanks for the help.