2/50,2/130,2/160

        Bit 0 is at the top and bit 7 is at the bottom; the patterns
        below are shown with bit 0 on the left.

            Pattern     Status                             Problem
            --------    ------                             -------
            00000000    After complete power-up sequence   System running OK
            00000001    Finished self-test, setting up to  Check CPU first,
                         boot                               then expansion brd
            00000010    Entering user watchdog routine     Software bug
            00000011    After local memory verified        CPU
            00000111    After diags, while setting up      CPU
            00001000    Blinks off and on while NMI OK     System running OK
            00010001    Testing context registers          CPU
            00100001    Testing constant data in seg map   CPU
            00100010    Testing addr dependency in seg map CPU
            00100011    Testing data lines in seg map      CPU
            00110001    Testing constant data in page map  CPU
            00110011    Testing data lines in page map     CPU
            00110010    Testing addr dep in page map       CPU
            01000000    Testing PROM contents              CPU or PROMs bad
            01010000    Testing SCC chip(s)                CPU
            01110000    Sizing memory before constant test CPU, then expansion
            01110001    Testing constant data in memory    CPU, then expansion
            01110010    Testing addr dependency in memory  CPU, then expansion
            01111111    Testing parity circuitry           CPU, then expansion
            10000001    Testing timer chip                 CPU
            11110001    Setting up memory after diags      CPU, then expansion
            11110010    Setting up maps after diagnostics  CPU
            11110011    Setting up frame buffer and video  CPU -- check video
                         clock jumper                       jumpers
            11110100    Setting up NMI or keyboard         CPU
            11111111    A reset sets LEDs to this state    CPU or PROMs bad, or
                                                            "bad device"

3/50
        Bit 0 is on the left. The pattern is shown as it appears on the
        LEDs.

            Pattern     Status                          Error
            --------    ------                          -----
            11111111    Resetting                       CPU or PROMs bad
            00000000    Test 0: CPU to SCC path         CPU board (SCC) bad
            10000000    Test 1: boot PROM               Boot PROM bad
            11000000    Test 3: context register        CPU board (MMU) bad
            00100000    Test 4: segment map RAM rd/wr   CPU board (MMU) bad
            10100000    Test 5: segment map RAM         CPU board (MMU) bad
            01100000    Test 6: page map RAM            CPU board (MMU) bad
            11100000    Test 7: memory data path        CPU board bad
            00010000    Test 8: bus error detection     CPU board bad
            10010000    Test 9: interrupt capability    CPU board bad
            01010000    Test 10: MMU read access        CPU board bad
            11010000    Test 11: MMU write access       CPU board bad
            00110000    Test 12: write to invalid page  CPU board bad
            10110000    Test 13: write to protected pg  CPU board bad
            01110000    Test 14: parity error check     CPU board bad
            11110000    Test 15: parity error check     CPU board bad
            00001000    Test 16: memory tests           CPU board bad
            00000001    Self-tests have found an error  See below
            00000010    An exception class error found  See below

        "Marching ones" (cycling through 10000000, 01000000, 00100000,
        etc.) indicates that Unix is running OK. On power up, it cycles
        through the tests in the chart above, then boots Unix. Pattern
        11111111 may also mean that a SCSI device was powered up prior
        to the CPU being powered up. If LED 7 (00000001) lights up while
        the tests are being performed, it indicates that the test
        failed. If LED 6 (00000010) lights up while the tests are being
        performed, it indicates that an unexpected error (bus error,
        address error, unexpected interrupt, etc.) occurred during the
        test. When all tests are finished, LED 5 (00000100) starts
        blinking to indicate that the ROM monitor is running and/or Unix
        is booting.

3/60,3/60LE
        Bit 0 is on the right. The pattern is shown as it appears on the
        LEDs.

            Pattern     Status
            --------    ------
            11111111    Resetting
            00000001    PROM checksum test
            00000011    Context register test
            00000100    Segment map read/write test
            00000101    Segment map address test
            00000110    Page map test
            00000111    Memory path data test
            00001000    Nonexistent memory bus error test
            00001001    Interrupt test
            00001010    Time-Of-Day clock interrupt test
            00001011    MMU protection/status tests
            00001110    Parity error test #1
            00001111    Parity error test #2
            00010000    Memory test
            10000000    Self-tests have found an error
            01000000    An exception class error occurred

        "Marching ones" (cycling through 10000000, 01000000, 00100000,
        etc.) indicates that Unix is running OK. On power up, it cycles
        through the tests in the chart above, then boots Unix. If LED 7
        (10000000) lights up while the tests are being performed, it
        indicates that the test failed. If LED 6 (01000000) lights up
        with the tests are being performed, it indicates that an
        unexpected error (bus error, address error, unexpected
        interrupt, etc.) occurred during the test. When all tests are
        finished, LED 5 (00100000) starts blinking to indicate that the
        ROM monitor is running and/or Unix is booting.

3/2xx
        Bit 0 is at the top and bit 7 is at the bottom; the patterns
        below are shown with bit 0 on the left.

            Pattern     Status
            --------    ------
            11111111    Resetting
            10000000    PROM checksum test
            01000000    DVMA register test
            11000000    Context register test
            00100000    Segment map read/write test
            10100000    Segment map address test
            01100000    Page map test
            11100000    Memory path data test
            00010000    Nonexistent memory bus error test
            10010000    Interrupt test
            01010000    Time-Of-Day clock interrupt test
            11010000    MMU protection/status tests
            00110000    ECC error test
            10110000    Cache data 3-pattern test
            01110000    Cache tag 3-pattern test
            11110000    Memory tests
            01001111    Initializing MMU
            00000001    Self-tests have found an error
            00000010    An exception class error occurred

        "Marching ones" (cycling through 10000000, 01000000, 00100000,
        etc.) indicates that Unix is running OK. On power up, it cycles
        through the tests in the chart above, then boots Unix. If LED 7
        (00000001) lights up while the tests are being performed, it
        indicates that the test failed. If LED 6 (00000010) lights up
        while the tests are being performed, it indicates that an
        unexpected error (bus error, address error, unexpected
        interrupt, etc.) occurred during the test. When all tests are
        finished, LED 5 (00000100) starts blinking to indicate that the
        ROM monitor is running and/or Unix is booting.