Re: (Fwd) Your B Series 8088 board reversal

From: Robert \ <kebernet_at_gmail.com>
Date: Sun, 5 Dec 2010 21:52:59 -0500
Message-ID: <AANLkTikP+dpWREKpQsgrh1UbGxJnXHUeEYAUVvrEc3zh@mail.gmail.com>
That is a fun factoid!

On Sun, Dec 5, 2010 at 2:46 AM, <Ruud@baltissen.org> wrote:

> Hallo allemaal,
>
>
> I just found this email, very interesting IMHO:
>
>
> ------- Forwarded message follows -------
> Send reply to:  <drshock@insectria.net>
> From:   "Edward Shockley" <drshock@insectria.net>
> To:     <ruud@BALTISSEN.ORG>
> Subject:        Your B Series 8088 board reversal
> Date sent:      Sat, 4 Dec 2010 20:55:19 -0500
>
>    Ruud,
>    Came across your reversal of Bo's 8088 board at
>    http://www.zimmers.net/anonftp/pub/cbm/schematics/computers/b/8088.txtand the
>    GIF of your schematic.
>
>    Have been doing a project where I wanted to get the 8087 coprocessor to
> work on this
>    board for some MS-DOS 1.25 software I felt like porting over. It never
> worked even back
>    in the day but nobody bothered to figure out why. Anyway, wanted to
> share a fix for a PCB
>    error that causes the 8087 to hang CP/M-86 or MS-DOS on bootup.
>
>    In your schematic you show pins 31 and 33 of the 8087 tied together.
> This is the error that
>    CBM made. The board has a silkscreen for "JP1" in the valley between the
> 8088 and 8087
>    but no PCB jumper was ever installed and only one solder pin hole was
> drilled for it. This
>    hard link between these two pins must either be cut, or cut and another
> hole drilled and
>    the jumper installed for the math coprocessor to work. It would be
> jumpered when not
>    installed, and the jumper removed if installed.
>
>    The 6509, as you noted, requests the bus from the 8088 across the RQ/GT0
> line. But only
>    when no 8087 is installed. When an 8087 is added to the circuit the 6509
> must instead
>    (with the design point the Commodore engineer(s) decided upon) request
> the bus thru the
>    8087 pin 33 (RQ/GT1) and let the 8087 pin 31 be connected to the 8088
> RQ/GT0.
>    Connected this way the 8087 will either request the bus for itself, or
> on behalf of the 6509
>    as necessary. The RQ/GT1 line of the 8088 remains non connected as
> original.
>
> If the fix isn't made then when the 8087 is installed the 6509 and
> the 8087 (during initilization of either CP/M-86 or MS-DOS) will
> both try to request the bus from the 8088 on the same line and this
> will lock the bus up and hang the B.
>
> With this change the 8087 can be installed and function as
> originally designed without problem - though with it added the
> board draws over 950mA!
>
>
> FYI,
>
> Edward Shockley
> http://www.insectria.org/b128.html
>
> ------- End of forwarded message -------
>
> --
>    ___
>   / __|__
>  / /  |_/     Groetjes, Ruud Baltissen
>  \ \__|_\
>   \___|       http://Ruud.C64.org
>
>
>
>
>
>
>
>       Message was sent through the cbm-hackers mailing list
>



-- 
:Robert "kebernet" Cooper
::kebernet@gmail.com
Alice's cleartext
Charlie is the attacker
Bob signs and encrypts
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x9E8759F8


       Message was sent through the cbm-hackers mailing list
Received on 2010-12-06 03:00:30

Archive generated by hypermail 2.2.0.