Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

An 80386 CPU board and RAM board for the S-100 Bus - possible use for CPM 86 programming

77 views
Skip to first unread message

monah...@gmail.com

unread,
Apr 15, 2013, 10:55:35 PM4/15/13
to
I would like to mention that Andrew Lynch and I have just completed work on the final prototype (V4!) of a new 80386 Master/Slave CPU S-100 bus board. This board is capable of reaching up to36MHz for its clock input when used with RAM on its daughter board (see below).

For those that have been following progress this has been a long project. It utilizes the 16 bit mode of the 80386 to address up to the 16MG of RAM the S-100 bus is capable of addressing. A second daughter board system with two overhead ribbon cable connectors is utilized to allow 32 bit addressing to one or more RAM boards. This "32-Bit S-100 Overhead Bus" as I call it, allows the CPU to run at its maximum speed (no wait states) and with the current static RAM chips in pipeline mode. The current daughter RAM board utilizes 16MG of static RAM. The plan is to next add high capacity DRAM boards.

It turned out that the 80386 splices very nicely into the S-100 bus. The CPU control signals are clean and tolerant. In fact the interface is simplifier than our earlier 8086 and 80286 boards. No bus controller or clock generator chips for example are needed.

The real fun in working with this chip is now one can take advantage of the 32 bit programming mode of the 80386 and its vast memory addressing capability. It took some time to understand how to switch the CPU into “protected mode” and I have written this up to make it easier who may like to start into this world for the first time.

For those with S100 systems that would like to use CPM-86 and get into 32 bit system hardware and software this board may be a great place to start. In the next few weeks we will be sending an order in for a production run of both bare boards. In all lightly-hood this will be a onetime batch/order, so if you think you would like to play around with a board like this in the future, now is the time to order a board. Also I should point out, this board is fairly sophisticated, I would not recommend it for people starting off with S100 boards. They should “cut their teeth” on say the 8088 CPU board first.

For more information about both boards please see here:-
http://s100computers.com/My%20System%20Pages/80386%20Board/80386%20CPU%20Board.htm
For information about 80386 Monitor software please see here:-
http://s100computers.com/Software%20Folder/80386%20Monitor/80386%20Monitor.htm

Steve Nickolas

unread,
Apr 16, 2013, 6:20:37 AM4/16/13
to
On Mon, 15 Apr 2013, monah...@gmail.com wrote:

> I would like to mention that Andrew Lynch and I have just completed work
> on the final prototype (V4!) of a new 80386 Master/Slave CPU S-100 bus
> board. This board is capable of reaching up to36MHz for its clock input
> when used with RAM on its daughter board (see below).
>
> For those that have been following progress this has been a long
> project. It utilizes the 16 bit mode of the 80386 to address up to the
> 16MG of RAM the S-100 bus is capable of addressing. A second daughter
> board system with two overhead ribbon cable connectors is utilized to
> allow 32 bit addressing to one or more RAM boards. This "32-Bit S-100
> Overhead Bus" as I call it, allows the CPU to run at its maximum speed
> (no wait states) and with the current static RAM chips in pipeline mode.
> The current daughter RAM board utilizes 16MG of static RAM. The plan is
> to next add high capacity DRAM boards.
>
> It turned out that the 80386 splices very nicely into the S-100 bus. The
> CPU control signals are clean and tolerant. In fact the interface is
> simplifier than our earlier 8086 and 80286 boards. No bus controller or
> clock generator chips for example are needed.
>
> The real fun in working with this chip is now one can take advantage of
> the 32 bit programming mode of the 80386 and its vast memory addressing
> capability. It took some time to understand how to switch the CPU into
> “protected mode” and I have written this up to make it easier who may
> like to start into this world for the first time.

To me, 386 is mainly useful for its ability to emulate stuff... it might
be possible to use the V86 mode to run normal PC software even though the
hardware does not at all resemble a PC, apart from the CPU.

-uso.

dott.Piergiorgio

unread,
Apr 16, 2013, 8:06:49 AM4/16/13
to
Il 16/04/2013 04:55, monah...@gmail.com ha scritto:
> I would like to mention that Andrew Lynch and I have just completed work on the final prototype (V4!) of a new 80386 Master/Slave CPU S-100 bus board. This board is capable of reaching up to36MHz for its clock input when used with RAM on its daughter board (see below).
>
> For those that have been following progress this has been a long project. It utilizes the 16 bit mode of the 80386 to address up to the 16MG of RAM the S-100 bus is capable of addressing. A second daughter board system with two overhead ribbon cable connectors is utilized to allow 32 bit addressing to one or more RAM boards. This "32-Bit S-100 Overhead Bus" as I call it, allows the CPU to run at its maximum speed (no wait states) and with the current static RAM chips in pipeline mode. The current daughter RAM board utilizes 16MG of static RAM. The plan is to next add high capacity DRAM boards.
>
> It turned out that the 80386 splices very nicely into the S-100 bus. The CPU control signals are clean and tolerant. In fact the interface is simplifier than our earlier 8086 and 80286 boards. No bus controller or clock generator chips for example are needed.
>
> The real fun in working with this chip is now one can take advantage of the 32 bit programming mode of the 80386 and its vast memory addressing capability. It took some time to understand how to switch the CPU into �protected mode� and I have written this up to make it easier who may like to start into this world for the first time.
>
> For those with S100 systems that would like to use CPM-86 and get into 32 bit system hardware and software this board may be a great place to start. In the next few weeks we will be sending an order in for a production run of both bare boards. In all lightly-hood this will be a onetime batch/order, so if you think you would like to play around with a board like this in the future, now is the time to order a board. Also I should point out, this board is fairly sophisticated, I would not recommend it for people starting off with S100 boards. They should �cut their teeth� on say the 8088 CPU board first.
>
> For more information about both boards please see here:-
> http://s100computers.com/My%20System%20Pages/80386%20Board/80386%20CPU%20Board.htm
> For information about 80386 Monitor software please see here:-
> http://s100computers.com/Software%20Folder/80386%20Monitor/80386%20Monitor.htm

why not attempting the development of a CP/M 386 or even a MPM/386 ? :)

Best regards from Italy,
dott. Piergiorgio.

dott.Piergiorgio

unread,
Apr 16, 2013, 8:08:55 AM4/16/13
to
Il 16/04/2013 12:20, Steve Nickolas ha scritto:

> To me, 386 is mainly useful for its ability to emulate stuff... it might
> be possible to use the V86 mode to run normal PC software even though
> the hardware does not at all resemble a PC, apart from the CPU.

hence is feasibile a MPM/386 based on multiple 1M TPA plus a 386 native
BIOS/BDOS/scheduler ?

monah...@gmail.com

unread,
Apr 16, 2013, 9:04:48 PM4/16/13
to
Actually Steve If you use our MSDOS-Support Board
(http://s100computers.com/My%20System%20Pages/MSDOS%20Board/MSDOS%20Board.htm)
and the 8086/80386 Monitor you can run a lot of PC-Software. It boots up for example a MS-DOS (V4.01) either the ZFDC or IDE boards. What is still needed is a VGA display. I have been working on a S100 bus XVGA display board for some time now. Has been quite difficult, but I think it is doable.

monah...@gmail.com

unread,
Apr 16, 2013, 9:06:27 PM4/16/13
to
On Tuesday, April 16, 2013 5:06:49 AM UTC-7, dott.Piergiorgio wrote:
> Il 16/04/2013 04:55, monah...@gmail.com ha scritto:
>
> > I would like to mention that Andrew Lynch and I have just completed work on the final prototype (V4!) of a new 80386 Master/Slave CPU S-100 bus board. This board is capable of reaching up to36MHz for its clock input when used with RAM on its daughter board (see below).
>
> >
>
> > For those that have been following progress this has been a long project. It utilizes the 16 bit mode of the 80386 to address up to the 16MG of RAM the S-100 bus is capable of addressing. A second daughter board system with two overhead ribbon cable connectors is utilized to allow 32 bit addressing to one or more RAM boards. This "32-Bit S-100 Overhead Bus" as I call it, allows the CPU to run at its maximum speed (no wait states) and with the current static RAM chips in pipeline mode. The current daughter RAM board utilizes 16MG of static RAM. The plan is to next add high capacity DRAM boards.
>
> >
>
> > It turned out that the 80386 splices very nicely into the S-100 bus. The CPU control signals are clean and tolerant. In fact the interface is simplifier than our earlier 8086 and 80286 boards. No bus controller or clock generator chips for example are needed.
>
> >
>
> > The real fun in working with this chip is now one can take advantage of the 32 bit programming mode of the 80386 and its vast memory addressing capability. It took some time to understand how to switch the CPU into “protected mode” and I have written this up to make it easier who may like to start into this world for the first time.
>
> >
>
> > For those with S100 systems that would like to use CPM-86 and get into 32 bit system hardware and software this board may be a great place to start. In the next few weeks we will be sending an order in for a production run of both bare boards. In all lightly-hood this will be a onetime batch/order, so if you think you would like to play around with a board like this in the future, now is the time to order a board. Also I should point out, this board is fairly sophisticated, I would not recommend it for people starting off with S100 boards. They should “cut their teeth” on say the 8088 CPU board first.
>
> >
>
> > For more information about both boards please see here:-
>
> > http://s100computers.com/My%20System%20Pages/80386%20Board/80386%20CPU%20Board.htm
>
> > For information about 80386 Monitor software please see here:-
>
> > http://s100computers.com/Software%20Folder/80386%20Monitor/80386%20Monitor.htm
>
>
>
> why not attempting the development of a CP/M 386 or even a MPM/386 ? :)
>
>
>
> Best regards from Italy,
>
> dott. Piergiorgio.

I'm not really a software guy, but I hope others will try this.

Steve Nickolas

unread,
Apr 17, 2013, 5:29:51 AM4/17/13
to
Probably. I think early versions of Concurrent DOS/386 might have been
something similar?

-uso.

Steve Nickolas

unread,
Apr 17, 2013, 5:30:58 AM4/17/13
to
On Tue, 16 Apr 2013, monah...@gmail.com wrote:

> Actually Steve If you use our MSDOS-Support Board
> (http://s100computers.com/My%20System%20Pages/MSDOS%20Board/MSDOS%20Board.htm)
> and the 8086/80386 Monitor you can run a lot of PC-Software. It boots up
> for example a MS-DOS (V4.01) either the ZFDC or IDE boards. What is
> still needed is a VGA display. I have been working on a S100 bus XVGA
> display board for some time now. Has been quite difficult, but I think
> it is doable.

4.01 on a non-PC? O_o I'd think it would have to be very well capable of
emulating significant parts of the PC architecture to be able to run 4.01,
not?

-uso.

Tom Lake

unread,
Apr 20, 2013, 7:08:30 AM4/20/13
to
wrote in message
news:d3234c30-d329-4765...@googlegroups.com...

I would like to mention that Andrew Lynch and I have just completed work on
the final prototype (V4!) of a new 80386 Master/Slave CPU S-100 bus board.
This board is capable of reaching up to36MHz for its clock input when used
with RAM on its daughter board (see below).

Now I need a front panel that has all the register and address lights and
switches for it!

Tom Lake

monah...@gmail.com

unread,
Apr 20, 2013, 1:23:45 PM4/20/13
to
I’m sure you are kidding Tom, but actually on my list is a new SMB (front panel board), the current one is somewhat limited, fine for slow Z80’s etc. but these faster CPU’s need latches for the HEX displays. Also I want to put is a circuit for handling multiple DMA’s and a few other things. That, and a new EPROM/RAM board are two on the list of S-100 board re-does.
One nice feature of the 80386 is the ability to single step, debug. I have a primitive example of this in the 80386 monitor itself.
John
0 new messages