Beige G3 Crashing...

[Ryoohki]

Ok I have this Beige and its crashing in OS 9.2.2 whenever I use the Network browser without having ethernet cord plugged in.

This one sure way to get it to crash but there are other ways but whenever it crashes I have to unplug the computer 8-o because the power button will not respond.

The beige is not overlocked. According to the ROM revision number it is a Rev. C. I have a very detailed list of the hardware but I won't post it unless it is needed.

I have reinstalled OS 9 many times and the problem still persists after reinstallation. However I have had OS X on the computer and that didn't crash like this (It was 10.2 so it crashed in different ways xD) so now you ask why I don't just use OS X well this Beige is affected by the OS X video card glitch so I am limited to 10.2.5 which is ok but i prefer 10.3.

Any assistance is greatly appreciated.

 

[equill]

Some extra data could be useful. DT or MT or AIO? How much RAM? PC100, or PC133 back-compatible? If a DT, low-profile DIMMs? Identical DIMM cards, or mixed? Why do you remove the EN cable? OS 9.2.1/9.2.2 installed with machine-specific CD and downloaded 9.2.2, or retail CD, or machine-specific CD from another Mac? On-board EN or an expansion card? How is TCP/IP configured? What PCI cards are installed? Does AppleTalk 'see' your EN card (on-board or add-in)?

de

 

[Ryoohki]

Its a MT

There are 3 Dims

64 MB not specified

64 MB not specified

128 MB PC-133 <--- think it was freezing before I added this stick

Tested memory for 3 hours with Gauge Pro no problems found

I have a retail copy OS 9.2.1 and I also have the 9.2.2 update

TCP/IP is configured to connect via Ethernet and configure using DHCP

I unplug the ethernet cable because i adopt old computers and i only have so many plugs on my switch as well as so many ethernet cables. I also swap it with the ps2 and i use the cable to hook up friends computers, the Nic is on the motherboard (built in)

vvv PCI card info is in the data below but in short USB and Firewire vvvv

Basic Hardware Info

Macintosh Model:

Power Macintosh G3 series/Macintosh Server G3 (ID = 510)

Personality Card:

Audio/Video card

Processor:

PowerPC G3 at 300 MHz

Co-Processor:

PowerPC Built-in

Physical Ram Size:

256 MB (262144K)

Logical Ram Size:

256 MB (262144K)

Keyboard Type:

Apple Extended Keyboard

Floppy Information:

Drive ID: 1

Device Type: 1.4M FDHD drive

Comprehensive Hardware Info

Macintosh Model:

Power Macintosh G3 series/Macintosh Server G3 (ID = 510)

Personality Card:

Audio/Video card

Processor:

PowerPC G3 at 300 MHz

Co-Processor:

PowerPC Built-in

Memory Manager:

PowerPC Built-in

Physical Ram Size:

256 MB (262144K)

Logical Ram Size:

256 MB (262144K)

Power On Hours:

70

Production Date:

Friday, June 13, 2008

Hardware Attributes:

VIA 1

SCC

Soft Power Off

SCSI 96_1

Universal ROM

Enhanced LocalTalk

PowerPC backside L2 cache (1024K) present

Keyboard Type:

Apple Extended Keyboard

ROM Size:

3 MB (emulation only)

ROM Version:

$77D

ROM SubVersion:

$45F2

ROM Checksum:

$78F57389

PCI Slot Information:

PCI Slot Name : F1

Card Name: ATY,mach64_3DUPro

Device Type: display

Board Model: ATY,GT-C

Board Version: 92

Board Vendor ID: 1002

PCI Slot Name : F1

Card Name: VDig

Device Type: VDig

PCI Slot Name : F1

Card Name: Codec

Device Type: Codec

PCI Slot Name : A1

Card Name: pci1045,c861

Device Type: usb

Board Version: 16

Board Vendor ID: 1045

PCI Slot Name : C1

Card Name: pci11c1,5811

Device Type: firewire

Board Version: 4

Board Vendor ID: 11C1

Mass Storage Devices Information:

SCSI ID: 7

Device Type: Macintosh CPU

Internal ATA:

ATA Bus: 0, device ID: 0

Device Type: Hard Drive

Name: Macintosh HD

Capacity: 18.6 GB (20020395520 bytes)

Blocksize: 512 bytes

Vendor ID: WDC WD200EB-00BHF0

Revision: 15.15M15

Serial Number: WD-WMA6K4656133

Internal ATA:

ATA Bus: 1, device ID: 0

Device Type: ATAPI CD-ROM Drive

Name: MacTest Pro (G3) ACRC 7.4

Capacity: 57.9 MB (60729344 bytes)

Blocksize: 2K (2048 bytes)

Vendor ID: SONY

Product ID: CD-ROM CDU611-25

Revision: 2.2c

Serial Number: Information not available

Internal ATA:

ATA Bus: 1, device ID: 1

Device Type: ATAPI ZIP Drive

Vendor ID: IOMEGA

Product ID: ZIP 100

Revision: 25.D

Serial Number: Information not available

Floppy Information:

Drive ID: 1

Device Type: 1.4M FDHD drive

Display Monitor Information:

Screen Size (pixels): 1152 x 870

Grays/Colors: Millions

Resolution (dpi): 72 x 72

External Monitor connected to built-in-video (Main screen)

 

[tomlee59]

I don't know if Rev. C ROMs fixed this problem, but ISTR that if the system resides in a partition larger than 8GB, much flakiness can occur if any system resources are stored beyond the 8GB limit. Perhaps this is irrelevant (and maybe even wrong altogether; I'd always partitioned my G3 drives to avoid even the possibility of such problems, so I'm relying on what I'd heard from others), but I'm passing it on FWIW.

 

[trag]

My WAGuess would be that one of your memory DIMMs is defective and the fault resides at an address which is always occupied by the operating system. GaugePro and its ilk cannot test portions of the RAM which are occupied by the OS.

You can try a couple of things. You could try removing one DIMM at a time and test whether the problem persists. You could also try moving the DIMMs around and retesting. However, keep in mind that you are probably only fully testing the DIMM which is in the middle. That is, the DIMM in slot 2 is fully tested but the ones in 1 and 3 are not.

The reason why I suspect a bad DIMM is the specificity of your failure. It could be that a bad location in RAM is used whenever you activate that particular feature.

 

[Ryoohki]

I don't know if Rev. C ROMs fixed this problem, but ISTR that if the system resides in a partition larger than 8GB, much flakiness can occur if any system resources are stored beyond the 8GB limit. Perhaps this is irrelevant (and maybe even wrong altogether; I'd always partitioned my G3 drives to avoid even the possibility of such problems, so I'm relying on what I'd heard from others), but I'm passing it on FWIW.

thanks for the suggestion and its defiantly worth trying but I think I will see if there is something else I can do first.

The HD is sort of a recent change however, i replaced it because the Quantum fireball was doing some really odd things and having data corruption issues. Come to think of it i remember this problem existing before the drive change because i remember having the drive become unreadable after an improper shut down. Oh ya thats when i was trying the mail application that the computer came with. was working fine and then freeze. The funny thing about that is that the crash created a 50 GB file on an 8 GB drive :lol: :lol: :lol:

 

[Quadraman]

You need to keep your Operating System files confined to the first 8gb of hard drive space or bad things happen. LEM recommends creating a 7gb partition and putting all your OS files on that and using any other partitions for your data and programs.

 

[Ryoohki]

You can try a couple of things. You could try removing one DIMM at a time and test whether the problem persists. You could also try moving the DIMMs around and retesting. However, keep in mind that you are probably only fully testing the DIMM which is in the middle. That is, the DIMM in slot 2 is fully tested but the ones in 1 and 3 are not.

Well i think we can count out ram as I just tried each stick one at a time and i still encountered the freezes with each one of the sticks.

What might me helpful is another beige owner opened network browser without the network cable plugged in to make sure that this isn't a flaw but a problem

 

[Ryoohki]

You need to keep your Operating System files confined to the first 8gb of hard drive space or bad things happen. LEM recommends creating a 7gb partition and putting all your OS files on that and using any other partitions for your data and programs.

while i take an hour to back up my files would you care to elaborate on "bad things" :lol:

Well looks like my usb hard drive will also crash it

AWWW crap where is my usb drive its not mounting in OS X now.....

 

[tomlee59]

The "bad things" or "flakiness" referred to include crashing, instability, failure to boot, etc.

The LEM article that Quadraman refers to is consistent with what I'd heard: Install the OS in a sub-8GB partition, or else you will be sorry. Initially all would be well if you installed the OS in a larger partition, but as soon as any important system resource (e.g., extension, etc.) got stored beyond that 8GB boundary, the bad, flaky things would come out to bite you. The precise nature of the badness would depend on what got written beyond the boundary.

 

[Ryoohki]

k the beige is reformatted however it just crashed and froze again... this time it was during shutdown the finder "unexpectedly quit" during the shutdown proses and i told it to shut down and complete freeze. Now who wasn't expecting the finder to quit i have no idea because i was expecting it this is one of the usual crashes i run into and that is a crash at shutdown however it doesn't happen all the time. I really need to get this thing fixed

:-/

 

[equill]

I am puzzled by the references to the <8GB spectre. Ryoohki is writing about OS 9.2.2, having suspended his use of the formerly installed 10.2.5. I'll not deny that the limit could exist for OS 9, but I have never met it in my Beige G3s running as G4s with OS 9.2.2, nor heard of it before. Given the presence of a rev. C ROM ($77D.45F2 in ASP), the behaviour strikes me as more like the result of a shot SDRAM DIMM, shot RAM slot, shot PMMU, or random AOG rather than a partition-size constraint. Easily tested, however, by partitioning the WD into 7.8GB (genuine) and The Rest 10.8GB and installing OS 9.2.2 anew (rather than from a copy or a clone of the present installation) in both partitions.

de

 

[tomlee59]

I happily defer to equill's data over hearsay. I've never directly tested the assertions from the LEM profile, so I have no direct experimental data to support or refute it. I've always just partitioned as if the assertions were true (couldn't hurt, might help).

I agree with equill that the symptoms are more consistent with some other problem. As usual, he sayeth much sooth.

 

[Ryoohki]

while reinstalling the ram sticks i noticed one of them sounded like crinkling cellophane however that stick appeared to work normally (if you can call it normal) with the computer running with just that one stick. But i think i might try removing that stick for a little while.

i should also try removing the pci cards but one thing at a time.

Btw how would I test the PMMU

 

[equill]

With extreme difficulty. It's one of those things that gets lumped in with 'MLB failure' as the only remaining (but unspecific) cause of malfunction. When you have eliminated the obvious, as Holmes was wont to say, whatever remains, however improbable, must be the cause. You are on the right course by testing one suspect at a time, although use of a known-good MLB would save a lot of time. Keep in mind that not only removal of a part but also substitution with a known-good part can be revealing. However, it is of some comfort if you get a POST chime at startup.

de

 

[Ryoohki]

Turns out that the Network browser crash is a bug not a flaw i got the same freeze from the iMac DV SE so i am guessing that the rest of the crashes were similar bugs in the OS 9 software.

Probably means i should start looking for a new video card so I can run OS X.

 

[Bunsen]

Not only does OS X need to be in a partition smaller than 8 GB, it must be the first partition on the drive. And said drive must be set as master, not slave or cable select.

It's true that this does not apply to OS 9 and below.

 

[madmax_2069]

Not only does OS X need to be in a partition smaller than 8 GB, it must be the first partition on the drive. And said drive must be set as master, not slave or cable select.
It's true that this does not apply to OS 9 and below.

i ran tiger on my unpartitioned 120gb HDD with the help of xpostfacto for a long time without issue. xpostfacto removes the 8gb limit if the drives larger then 8gb. if you was to try to install OS X (testing using 10.2) on the 120gb hdd without xpostfacto it will not allow you to select where to install to. with xpostfacto (version for tiger) you can install on the 120gb HDD.

like he said above the limit isnt there for Mac OS, i really think it was a software implemented attempt from Apple to persuade people from installing OS X onto a Beige (old world Mac) even tho 10.2 supported the Beige. if OS X would not install on the beige that isnt in stock config (HDD bigger then 8gb) and the person didnt know about the limit then they will either A get a new Mac or stick with the old Beige and Mac OS. remember that when OS X is running it over rides the stock rom in the Beige with rom in ram so it dont matter what rom version you have.

also like he said it has to be set to master (sometimes single master mode if its a option) i found that out the hard way with my first attempt to install 10.2 on my Beige the first time. my HDD was set to master but i kept getting cant open across the screen in the openfirmware screen when it would try to boot up. then someone helped me to diagnose the issue and asked me if the HDD was put into single master mode and that if it wasnt to try it in that mode. when i did 10.2 booted up and ran fine untill the 10.2.8 update.

but my 120gb HDD could be set to single master or master with slave present and 10.2 didn't have a issue booting.

but my Beige G3 in OS 9 had stability issues as well but in OS X it ran like a champ with no issues (sept for the monitor going to sleep and not waking up after the 10.2.8 update, which there is a work around for on Apples site) i have the AIO model, the issue with the monitor going to sleep and not wake up is also there in 10.3 and 10.4 but the work around works for those as well.

 

[Ryoohki]

but my Beige G3 in OS 9 had stability issues as well but in OS X it ran like a champ with no issues (sept for the monitor going to sleep and not waking up after the 10.2.8 update, which there is a work around for on Apples site) i have the AIO model, the issue with the monitor going to sleep and not wake up is also there in 10.3 and 10.4 but the work around works for those as well.

Yep same issue here however i haven't had any luck with the work arounds how did you get it working? Only solution i have heard of is a mac compatible PCI video card. If x post facto removes the 8 gb limit that explains why i never had stability issues with OS X besides the normal instability, i had to use it because I have a dvd so i instal from an external firewire drive. Can't believe i might finally get to the bottom of the video issue [] ]'> unless the work around is a new video card in which case i need to find one first

*Flips through Apple's site* Link might be helpful [] ]'> *Frantic searching*

---------- few minutes later----------

Could this be it?

http://docs.info.apple.com/article.html?artnum=25564

Sort of odd though says more than 192 mb ram but when i was messing round with osx before i had 128 mb... mabey just the one stick instead of the 2 64 mb would help

i was reading through this post on the apple message board:

http://discussions.apple.com/thread.jspa?threadID=606936&tstart=360

Looks like there are allot of cables and drive settings to be checked as well luckily everything in my configuration is stock or added on I will try again tomorrow. Wish me luck...