12

After removing the Intel DX 486 card from my RiscPC (running RiscOS 3.6), the computer doesn't start up properly. Instead, it displays the text POST in red.

What does this error message mean?

  • Welcome to Retrocomputing Stack Exchange. It's good to see another person who's still tinkering with their RiscPC; this is only the second question about the machine on this site! – wizzwizz4 Oct 4 '16 at 18:54
  • Have you tried holding down Shift during boot? This will activate some sort of Safe Mode where it skips or uses default setting for boot, but I'm not sure what it's called. – wizzwizz4 Oct 4 '16 at 19:03
9

What does "POST" mean?

On the RiscPC, "POST" stands for Power On Self Test. The POST sequence is performed on computers running RiscOS 3, and ensures that the computer is working properly before it starts up. If the computer fails this test, an error message will be displayed on the screen, informing you that there was a POST error.

How do I know what the problem is?

You may notice your floppy-drive light flashing. After a POST error is detected, the floppy-drive light will output a sequence of flashes a little like Morse Code. The error code is a 32-bit binary number, where a long flash represents 1 and a short flash represents 0. Here are some common fatal error codes and their causes:

  • 0000 0000 0001 0000 1001 or 0000 0000 0001 0001 1001
    The CMOS checksum is invalid. This can be fixed by resetting the CMOS data (see below). This error is often caused by a flat CMOS battery, so if the error persists after resetting the CMOS data, you might need to replace the battery.
  • 0001 0000 0000 0100 1001 or 0001 0000 0000 0101 1001
    The CMOS chip is unreadable. This probably means that your CMOS battery has leaked and damaged the CMOS chip itself. This is harder to fix, requiring not only a new battery but also a CMOS chip.
  • 0010 0000 0001 0100 0001
    The RAM Control line has failed. This is a really big problem. Most likely your CMOS battery has leaked and damaged the RAM Control line, rendering the RAM unusable. This is difficult, but not impossible to repair.

Note that these error codes all have 0s at the beginning. You don't need to count exactly how many there are of these, so long as you have the end part of the code.

How do I reset the CMOS data?

To reset the CMOS data to its default values, follow these instructions:

  1. Turn off the computer if it is on.
  2. Press and hold the DEL key.
  3. Turn the computer back on.
  4. Keep holding the DEL key until you see a black screen with a red border.
  5. Release the DEL key. The CMOS data is now reset to sane defaults.

What if I have a different error code?

The error codes are designed so that each binary digit corresponds to a specific fault code. Some of the fault codes aren't errors; these are just informative. If a code is an error the text will be bold. The bits are numbered from right to left, where bit 0 is the right-most digit (the 1 in 0000 0000 0000 0000 0001).

  • Bit 0: Self-test due to power-on. This is normal.
  • Bit 1: Self-test due to interface hardware. This shows that the POST test wasn't started directly from the boot sequence.
  • Bit 2: Self-test due to test link. This means that you manually triggered the POST test (I think!)
  • Bit 3: Long memory test performed. This just tells you that nothing went wrong before the long memory test.
  • Bit 4: ARM3 fitted / ARM ID read and not ARM2. This gives information about the hardware. If you have an ARM2 installed and this shows up, there might be a problem, but likely not a major one.
  • Bit 5: Long memory test disabled. This is not a problem; it just means that the long memory isn't tested.
  • Bit 6: PC-style IO world detected.
  • Bit 7: VRAM detected.
  • Bit 8: CMOS RAM checksum error. The CMOS checksum is invalid, which means that the CMOS memory is corrupt. This can be fixed by resetting the CMOS data (see above). This error is often caused by a flat CMOS battery, so if the error persists after resetting the CMOS data, you might need to replace the battery.
  • Bit 9: ROM failed checksum test. This means that your ROM is corrupt. This might not be significant enough to cause much of a problem, but expect erratic behaviour and the occasional crash.
  • Bit 10: MEMC CAM mapping failed.
  • Bit 11: MEMC protection failed.
  • Bit 12: IOC register test failed.
  • Bit 13: Reserved code (RISC PC). This code should never appear. If it does, please ask a question about it!
  • Bit 14: VIDC Virq (Video Interrupt) timing failed. This means that things probably won't display properly on the screen, because the "draw screen content" timing doesn't line up with the actual screen refresh.
  • Bit 15: VIDC Sirq (Sound Interrupt) timing failed. This is similar to the Virq failure, but for sound. Audio might sound stuttery, or play in segments that are the wrong way around (e.g. 2143658709 instead of 1234567890).
  • Bit 16: CMOS unreadable. This probably means that your CMOS battery has leaked and damaged the CMOS chip itself. This is harder to fix, requiring not only a new battery but also a CMOS chip.
  • Bit 17: RAM control line failure. This means that the RAM control line is probably damaged, which is difficult to repair.
  • Bit 18: Long RAM Test failure. The long memory test from bits 3 and 5 has failed.
  • Bit 19: Reserved code (RISC PC). This code should never appear. If it does, please ask a question about it!
  • Bits 20-32: Unused codes. These should all be 0, if they are shown at all.
  • If you want to know how to replace the CMOS battery or chip, or fix the RAM Control line, please ask a follow-up question. – wizzwizz4 Oct 4 '16 at 20:17
0

i found following resource for Archimedes POST faults which might help:

http://www.retro-kit.co.uk/page.cfm/content/Common-faults-on-the-Acorn-Archimedes/

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.