SE/30 Only Boots From System 6

[stachowiak]

I'm repairing an SE/30 logic board for a friend and have a problem with booting up with system 7.  I don't have any history with this board, but it had some corrosion in the battery holder but was otherwise very clean.  I have successfully recapped and repaired over 50 logic boards, and this board is the one of the cleanest I've ever encountered.  I have a working SE/30 that I use for troubleshooting and am using this to test this board, so I know that the analog board, hard drive and floppy are good.  I've also installed the good Ram and Rom Sim from my good board into the problem board,  Booting from a system 7 hard drive hangs up just after the "Welcome To Macintosh" screen appears.  If I boot from a system 7 floppy takes me to a blank finder screen with no text or icons.  The system will boot from a system 6 floppy and will mount the system 7 hard drive.  When I run my Snooper test program that's on the hard drive it fails the memory test with a "Buss Error" and locks up the machine.  I've checked continuity from the ram sockets to the mux and bus connections and haven't found any open connections.  I've also traced out all of the SCSI chip pins without finding anything.  A Google search found a listing for a similar problem on this forum from April of 2017 but it doesn't seem to be here anymore.  Any suggestions would be appreciated.

 

[Crutch]

Is that a stock SE/30 ROM SIMM?  What version of System 7 are you running?  I assume you confirmed the System 7 hard drive runs fine on your other known good SE/30?

 

[LaPorta]

Yes, is this a known-working 7 installation?

 

[stachowiak]

I tried to make it clear that all the components connected to the problem board are good.  They all work properly with my several good SE/30 logic boards.  The ROM Sim is the stock Apple version and I tried several good ones that I have with the same results.  The hard drive has system 7.1 installed.  It's not a machine specific installation that I use it for testing and works properly when connected to a good Plus, SE, SE/30, Classic, or Classic II logic board.  The problem is definitely related to the one SE/30 board that I'm working on.  I wish it wasn't a board problem as it would be a lot easier to resolve.

 

[joshc]

I have two SE/30s and both of them misbehave a lot if the ROM SIMM is not seated very tightly into the ROM socket. The only way I got mine to actually behave consistently is by hot gluing the ROM SIMM in. Do you think you may have a similar problem?

I would say the same for the RAM, is any of it sitting loosely in the RAM sockets?

With one of my SE/30s, the ROM SIMM unseated itself while the machine was on! That was the point at which I got the hot glue out...no problems since.

I might be far off the mark here - hopefully it's something as simple as the above rather than component specific failure.

Also, have you tried booting from System 6.0.7/6.0.8 instead? No idea why that would help but something to try...

 

[joshc]

Sorry, you had indeed already mentioned it booting from System 6 OK.

I am struggling to see how this is a hardware issue in that case. Have you tried booting from a different System 7 drive or a System 7 boot/install disk?

Did you try booting System 7 with extensions off?

 

[stachowiak]

I won’t be getting back to the board until tomorrow.  I’ve tried my 7.1 and 7.5 hard drives, both boot up to the “Welcome to Macintosh” screen and then locks up at that point.  It won’t boot from any system 7.X disk tools floppy either.  Both of my system 7 hard drives as well as my floppy disk work perfectly with 2 of my good SE/30 logic boards.  Just to confirm things, I just installed one of my good SE/30 boards along with the ram and Rom Sim from the problem board and it boots up perfectly.  I’m having a very hard time thinking that this is a software problem.

Thanks,

Dick

 

[superjer2000]

Have you tried booting with extensions off or using extensions manager to see if you can narrow down what hardware is being called at the point in the boot process that it freezes?  

Have you tried zapping the PRAM?

 

[stachowiak]

Thanks to everyone that has responded to my issue so far.  I worked on the board more today with no success.  It boots fine from a floppy with system 6.08 and mounts the hard drive that I use for testing.  This SE/30 board will not boot from a hard drive or floppy that have any version of system 7 with or without extensions.  The pram has also been zapped with no affect.  Just to confirm that my system 7.1 hard drive and 7.1 floppies are good, today I used them without any problems on 4 other logic boards I have.  Last year I had a Classic logic board that had the exact same problem but I don't remember what I did to fix that one but it involved repairing rotted traces.  System 7 must make use of something in the circuitry that system 6 doesn't.  This board has been properly recapped and is one of the cleanest SE/30's I've ever worked on.  Other than a very small amount of corrosion on one of the battery contacts, there is not a spec of green on any of the chip legs.  Also when I removed the old caps, none of them showed any sign of leakage and I'm positive that this board has never been recapped.  As I said before, I have no history with this board.  The person I'm fixing it for told me it didn't work (he also has no history with it other than he's owned it for about 25 years and wanted to get it going again) and asked for a recap, so I never tried it to see what it did or didn't do before I started.

I've poured over many post on this site and saw several references to systems that wouldn't boot with the suggestion to see if it will boot under system 6, so hopefully my problem means something to someone. Unfortunatly the original poster never posted if they had solved the problem and what they ended up doing.

Thanks,

Dick

 

[Crutch]

I know this is unhelpful but my gosh that is weird. I can guarantee you no part of the SE/30s logicboard circuitry was designed special purpose for System 7 or is uniquely used by System 7 but not System 6. (Part of the reason I can guarantee this is, the SE/30 came out a year before System 7!!).  Does it always hang at the exact same spot? Can you drop into a debugger?

 

[stachowiak]

It happens at the same point each time.  A few seconds after the "Welcome To Macintosh" appears, the mouse pointer shows up in the top left of the screen. and the hard drive activity stops.  Moving the mouse does move the pointer on the screen. I haven't hit the interrupt switch but I will try that later tonight.  I also tried running Mactest and it passed all the test.  When I run Snooper it fails the ram test with a "Buss Error".  I'm running Mactest and Snooper from the hard drive but booting from a system 6.08 floppy.  I've already spent way too much time on this but I'm too anal to give up.  Is it possible that system 7 uses more of the address lines than system 6?

Thanks,

Dick

 

[LaPorta]

This is going to sound weird, but...how much RAM do you have installed on the board? Does either installation have Mode32 or equivalent on there to address more than 8 MB of RAM? Is there any chance that 7.1 is addressing a higher RAM address than 6, and that address has a physical issue on the board, be it connectivity or the like? Something that might not fail the startup RAM test outright. This is going way out in left field, but I am trying to see what might actually cause this sort of issue. Your Buss Error probably indicates a hardware issue as you are saying. What I'm trying to theorize is that System 6 may not "see" it because of it's nature. I see that you tried the RAM on another machine and it worked fine. This sounds more like it is a hardware problem manifesting in software because of the different resources the softwares are utilizing. 

 

[Crutch]

I thought of that too, but he said it happens with extensions off (therefore no MODE32).  If he were running a modded ROM this would make more sense but we ruled that out also.

If the mouse is moving, the CPU is still working and talking to the RAM, and if you’ve seen Welcome to Macintosh should should be able to drop into Macsbug.  I am still betting this is somehow a software issue somehow ... I concede I am very likely wrong ... but dropping into Macsbug will be insightful.

 

[stachowiak]

I know it sounds like a software issue but for that to be the case, all of the following would have to be true:

1 - Over 50 logic boards that I have repaired using my testing drive with system 7.1 have some sort of hardware problem that ignores a bad system folder.

2 - All my original Apple provided System Tools floppy disk have a corrupt system on them.

3 - Only my system 6 based floppies are good.

I've never used a debugging tool so I have study up on that.  Right now I'm thinking that system 7 must be trying to load into an area in memory that has a problem on this board.  I appreciate all of the post I've gotten so far and will work on this till it's resolved.  I'll also post the fix once found to help the next person that encounters this.  I hate when people never post what worked for them to solve a problem.

Thanks,

Dick

 

[Bolle]

Well, to rule out a RAM problem you could try to put in just 2 or 4MB in total.

You said you got corrosion next to the battery. Check if the PRAM/RTC chip is working correctly (time advances and PRAM settings are actually remembered)

I had a problem booting certain system folders when communication to the PRAM didn’t work.

What else did you do to the board? I once had a board where I suspected the CPU to be bad. (it was bad but that doesn’t matter here)

I installed a socket after removing the old CPU and put one of my spares in. The spare CPU I got was one with a golden top while the original CPU had a black plastic case. The socket I installed had a pin layout that exactly matched the plastic packaged CPU. Turned out the gold/ceramic packaged CPU has a few more undocumented pins that went into blind holes into the socket and the result was that the machine would only boot System 6 and fail loading System 7.

As soon as I installed the two missing pins in the socket the board worked just like you would expect with the ceramic 030 and booted System 7 just fine.

 

[Crutch]

I've never used a debugging tool so I have study up on that.  Right now I'm thinking that system 7 must be trying to load into an area in memory that has a problem on this board.  I appreciate all of the post I've gotten so far and will work on this till it's resolved.  I'll also post the fix once found to help the next person that encounters this.  I hate when people never post what worked for them to solve a problem.

Yeah it’s probably not a software problem given your history.  But it really, really feels like one.

Install Macsbug in your System Folder.  When the freeze happens, hit the Interrupt button on your programmer’s switch.  Assuming Macsbug then comes up:  you must be in an infinite loop.  The system is waiting for something to happen that isn’t happening - very possibly for something to come back from the hardware.  Possibly even for something stored in PRAM, in fact, as Bolle implied!  Even if it’s a hardware problem, figuring out where the Mac is getting stuck could help pin down the hardware issue.

Try “IP” (instruction page) to get a list of instructions around the program counter.  If you want, take a photo of the result with your phone and post it here.  Someone might be able to help interpret the results.

 

[techknight]

The other thing is System 7 requires more memory to run. So if the high order 74F253 muxes are bad, your gonna have problems accessing and addressing more RAM past a certain point. which WILL crash system 7. 

 

[LaPorta]

To help diagnose, why not try and start System 6 with Multifinder and see if that uses enough to cause it, if not, launch multiple programs and see if it occurs.

 

[Crutch]

The System Heap lives in low memory with Application Heaps piled on top of it ... the stack lives in high memory and grows downward from BufPtr (this is true in both Systems 6 and 7).  If very high memory had a problem, there would be issues accessing the stack even in System 6.

But I suppose it’s possible you have an issue somewhere in the middle of RAM, where System 7 is trying to put the Finder but System 6 doesn’t get that high.   Then LaPorta’s idea would be useful ... MultiFinder with a bunch of apps running will indeed eat up the same bits of RAM as System 7 with a bunch of apps running.  But if this were indeed your problem, running System 6 would very likely not be long-term stable either.  So I am a bit skeptical.

I have an old extension called MemHell that loads a bunch of junk all over RAM, continually allocates & disposes relocatable blocks, etc. to make it easier to find/debug memory leaks in your code.  I will see if it works as I remember and if useful I will upload it ... I don’t see it on any of the Mac software repo sites.