1. Forum
  2. Usenet
  3. COMP.OS.OS9

  • MVME167s don't want to boot

    From Ralf Kiefer@21:1/5 to All on Mon Aug 8 18:51:04 2016
    Hi,

    there are some problems to start some(!) MVME167.

    EPROM sockets U1 and U2 are populated with the Motorola 167Bug, the
    sockets U3 and U4 with the Microware stuff running OS-9/68k V3.0. They
    are marked as "1.01" probably by Dr. Keil in Germany in the 1990s.

    After reset:

    | Copyright Motorola Inc. 1988 - 1994, All Rights Reserved
    |
    | MVME167 Debugger/Diagnostics Release Version 2.2 - 01/14/94
    | COLD Start
    |
    | Local Memory Found =06000000 (&100663296)
    |
    | MPU Clock Speed =33Mhz
    |
    | ROMboot in progress... To abort hit <BREAK>

    That's what I exspect. The configuration should start the OS-9 ROM
    without user action. I exspect so see the OS-9 boot menu:

    | OS-9/68K System Bootstrap
    |
    | BOOTING PROCEDURES AVAILABLE ----- <INPUT>
    |
    | Boot from Teac SCSI floppy drive - <fs>
    | Boot from SCSI(SCCS) hard drive -- <hs>
    | Load Bootfile from ROM ----------- <lr>
    | Boot from ROM -------------------- <ro>
    | Reconfigure the boot system ------ <rc>
    | Restart the system --------------- <q>
    |
    | Select a boot method from the above menu:

    Very often (estimated 90%) I get this rubbish instead of the boot menu:

    | x~xxxxxxx~x~xs~~@<~`~xfx~~x~`f`x@sf~`
    | f~~@|f~~@|xxxxxxxfffxx83x<~`xx~xf~~`@|~~~@xO~p~xxx~~xxxx~xx<3~`@<x~
    | x~~~~
    | ~~~xxxxx
    |
    | x
    | ~~xxxxx0~xx<`xxs`xxx~@<~@xxxxxxxxxxxxx<xx<~`@xx~xBs@xxxxxxxxxxxxxx
    | xxxxxxxxx~xxf~
    ...

    When typing I get an echo, but not what I typed in. So I tried to test:
    - other terminal: the same rubbish
    - other baud rate: no change from 1200Baud to 38400Baud (I tried just
    the typical baud rates)
    - changed the EPROMs because they are old and sometimes they seem to
    forget a bit, or two ... I burnt a new set.
    - changed the NVRAM chip: I know that some of them are dead (battery),
    but I don't have new ones, just some other which seem to work fine
    - the power supply should be good, if not the voltage monitoring would
    set the error LEDs.

    Are there any ideas to locate and fix the error(s)?

    I want to run the Motorola 167bug and OS-9 V3.0(.3).

    TIA, Ralf

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Marc@21:1/5 to All on Wed Aug 10 18:36:32 2016
    Le 08/08/2016 à 18:51, Ralf Kiefer a écrit :
    Hi,

    there are some problems to start some(!) MVME167.

    - the power supply should be good, if not the voltage monitoring would
    set the error LEDs.


    TIA, Ralf


    Hi Ralf,

    Could you check the ripple voltage on the +5V rail with a scope?

    Marc

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Ralf Kiefer@21:1/5 to Marc on Thu Aug 11 18:08:13 2016
    Marc wrote:

    Could you check the ripple voltage on the +5V rail with a scope?

    Yes, I did. Worst case when running OS-9 and "dchecking" the harddisk:
    +/-30mV. After reset running 167-Bug and Microware ROM without activity
    of the harddisk or other VMEbus cards: +/-20mV. This seems to be
    acceptable. The power supply is a little bit oversized :-) Actually the
    cards consume approx. 9A @5V, the power supply can deliver up to 30A @5V
    (18 years ago, but moderate "mileage").

    The scope was connected to Vcc and GND of the 48T08 (NVRAM).

    Meanwhile I examined 15 MVME167s: most of the 48T08 internal batteries
    are dead. I opened the first one and installed an external fresh
    battery. There is no difference of this rubbish send to my terminal
    depending on the state of the NVRAM chips.

    But probably this is a "regular" timing problem of the Microware
    software, means an old problem. When I plug in the Microware ROMs into
    sockets U1 and U2 without Motorola's 167-Bug the Microware code
    initializes the I/O in a correct way. I don't see the rubbish in this
    case.

    And there seems to be another problem at the interface between Motorola
    code and Microware code. Nearly every time after cold or warm reset the Motorola code starts twice.

    Conclusion: I will live with that situation. If I need a system booting
    save after reset I will insert just the Microware code without
    Motorola's 167-Bug.

    Ralf

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Ralf Kiefer@21:1/5 to bkorosin@gmail.com on Tue Aug 30 22:53:40 2016
    <bkorosin@gmail.com> wrote:

    Did you try running the board with J1 jumper on pins 3-4 removed to force
    default setup / operation parameters in ROM? Alternatively you can also
    depress RESET and ABORT switches at the same time.

    I didn't changed jumpers. The configuration in NVRAM is ok. By pressing
    the Abort switch after pressing the Reset switch it will influence just
    the Motorola software (167-Bug) but not the Microware part.

    The same with the MVME177 runs good. It's nearly the same software in
    the ROM/FLASH. And I found a note (my own one :-) ) from 1997 which
    mentioned this problem at that time too.


    Then use cnfg;m and env;d to rewrite the configuration of the board.

    No problem with that. If there is trouble with the battery of the 48T08
    and the NVRAM lost its content 167-Bug and Microware bootstrap tell
    something like "CRC error". My terminal (a Mac) can send a proper
    S-record file to fill the NVRAM instead of typing lots of bits and bytes
    to configure the NVRAM manually.

    Ralf

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From bkorosin@gmail.com@21:1/5 to Ralf Kiefer on Sun Aug 14 09:24:32 2016
    Hi Ralf,

    Did you try running the board with J1 jumper on pins 3-4 removed to force default setup / operation parameters in ROM? Alternatively you can also depress RESET and ABORT switches at the same time.
    Then use cnfg;m and env;d to rewrite the configuration of the board.

    Best regards,
    Bo/

    On Monday, August 8, 2016 at 6:51:06 PM UTC+2, Ralf Kiefer wrote:
    Hi,

    there are some problems to start some(!) MVME167.

    EPROM sockets U1 and U2 are populated with the Motorola 167Bug, the
    sockets U3 and U4 with the Microware stuff running OS-9/68k V3.0. They
    are marked as "1.01" probably by Dr. Keil in Germany in the 1990s.

    After reset:

    | Copyright Motorola Inc. 1988 - 1994, All Rights Reserved
    |
    | MVME167 Debugger/Diagnostics Release Version 2.2 - 01/14/94
    | COLD Start
    |
    | Local Memory Found =06000000 (&100663296)
    |
    | MPU Clock Speed =33Mhz
    |
    | ROMboot in progress... To abort hit <BREAK>

    That's what I exspect. The configuration should start the OS-9 ROM
    without user action. I exspect so see the OS-9 boot menu:

    | OS-9/68K System Bootstrap
    |
    | BOOTING PROCEDURES AVAILABLE ----- <INPUT>
    |
    | Boot from Teac SCSI floppy drive - <fs>
    | Boot from SCSI(SCCS) hard drive -- <hs>
    | Load Bootfile from ROM ----------- <lr>
    | Boot from ROM -------------------- <ro>
    | Reconfigure the boot system ------ <rc>
    | Restart the system --------------- <q>
    |
    | Select a boot method from the above menu:

    Very often (estimated 90%) I get this rubbish instead of the boot menu:

    | x~xxxxxxx~x~xs~~@<~`~xfx~~x~`f`x@sf~`
    | f~~@|f~~@|xxxxxxxfffxx83x<~`xx~xf~~`@|~~~@xO~p~xxx~~xxxx~xx<3~`@<x~
    | x~~~~
    | ~~~xxxxx
    |
    | x
    | ~~xxxxx0~xx<`xxs`xxx~@<~@xxxxxxxxxxxxx<xx<~`@xx~xBs@xxxxxxxxxxxxxx
    | xxxxxxxxx~xxf~
    ...

    When typing I get an echo, but not what I typed in. So I tried to test:
    - other terminal: the same rubbish
    - other baud rate: no change from 1200Baud to 38400Baud (I tried just
    the typical baud rates)
    - changed the EPROMs because they are old and sometimes they seem to
    forget a bit, or two ... I burnt a new set.
    - changed the NVRAM chip: I know that some of them are dead (battery),
    but I don't have new ones, just some other which seem to work fine
    - the power supply should be good, if not the voltage monitoring would
    set the error LEDs.

    Are there any ideas to locate and fix the error(s)?

    I want to run the Motorola 167bug and OS-9 V3.0(.3).

    TIA, Ralf

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)