6360 locks up when copying - do i know why?

[Ncc74656]

my older macs (68040 and 80mzh PPC) did not have any issues copying files. however my 6360 both with and without G3 card - will lock up when copying a file and then unlock, copy some, and lock up again.

i tested different CD drives, HD's, Operating systems (8.1/8.5/8.6/9.0.4/9.1).
while running mactest pro - RAM testing will lock the system when testing an address and unlock when it moves to the next, then lock up again.

these systems have a 40mhz bus - do they also have the same split 64 lanes between two 32 lane bus's on the board? so when the memory controller is maxed out is the ADB port unable to Rx/Tx?

would the system having 136MB of ram be causing the lockups as the memory buffer is that much larger and thus the flooded system controllers are noticeable? where as smaller ram or slower CD drives might mask this design fault?

the system locks up when copying via HDD-HDD or CD-HDD but seems to do it less often with smaller files.

 

[LaPorta]

Where does the ADB port come into play? Do you mean the entire machine actually freezes (i.e. no mouse movement possible)? With all you mentioned, across all HDs, CDs, Systems, etc...I would suspect a RAM issue. EIther the RAM itself, or some sort of intermittant contact. Also would think that is more likely when you say it does it less with smaller files (meaning less memory swapping necessary).

 

[Ncc74656]

yea, the i/o stops. the ram is new? its from a site that packaged it as new anyway. i have more sticks, smaller so i can test things out. im running the hardware tests now - will be well into tomorrow until i know more.

so do these systems not have the issues i listed? its not a limitation of the split bus or large ram ammount?

 

[LaPorta]

There's no Mac I am aware of that actually freezes during copying. Slow, jerky mouse movement perhaps, but not freezing.

 

[Ncc74656]

ok, thats good to know. thanks. gives me a better idea of what im dealing with.

i have no hard locks in purpiturity and i do not have any other anomolus issues. the hardware quick tests so far - are passing so... im not suspecting a capacitor issue as id expect that to lead to more HARD failures.

i have an SSD and adapter that im using - it does NOT freeze when transplanted into a 6300CD so this leads me ot believe my issue is NOT in the hard drive.

perhaps a bad HDD cable? or maybe an OS conflict im not aware of. maybe ill try a full clean install

 

[Ncc74656]

finished an over night hardware scan. all ram passed its tests. system locked up while testing each memory block and then unlocked when changing blocks.
im going to swap different ram sticks and leaving a dimm slot open as well. see if there is a configuration that hte system does not lockup with

 

[LaPorta]

Possibly photos of the slots themselves could aid in any physical issues.

 

[Byrd]

Try an old spinning HD in your 6360; SSDs aren't always infallible in early IDE implementations

 

[LaPorta]

I thought the OP stated they tried different HDs?

 

[Cory5412]

In general, Macs prioritize keyboard and mouse input above a lot of other things, although Classic Mac OS is also... not very good software, by modern standards, so it doesn't necessarily surprise or concern me that it looks like it's just hard-locking for a while while it's busy.

If it eventually unlocks, then I'm kind of tempted to say "is it impacting anything else? If it reliably unlocks and ultimately finishes, and everything else runs fine, then I don't really see any problems here."

do they also have the same split 64 lanes between two 32 lane bus's on the board

The 6360 is an updated PCI-based architecture from the 6200 and 6300, but the "right 32, left 32" thing was never true for any system it's talked about. The site that published it didn't bother to get the correct information (which was available at the time if you wanted it) and dismissed notes from people who actually had the machines that they weren't all that bad.

The architecture of the 6200 and 6300 is best described as "Mac 630 with a PPC 603 upgrade soldered on" -- it was a common architecture in budget oriented PCs too.

Entertainingly, the machines aren't even that bad. If you put 7.6.1 or 8.1 and Speed Doubler 8 on a 6200, it becomes basically average for the era, comparable with their contemporary 66-80MHz PPC 601s. The 6300 would benefit from this but itself contains a larger 32k L1 cache (matching basically every other PPC until the G3) which improves things significantly as well.

The PowerBooks 1400, 2300, and 5300 also share this architecture.

 

[Ncc74656]

So I spent a couple hours today diagnosing and troubleshooting. I stumbled across what I believe demonstrates an operating system level error.

I had the same operating system and hard drive in a 6300, it did not present these problems. That led me to believe that wasn't the issue. However...

I had another IDE Drive lying around that I stuck in, Mac OS 8.1. It did not demonstrate a hard-luck condition when copying a CD to the Dr. So I made a ramdisk, I copied it's operating system into RAM and then plugged my SSD back in, I am currently copying a CD to my SSD without hard locking. So, it only hard locks when I'm booted from the SSD in 8.6 and 9.1

I believe my next step is to reinitialize the ssd, make new partitions, new hard drive driver and see if a clean install solves this

 

[LaPorta]

So there IS an SSD! Your first post said nothing about SSD; that would have led me to what Cory said.

 

[Ncc74656]

i zapped the Pram a few times and that seems to have solved this issue for hte most part. i can multi task again while copying things in the background.
i need to look into how PR works.

do any of you have a link to how this systems bus was designed? cory said the left/right 32 lane was never a real think and id love to know how they actually function. people bash on these systems SO MUCH and i want to understand what their true design faults are.

 

[CC_333]

people bash on these systems SO MUCH and i want to understand what their true design faults are.

Basically, as @Cory5412 says, they're pretty average for the time. About the only major flaw I can think of was that the 603 chip used in the 6200 had insufficient cache for the 68k emulator that was needed for System 7 to properly run (much of the OS before 8.5 or so was actually not fully PPC native and needed to run under emulation on Power Macs).

Anyway, as a result, the emulator couldn't run as efficiently, and thus caused the whole OS to feel very slow, hence the 6200's bad reputation.

Aside from that, pretty much what Cory said. The 6200, architecturally, is basically a Performa 630 with a built in PowerPC upgrade. If that's a design fault, it's only because people tend to overestimate it and compare it unfairly to its contemporaries; it's perfectly acceptable for what it is (a low-end budget machine meant either for schools or for first-time Mac users), and the main reason it's slow is the 603's skimpy L1 cache. If anyone Mac shopping at the time wanted a better machine, they bought a better Mac. Why else did Apple have so many different models? So buyers had choices, of course! It was an admirable, albeit misguided and severely mismanaged attempt at maximizing the Mac's market share (the clone program was similar, except it ended up backfiring severely for Apple, because, ironically, it worked too well: the clone makers offered many high-end Power Mac-spec machines at near Performa prices, and Apple couldn't compete).

Anyway, later Performas, such as the 6360 mentioned, and other later 6xxx models, received an improved PCI-based architecture including the enhanced 603e with its bigger L1 cache, so they performed much better, if only because they could run Apple's 68k emulator, and thus System 7, more efficiently.

c

 

[Ncc74656]

well my 630CD was our first system. i think i was 6 or 7? mom was shocked i was able to hook it up with out reading a manual. i thought it was rather obvious, everything had markings - it was like lego.
aol 2.7 iirc, global village model (2800 baud). the sales woman told us "this system has 8MB of ram! its more than you will ever need"
learned how to script on that system. found escape velocity and wasted THOUSANDS of hours taking over the galaxy.

i thought it was a good system, fast, ran well for everything i needed it for.
its too bad apple broke up their lines so much. they could have standardized, probably made manufacturing cheaper, and given their users more longevity and smoother upgrade path into more apple products. instead of jumping ship onto PC's.

 

[CC_333]

its too bad apple broke up their lines so much. they could have standardized, probably made manufacturing cheaper, and given their users more longevity and smoother upgrade path into more apple products. instead of jumping ship onto PC's.

Yes, it is.

They did eventually streamline their lineup around 1997-1998 when they bought Steve Jobs' NeXT and he became CEO (this move, among others, such as releasing the iMac (and eventually the iPod) and cancelling the clone program, arguably saved Apple by bringing it back from the brink of bankruptcy), but had Apple done some of these things a few years earlier, it's very possible it would've been in much better shape. Much of the damage, unfortunately, was self inflicted I think, due largely to a near-fatal combination of mismanagement and shortsightedness at various levels throughout the company.

c

 

[Cory5412]

Great to hear this is working better now.

A bunch of the developer notes are here https://cdn.preterhuman.net/texts/computing/apple_hardware_devnotes/ and this is a good spot to learn about the design of the actual machines and the architecture. There's block diagrams of most machines and Apple goes into at least some detail about what each component is or what function it serves in the architecture.

The 6360/6400/6500 should be covered (more or less) by the 5400 document and the 6200/5200 are together in a document called 5200 and 6200.

W/re the 6200 -- most of why this perception/misinfo persists now, in addition to the original site that published the info never going back on it or revising it, is people who are emotionally invested in the idea that some macs are bad enough that they're outright worth pretending they never existed, or that Apple should never have built them.

they could have standardized, probably made manufacturing cheaper, and given their users more longevity and smoother upgrade path into more apple products. instead of jumping ship onto PC's.

Maybe -- probably even, depending on what you're thinking of in particular.

By the late '90s, Apple didn't really mean in-band upgradeability to be used for life extensions, The single shining piece of evidence that they were thinking of that is the 604e@233MHz CPU upgrade and to be honest I think a lot of why that upgrade existed is an acknowledgement that the 9600/233 and the 9500/132 are literally the same computer save for which CPU is installed, so Apple can in theory retain some goodwill by letting you get the faster machine.

That's not a completely unique moment in Mac history, but it's also not common.

CC mentioned this but Apple did eventually rationalize the product line a lot. Something like forty or so models were coalesced into around 7-10 or so, mostly accounting for different physical form factors to meet different needs.

The iMac G3 is often credited with saving Apple, but I think the G3 CPU on its own did a lot of the groundwork on that one. From early 1997 to late 1998, Apple managed to cut the retail price on almost everything it sold by a thousand dollars or more while replacing big machines with smaller ones that were significantly faster, and simplifying nearly everything about the company and the platform in the process.

 

[jajan547]

Take a look at this I scanned it to archive.org has Information for the 6200/6360 lineup that may help was previously not online. Starts on page 112.