> I'm working on a project that I'd like to have an 'old-looking' computer
> running a device as a sort of 'process controller'. I really want it to
> have a 'blinkenlights' operators panel on it. As I don't have a
> commercially made computer with a suitable frontpanel on it, I was
> thinking about taking something I have and 'modifying' it to be suitable.
Well, one thing would be to define what are the functions of a
Blinkenlights panel. To me it must allow at least start/stop of
the CPU on hardware level and memory modification without help
of the CPU. A nice feature is singlesteping without a monitor
programm, but hard to do for most microprocessors (here comes
the need for ICE hardware). Furthermore a real Blinkenlights
should give you a visual representation of all CPU registers
at all time (if switched on). Again a feature not realy possible
on microprocessors.
Well, I'm biased from working with /370ish machines from the time
before service processors came up :)
> 2) 8-bit ISA interface card and a 'suitably old' PC - Maybe an IBM 5150,
> maybe just a 486 - if I do a 5150, i'll be writing a bunch of assembly and
> avoiding using an OS. If I choose a 386/486 (I have spare 486s) I can
> work on a driver for it in Linux so I have some sort of usable OS
> available.
> 3) Apple //e card. I've got no idea how easy this will be to do - I don't
> really know 6502/apple assembly at all, nor the bus.
> #2 is the least amount of work for me, #3 some more work, #1 the most
> work. Has anyone ever create an 'operators panel' for a PC or Apple 2?
> I've got some ideas on how to do it (will definately require some
> soft/firmware to make it work nicely). Is anyone else interested in
> something similar? I was thinking about making something like a IMSAI
> front panel - using the system's NMI to trigger the software to 'activate'
> the panel.
I'd rather take the #2 aproache, expecialy since the Apple slot
offers all signals needed to capture the bus. The PC is, especialy
with it's modern versions always a bit complex to do. Especialy
hen it comes to DMA (which is AFAIR only possible with support
software - read drivers).
Also the A2 is by far the more simple machine to handle. And you'll
get all CPU signals. All you need is a small sprogramm for single
step and readout of the CPU registers. There is a 256 byte programm
space associated to each I/O card, and 2 K wich can be requested by
the I/O card. Usualy the 256 should be enough for named programm.
furthermore you got some address space for I/O.
If I think along this line, some 'debugging' cards for the Apple
come to mind ...
I rather suggest some kind of abstract interface, where the card
hardware gives their signals on a fixed set of lines, and a pair
of ports where the software pushes a specific data structure. On
the other end of this interface the lines are either connected to
lights by a wrap field (bus signals etc.), and a microprocesser
recives the data structure and displayes it on LED displays
according to some configuration data.
The basic idea is a system dependant part which has to be made up
for each host system and a nice 'Front Panel Box' which can be
used for various systems.
Anyway. If you like to follow this trail, give me a note.
Gruss
H.
--
VCF Europa 3.0 am 27./28. April 2002 in Muenchen
http://www.vcfe.org/
Received on Mon Apr 15 2002 - 08:14:23 BST