Hi again from Norway!
This time, I'm trying to fix my SE (first "revision" with resistor RAM config) as I have resurrected the SE/30 so I can make proper 800kb floppies ( https://68kmla.org/bb/index.php?threads/se-30-crt-vertical-foldover-woes.50130/ ).
Long story short, the SE's previous owner took data security seriously and drilled a hole in the "hard drive", the hard drive being the bottom of the case...
Luckily, only a trace for the non-socketed CPU's D3 (pin 2) got hit lightly and was "easily" repaired (C47 or was it C48 also got a bit mangled. Couldn't fin it shorted towards anything critical):
For more detailed pictures, look over at my VCFED post almost a year ago:
forum.vcfed.org
I have removed the battery (no signs of leakage), cleaned thoroughly the SIMM sticks with fiber glass pen and electro spray on q-tips (dirty!).
Also, all socketed chips including the PLCC have been lifted and got the same treatment with a good spray in the sockets.
Symptoms:
Before cleaning the SIMM sticks I got RAM errors when testing in Snooper. Now gone.
In fact, I have just run Snoopers RAM test with all algos turned on for over 10 successful loops and no issue.
All other Snooper tests pass, except for the ones that require a loop back dongle.
The issue appears when I test the CPU, or try the "heavier" math benchmarks. Integer is fine, but floating point and functions makes it crash / lock up (no mouse movement, or "busy ball" jumping fast from top to bottom), just when it is about to show the results. Same thing with the "regular" CPU test (99.9%-ish done).
The CPU can also lock up after boot, sitting too long idle.
I also just got a critical error message in Snooper: "Coprocessor not installed. Restart"
A year ago I also got error messages in OS 6.0.8 on a floppy I made with a greaseweazle about address bus errors.
Is this just Snooper 2 playing tricks with me by not being too SE friendly?
Also, on an SE, does the top horizontal lines on the display flicker when testing the RAM?
Mine does. The horizontal width of the vertical bars do seem to differ during the test. My SE/30 doesn't.
Before I head in with my head first into a pool of red herrings, what other diagnostic programs and tools are worth their salt?
The CPU is faulty?
The decoupling cap traces that got mangled causes this? Both sides of the cap shows continuity with its neighbour friends.
Apocrypha:
Did you know that a not-completely-settled-and-ready-because-it-is-still-cold-and-grumpy "Apple" 40MB SCSI HDD in an SE/30 can give RAM errors in Snooper booted from floppy?
Took me some time. Literally.
The SIMM sticks and sockets are now clean though.
This time, I'm trying to fix my SE (first "revision" with resistor RAM config) as I have resurrected the SE/30 so I can make proper 800kb floppies ( https://68kmla.org/bb/index.php?threads/se-30-crt-vertical-foldover-woes.50130/ ).
Long story short, the SE's previous owner took data security seriously and drilled a hole in the "hard drive", the hard drive being the bottom of the case...
Luckily, only a trace for the non-socketed CPU's D3 (pin 2) got hit lightly and was "easily" repaired (C47 or was it C48 also got a bit mangled. Couldn't fin it shorted towards anything critical):
For more detailed pictures, look over at my VCFED post almost a year ago:
Macintosh SE main PCB repair
Hi! Just rescued a Macintosh SE and a Toshiba T1000 from a friends attic. The SE had miraculously no battery (removed!) or cap leakage whatsoever (T1000 has a different story...) A catch with the SE though is that the previous owner, before it was donated to my friend, "drilled a hole in the...
forum.vcfed.org
I have removed the battery (no signs of leakage), cleaned thoroughly the SIMM sticks with fiber glass pen and electro spray on q-tips (dirty!).
Also, all socketed chips including the PLCC have been lifted and got the same treatment with a good spray in the sockets.
Symptoms:
Before cleaning the SIMM sticks I got RAM errors when testing in Snooper. Now gone.
In fact, I have just run Snoopers RAM test with all algos turned on for over 10 successful loops and no issue.
All other Snooper tests pass, except for the ones that require a loop back dongle.
The issue appears when I test the CPU, or try the "heavier" math benchmarks. Integer is fine, but floating point and functions makes it crash / lock up (no mouse movement, or "busy ball" jumping fast from top to bottom), just when it is about to show the results. Same thing with the "regular" CPU test (99.9%-ish done).
The CPU can also lock up after boot, sitting too long idle.
I also just got a critical error message in Snooper: "Coprocessor not installed. Restart"
A year ago I also got error messages in OS 6.0.8 on a floppy I made with a greaseweazle about address bus errors.
Is this just Snooper 2 playing tricks with me by not being too SE friendly?
Also, on an SE, does the top horizontal lines on the display flicker when testing the RAM?
Mine does. The horizontal width of the vertical bars do seem to differ during the test. My SE/30 doesn't.
Before I head in with my head first into a pool of red herrings, what other diagnostic programs and tools are worth their salt?
The CPU is faulty?
The decoupling cap traces that got mangled causes this? Both sides of the cap shows continuity with its neighbour friends.
Apocrypha:
Did you know that a not-completely-settled-and-ready-because-it-is-still-cold-and-grumpy "Apple" 40MB SCSI HDD in an SE/30 can give RAM errors in Snooper booted from floppy?
Took me some time. Literally.
The SIMM sticks and sockets are now clean though.
