A few thoughts:
- I2C lines do indeed idle high.
- There's something funny with the supply to the RTC: The +5V supply should be stiff enough override the battery voltage when the power is on. That suggests there's a higher-than-expected resistance somewhere likely D2? (or R130/R212 are too low...). The RTC should have most of 5V on it to allow accesses.
- The POST code source used to be in the ROOL sources: It's no longer built, but the code should still linger in the archives somewhere (it was under Sources/Kernel/TestSrc). According to a copy I saved a few years ago you're stuck in the RAM test:
Code:
;; Border colours used for self-test indicators; [ VIDC_Type = "VIDC20"C_2NDBANK*&40000000+&C00070 ; jumping to image in 2nd ROM bank (Dark blue)C_ARMOK * &40000000+&7000C0 ; testing ROM (Magenta)C_RAMTEST * &40000000+&C07000 ; testing RAM (Blue)C_FAULT * &40000000+&0000F0 ; failed tests(Red)C_PASSED * &40000000+&70C000 ; all passed (Green)C_WARMSTART * &40000000+&707070 ; not tested (Mid grey)C_DEFAULT * &40000000+&000000 ; default (Black) ];; - The POST code is clearly running from ROM - or at least starting to - but must be crashing before it can flash the floppy (which occurs at the same time that the screen goes Green or Red). That likely means the POST box won't tell you much more than you already know, but could provide some clues. You'll certainly see some output from it though, given it's got that far.
- From memory, though, having the POST box present causes it to do a more detailed memory test than it otherwise would.
- The fact that it's crashing with no RAM fitted suggests it could be falling over while trying to detect the RAM size/configuration?
Statistics: Posted by wiggy — Wed Jul 31, 2024 12:53 pm