Re: Help with a 8050 drive

From: Ethan Dicks (
Date: 2007-05-14 16:32:21

On 5/14/07, Mark Zablocki <> wrote:
> Hi Ethan,
> I'm using new Athana 5.25 disks with the 8050. They should work fine.

I don't know that brand.  What are the published specs for it
(coersitivity of the media, or at least the marketing "DSDD" or "DSHD"
letters)?   If these disks are for PCs, it would be nice to eliminate
the media as a source of problems.

> I found the following information in the 8050 Troubleshooting Guide at
>      PROBLEM: drive appears to load but won't run.
>      SOLUTION: check 6532s and UK3 controller ROM.
>      PROBLEM: drive won't write/format.
>      SOLUTION: check write protection tab in disk and switch on drive.
>                      check write circuitry on analog board.
>                      check write enable on digital board.
> Unfortunately, I have no idea where these circuits are located or how to check them. I don't even know where to find parts.

the 6532s are just that - 40 pin parts on the drive's main board with
a part number of "6532" stenciled on the chip.  The UK3 controller ROM
is a ROM chip (24 or 28 pins) at the position marked "UK3".  As for
"write circuitry" and "write enable", you are going to need the
schematics ( and tools like a multimeter and/or an
oscilloscope.  A continuity tester might help with the write
protection switch, but I think you would get an immediate error trying
to format a disk with the write protect circuit "on".  You would not
get an immediate error if there were a fault in the write circuity on
the analog board, but the format would still fail.  You really should
see if you can get a known-working 8050 disk to read-test.  That would
at least let you know that large portions of the drive are working.

As for parts, well... ROMs can be re-burned with EPROMs and you can
use pin adapters to use modern chips.  replacing a 6532 is a little
tougher.  An analog fault on the analog board (bolted to one of the
drives) should be easy to get parts for, once you determine what has
failed.  The only source I can imagine now for whole boards would be
another 8050, since nobody has stocked spare boards for PET drives in
a long time.

> I appreciate any and all help.

If you don't have component-level debugging tools, you are going to
find it hard to diagnose component-level faults.  Back in the day, a
tech would put the drive on the bench, hook up a computer, issue some
test commands (or run pre-fab C= diagnostic software) to make the
drive read continuously, or write continuously, while poking at the
boards with an oscilloscope.  Eventually, that got too expensive, so
they just swapped boards and sent dead boards back to a repair depot
where a tech spent a few minutes on the board to see if it was easy to
fix or not.  Easy ones got fixed; hard ones got scrapped.

There is no simple one-size-fits-all set of procedures to fix a disk
drive - you have to start from what works and narrow down the search
until you identify the specific part that's failed.  When it's more
than one part, the job is harder.  Fortunately, in your case, you seem
to be able to communicate with the IEEE CPU, so you can try to issue
specific commands to the drive to see what works and what does not.
If this were a 1541, the information is out there to tweak things at a
very specific level (like being able to write specific bad tracks to
replicate copy protection schemes), but, unfortunately, the same level
of detail is not available for the PET drives as for the C-64.

To sum up - try to find a known good disk.  If it won't read, you'll
have to look for the fault in a different place than if the drive
reads but won't write.  If you have to turn to someone on the list for
a disk, try getting 2 - one to try reads, the other to try formats.
See if you are having a media problem or a drive problem.  All 5.25"
disks are _not_ alike and _not_ interchangable.  You have to match the
media type for the drive model.  It's not as simple as with 3.5"
drives (where there are, essentially 3 types - commonly called "720K",
"1.44MB", and "2.88MB", though those names are somewhat inaccurate
from a technical standpoint).


       Message was sent through the cbm-hackers mailing list

Archive generated by hypermail pre-2.1.8.