[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference hydra::amiga_v1

Title:AMIGA NOTES
Notice:Join us in the *NEW* conference - HYDRA::AMIGA_V2
Moderator:HYDRA::MOORE
Created:Sat Apr 26 1986
Last Modified:Wed Feb 05 1992
Last Successful Update:Fri Jun 06 1997
Number of topics:5378
Total number of notes:38326

4113.0. "2090A woes" by CSSE32::SMITH (Reality, just a visible imagination?) Thu Sep 13 1990 10:25

Yesterday my hard drive failed to boot! Yeech...  Last night I had some time
to troubleshoot and it look like my 2090A bit the big one.  My system is a
2000 with 2090A and 2 20MB ST506 drives.  

I'm looking for suggestions.  I'm not auto booting (1.2 roms) so I boot my 
floppy, BindDrivers and transfer control to my HD0's (FFS) S:Startup-Sequence
to finish the boot.  Currently I hang the system in the BindDrivers command.
I booted up a normal floppy (no harddrives) and tried issuing the BindDrivers
command from a CLI window, same effect.  Normally when BindDrivers executes
after a warm boot the drives gronk (drive lite on) for a short time. I get no
noise or lite.

Here's what I did in an attempt to isolate the failing hardware:

1. Recabled/terminated to one drive and tried BindDrivers from CLI, no change.
   I also tried each drive as drive 0 and drive 1 (reset unit number of course).

2. With only one drive, used a new set of cables, no change.

3. Moved the 2090A to another slot, no change.


Assuming I haven't missed anything it looks like I'll need to replace/repair
my 2090A.  Any suggestions on getting it repaired?  Anyone have a 2090 or
2090A they would like to sell or lend so I could verify it is the failing unit?
I set mail to Keith Barrett who listed a 2090 for sale a short time ago.
  
I was really hoping the controller was ok and the one of the drives was bad as 
I would like to upgrade to a larger SCSI drive.  Oh well.

Any help appreciated,

...Ed
T.RTitleUserPersonal
Name		DateLines
4113.1don't give up yetSAUTER::SAUTERJohn SauterThu Sep 13 1990 11:274
    Before you give up on the board try an old DEC field service trick:
    remove the board, clean the contacts, and replace it, making sure it
    is well seated.
        John Sauter
4113.2Pink EraserCSSE32::SMITHReality, just a visible imagination?Thu Sep 13 1990 11:458

Thanks John, forgot to mention,

I tried the old pink eraser trick last nite without success.  I also tried 
reseating the many socket'd chips.  

...Ed ;'(
4113.3MSVAX::BARRETTI must not waste diskspaceThu Sep 13 1990 12:0023
    These are just shots in the dark, but:
    
    
    Does the LED of the controller stay lit when the failure happens?
    If so, I have seen this occur when memory situations occur. Have
    you tried control/D immediately in your startup and doing a manual
    binddrivers command?
    
    Make sure that both the device and the .info file are in the expansion
    drawer -- it will not work if the info file is missing.
    
    If you take the autobooting chip out does it work? (I believe there
    is also a jumper that has to be moved to disable autobooting).
    
    Is your LED shorted out or causing a direct connection?
    
    When you boot the prep disk; does it allow a prep of res0:?
    
    
    
    Keith
4113.4Time to 2090 shop...CSSE32::SMITHReality, just a visible imagination?Thu Sep 13 1990 12:3444
 
Hi Keith,

Looks like I'll be buying your 2090...

  
>    Does the LED of the controller stay lit when the failure happens?
>    If so, I have seen this occur when memory situations occur. Have
>    you tried control/D immediately in your startup and doing a manual
>    binddrivers command?
 
No the LED doesn't lite at all.  It will however lite if I remove the cables
so it is functional.  ^D will not abort the BindDrivers, I've tried.  The only
way to get control of the CLI window is reboot.

   
>    Make sure that both the device and the .info file are in the expansion
>    drawer -- it will not work if the info file is missing.

Both are there, I did have that problem last nite when trying a manual 
BindDrivers and had to copy them to the expansion drawer.
    
>    If you take the autobooting chip out does it work? (I believe there
>    is also a jumper that has to be moved to disable autobooting).

Never had the chips in as I still have 1.2 roms (waiting for 2.0).  I've
always booted DF0:, BindDrivers and switch control to HD0:S/Startup-Sequence.
    
>    Is your LED shorted out or causing a direct connection?

No.
    
>    When you boot the prep disk; does it allow a prep of res0:?

This is tonites step.  I've been reluctent to try this as I will lose my data
and force a backup (which is a little old).  I suspect it won't work since
I do not seem to have any access to the drive.


Thanks for your help,

...Ed
4113.5SDOGUS::WILLIAMSTOPGUNThu Sep 13 1990 12:4510
    Before you try preping res0, find someone (anyone) with a 1.3 2000
    and plug your 2090a into that machine and try booting.  1.3 Kickstart
    solves A GREAT DEAL OF PROBLEMS!
    
    If that doesn't work, then try a new (different) board.
    
    If that doesn't work, then try a prep (yuck).
    
    TOPGUN
4113.62090A brokeCSSE32::SMITHReality, just a visible imagination?Fri Sep 14 1990 09:5731
Well I tried a PREP RES0: last night and of course it doesn't work because
the BindDrivers command must be done first to load the HDdisk driver and this
is where my system hangs trying to access (init?) the drive.  

Re .5

>    Before you try preping res0, find someone (anyone) with a 1.3 2000
>    and plug your 2090a into that machine and try booting.  1.3 Kickstart
>    solves A GREAT DEAL OF PROBLEMS!

Too late ;')  Although 1.3 may solve a lot of problems it's not the reason
my 2090A isn't working.  I have been running this configuration with 1.2
ROMS for 2 years now.  I have neglected to update to 1.3 since autobooting
the hard disk was not a big deal to me.  

If some in the Hudson/Nashua area with a 2000 would be willing to let me plug 
in my 2090A that would be a great test.  As most owners don't like poking around
inside with the hardware I don't expect any offers.


My current plan is to get another 2090/2090A for a reasonable price.  This will
solve my immediate problem.

Has anyone had a PCB repaired?  I would like to get my 2090A repaired and would
appreciate any suggestions as to where to take or send it.  Does CBM have there
own repair shop I could send it to?


Thanks for all the suggestions,

...Ed
4113.72090 fixed it!CSSE32::SMITHReality, just a visible imagination?Mon Sep 17 1990 23:2114
    Well I just finished plugging in a used 2090 (thanks Keith) and Amy is
    back on the air.  Yea!  I had a good look at my 2090A and there is no
    burns, bad solder connections, broken etches, etc. that are visable. 
    Since it's such a solid failure it should be simple to troubleshoot.
    
    Even without a printset I expect I may be able to isolate the failing
    chip by scoping both modules look'n for steady state differences.  One
    problem, now I no longer work in a hardware group I no longer have
    access to a scope.  Anyone in the Nashua area have a scope or access to
    one that I could use for a weekend?  How about access to a Huntitron
    (sp) Tracker?  This would find a stuck at fault quickly.
    
    
    ...Ed