765A's can be had from JDR MicroDevices for
1.29 USD, in stock and available in 40 pin. Not
uncoincidentally, they have it in their 'IC/Z80 &
Support' section. And in *40 pin* (gasp!). I do
not work for them; I am a satisfied customer.
They were my one-stop shopping source when
I got bit by the Z80 bug. They have IMHO a
pretty decent selection of microprocessors
from yesteryear, and while the prices aren't
bargain-basement, I do like their IC section's
organization. Well, enough lip-service.


>I once had to read a Floppy with 128 Bytes in DD,
>which would have been impossible on the 765.

I am confuzzled. The spec line on the 765 from
JDR lists it as an 8MHz SS/DD FDC. Re-reading
my doc's on the specific chip I have, it seems like
it will read 128-byte sectors, but only in FM mode.
I'm no expert on the subject, so please feel free to
edumacate me. If using the 765 is going to cramp
our style in any way, then I am for using the WD
chip, esp. in light of Don's revelation of their ]
availability from BG Micro (I've done business w/
them, too, and was pleased).

As I stated, I'm no expert on FDC's- in fact,
part of my motivation for this project is to
learn about such things from the very basest
level- and frmware just doesn't get much more
basic. I say let the debate begin- 765 vs WD17xxx
(what is 27xxx? upgrade/advanced features?).
Pro's and Con's of each.I'm off to research the WD
device(s).

TTFN,
Tarkin

On Sun, 16 Apr 2006 09:57:35 -0500, wild bill
wrote:
>
>There is no better Z-80 implementation anywhere
>than the BB II. It has all the support chips: the CTC chip;
>the DMA chip; the PIO chip; the full blown SIO chip,
>and is the classic 'everything in the Intel databook
>implementation' with the addition of a SASI interface
>and the ability to program it's own BIOS ROMs.
>

I think I'm confusing it with a Bonneville stock market
data receiver/decoder I took apart some time ago.

If it matters, I just hauled out one, and I don't see a
PIO chip anywhere. And, there are TWO of the CTCs.
Ferguson used TTL for the assorted parallel I/Os.

And, there is a 6845 video controller, which might
be of interest to some ...

Oh, and Allison - I think we were talking Z-80, not 8080.

S-100? gimme a break! A proper Z-80 would be all
on one board, because the BUS was not designed
to allow spreading the support chips across it.

The interrupt support in S-100 is little more than
putting a garage door opener on a commode.

The Z-80 support chips were DESIGNED to allow
interrupt driven, higher performance operation.

Not that many people ever WROTE anything that way.

If all you want is a cheap Z-80, get a Timex Sinclair.

Bill

I have a timex-sinclair. It's pretty cool. But
I don't feel like I can get at it's guts and do
some real I/O without knackering it right
proper.

CP/M, striclty speaking, was an 8080 OS,
the 68000, Z8000, and 8086 vs's notwithsatanding.

The various 'aftermarket' CCP's and such all
leveraged some aspect or other of the Z80
beyond what features the traditional 8080
CP/M offered.

I think that there is a real possiblity here of
creating a project that will offer a wide range
of challenge possiblities- from the
'ho-hum another Z80 board w/ CP/M' to
a complex, CPLD-boosted microcomputer
leveraging the full power of the Z80.

My personal desire is to is help to contribute
to a design which is flexible enough to encompass
that broad spectrum.

I really do like the P112's design; it's simply
out of my budget at the moment. I do have
a bunch of Z80 and related parts, though;
so I'd like to make something and contribute
to it's design. Er, that is, I am/was planning
on build ing a Z-80/CP/M beastie anyway;
why not collude with others who are doing
the same thing?

And please, do tell me more about that 6845
VC chip......

p.s. Did anyone have anything to contribute
on the FDC issue??

TTFN,
Tarkin

On Tue, 25 Apr 2006 19:34:01 -0500, wild bill
wrote:

>On Sun, 16 Apr 2006 09:57:35 -0500, wild bill
>wrote:
>>
>>There is no better Z-80 implementation anywhere
>>than the BB II. It has all the support chips: the CTC chip;
>>the DMA chip; the PIO chip; the full blown SIO chip,
>>and is the classic 'everything in the Intel databook

Intel had nothing to do with Z80.

>>implementation' with the addition of a SASI interface
>>and the ability to program it's own BIOS ROMs.
>>
>
>I think I'm confusing it with a Bonneville stock market
>data receiver/decoder I took apart some time ago.
>
>If it matters, I just hauled out one, and I don't see a
>PIO chip anywhere. And, there are TWO of the CTCs.
>Ferguson used TTL for the assorted parallel I/Os.
>
>And, there is a 6845 video controller, which might
>be of interest to some ...
>
>Oh, and Allison - I think we were talking Z-80, not 8080.

Of course what you describe is almost a Kaypro.
Kaypro, Z80, 2SIO, 6845, FDC and PIO on one board.

The amproLB and the a few others I referred to are z80 and
better yet Z180(64180). Thats a limited subset of examples
of Z80 based systems that were very good.

AmproLB+ had Z80, SIO, CTC, FDC and SCSI on a far smaller
footprint than the BB. It fit on top of a 5.25" floppy!


>S-100? gimme a break! A proper Z-80 would be all
>on one board, because the BUS was not designed
>to allow spreading the support chips across it.

Yes, there were complete systems on a board using Z80, 128k
ram, SIOs, FDC on S100 too. The upside was you could extend
to multiple CPUs or add tons of IO as needed.

>The interrupt support in S-100 is little more than
>putting a garage door opener on a commode.

You really missed the point. All the bus did is provide 10
lines specifically for interupt interface and control. Interrupts
were a hardware issue on the boards used. Many
had maskable vectored interupt systems some with Z80
peripherals and more did it without using plain TTL.

However, most software never used it or if they did it was in
a trivial way.

>The Z-80 support chips were DESIGNED to allow
>interrupt driven, higher performance operation.

At very high cost.

I've implmented the Z80 vectored (mode2) with just a bit of
TTL and non-Zilog IO. I've also modified more than a few
Z80 boards (trivial mod) so that the existing 8080 style
vectored interrupt (RSTn) hardware could be used
for Mode 2.

I'll never argue against Z80 mode 2 interrupts. That was one of the
more useful improvents to the base 8080 of it's origin. For CP/M
having interrupts that were not vectors to the bottom of ram were
more useful. While the 8080/8085 (and Z80!) can use the 8259
vectored interrupt controller. Though anyone that programmed
that part knows it's a pain to use for small systems. However if
you need more than a few interupts than a few the cascadeability
is hard to beat.

Z80 DMA, at best an ok part. It's biggest issue is 4mhz was the
fastest. That's painful considering Z80s faster than 4mhz were
available. Many designs used the 8257 or 8237 as they were
faster and flexible enough while cheaper and more available.

>Not that many people ever WROTE anything that way.

That was also a problem.

>If all you want is a cheap Z-80, get a Timex Sinclair.

Toy, fun to play with.


Allison

On 25 Apr 2006 19:11:59 -0700, "Tarkin" wrote:

>I have a timex-sinclair. It's pretty cool. But
>I don't feel like I can get at it's guts and do
>some real I/O without knackering it right
>proper.
>
>CP/M, striclty speaking, was an 8080 OS,
>the 68000, Z8000, and 8086 vs's notwithsatanding.
>
>The various 'aftermarket' CCP's and such all
>leveraged some aspect or other of the Z80
>beyond what features the traditional 8080
>CP/M offered.

I have to agree, and worthwhile improvents at that.

>I think that there is a real possiblity here of
>creating a project that will offer a wide range
>of challenge possiblities- from the
>'ho-hum another Z80 board w/ CP/M' to
>a complex, CPLD-boosted microcomputer
>leveraging the full power of the Z80.

Closest I've come to that is multiple Z80s on S100.
Or using slaves like 8741/2 to take tasks from the
z80.

>My personal desire is to is help to contribute
>to a design which is flexible enough to encompass
>that broad spectrum.

Everyone has a valid opinion on what that should be.
Thats both good and bad as the directions you can go are
oppsites!

>I really do like the P112's design; it's simply
>out of my budget at the moment. I do have
>a bunch of Z80 and related parts, though;
>so I'd like to make something and contribute
>to it's design. Er, that is, I am/was planning
>on build ing a Z-80/CP/M beastie anyway;
>why not collude with others who are doing
>the same thing?

I always thought the P112 was cool but in some repects
it was a closed system where adding IO was going to be
harder.

>And please, do tell me more about that 6845
>VC chip......

Popular, I don't use it.

>p.s. Did anyone have anything to contribute
>on the FDC issue??

Memory lapse? What FDC issue. Local (here in the usa)
suppliers have 765s and I've seen WD1793s too. Bluntly,
both require a lot of TTL (or the smc9229 instead) to make
them go.

Opinion:
If you have atleast one system that satisfies the floppy needs
then just wite or find some serial file transfer code. And forget
a floppy.

A CF card is cheaper and easier to implement
once your past the socket. I've found over time while
floppy is handy for old file recovery it was a pain in 1976
(size and speed) and is still a pain (size and speed) and all
systems I build now floppy is omitted as I can serial copy files
from existing floppy based Z80s or PCs. Booting is handeled
with big (32kb or larger) EEproms or similar as it's easy to put
CP/M in a rom and copy it to ram rather than depend on a
floppy(or other disk).

CPLDs/FPGAs are cool devices but mounting and programming
are issues most people will not want to tackle.

A modern styled CP/M system can be based on 40pin Z80s
and still be open enough to modify without having construction
or parts constraints.


Allison

>Closest I've come to that is multiple Z80s on S100.

I know. I read a thread a while back, before 'subscribing'
to this group, where you talked about programming for
both co-processing and shared-memory access w/
8-bit(IIRC) processors. Compelling stuff!

>Thats both good and bad as the directions you can go are
>oppsites!

True enough. I like to think of it as sort of like
Monster Garage; taking bits, peeling them back
to bare iron, and then modding it to your heart's
desire.

I'll say this again: I believe that if the design is
both simple and elegant, with specific I/O
ports and/or mem. addy's reserved for specific
functions, the 'black-box' approach should
work for the glue logic. If you're a poor hick
like me, you use TTL chips harvested from
the junk-o-copia and spares bin (and R*dio
Sh*ck, that 7-11 for your component needs,
they're unloading all sorts of IC's for a
comparative bargain). If you're an EE type,
or whiz-bang college student, or a have
a bigger hobby budget(not just $$, but Tau,
which in some equations = $$), you do it
in a CPLD, and use the rest of the macro-cells/
gates/whatever for your own purpose, for
example cascading to another device for video.

>Memory lapse? What FDC issue
>>As I stated, I'm no expert on FDC's- in fact,
>>part of my motivation for this project is to
>>learn about such things from the very basest
>>level- and frmware just doesn't get much more

>>basic. I say let the debate begin- 765 vs WD17xxx

>>(what is 27xxx? upgrade/advanced features?).
>>Pro's and Con's of each.I'm off to research the WD
>>device(s).

My main beefs w/ the 765 is this nugget from the
NEC 765A spec sheet (the chip I have):
"The µPD765B does not require synchronisation
between the CLK and WCLK inputs."
The CLK I've used successfully (on another project,
not floppy-related) was a 3.59xxxMHz
jobby, most likely intended for TV use. I was wondering
how in blazes I was going to sync it w/ a 250KHz or
500KHz!! Or is it as simple as just gettting a new integral
(i.e. 4.00MHz) CLK.

>Booting is handeled with big (32kb or larger) EEproms

That's what I did on the my other project- a 256-KB
multi-sector FLASH jobbie scavenged from a dead
mobo- worked great. Tied the upper unused bits
lo for testing, eventually planned for some latching[1]
scheme to use it as a sort of flash-ram-disk.

>A CF card is cheaper and easier to implement
>once your past the socket.

I've considered that. I also have a 32 or 64 MB
disk-on-chip, and wondered if it didn't beg to
be used on this sort of project. Would like
to see some more data on the CF idea.

I rigged my nano-system to boot from
the FEEPROM; it copied the relevant
code+data into the upper 32KB of RAM,
which included copy-down code, jumped
to the upper 32K, did the I/O voodoo to swap
in the lower 32K of ram, and the copy-down
code copied the actual intended executable
code+data back down to 0000h, RST 0,
and voila! Sticking with that cumbersome model
has but one advantage: it doens't matter
what's providding the data in the lower 32K (
or whatever size rams one uses) on boot;
just copy high, bank in the ram, copy low,
RST 0, and presto!
[1]:which was how I programmed it- using 3 latches,
a 2-to-4 line decoder, and some inverters, and a
PC parallel port. Not the simplest design, but
it worked, and could program every bit. I learned
a lot from that project.

>Opinion:
>If you have atleast one system that satisfies the floppy needs
>then just wite or find some serial file transfer code. And forget
>a floppy.

I'll take that opinion and use it to make my case for
tty/SIO/SCC as a necessity. Floppy seems like
ars gratia ars, or just a plain ol' pain in the ars.

> but mounting and programming
>are issues most people will not want to tackle.

Which also has prevented me from working
w/ surface mount.

Thanks for you're input Allison!!

TTFN,
Tarkin

Allisonnospam@nouce.bellatlantic.net wrote:

> I always thought the P112 was cool but in some repects
> it was a closed system where adding IO was going to be
> harder.

Closed? Last time I checked, everything about it is freely
redistributable. The OS (ZSDOS) is GPLed. I don't know how design
files can be GPLed, but Dave Brooks told me he wants them distributed in
a similar spirit.

--
David Griffith
dgriffi@cs.csbuak.edu <-- Switch the 'b' and 'u'

On 26 Apr 2006 20:42:45 -0700, "Tarkin" wrote:

>>Closest I've come to that is multiple Z80s on S100.
>
>I know. I read a thread a while back, before 'subscribing'
>to this group, where you talked about programming for
>both co-processing and shared-memory access w/
>8-bit(IIRC) processors. Compelling stuff!
>
>>Thats both good and bad as the directions you can go are
>>oppsites!
>
>True enough. I like to think of it as sort of like
>Monster Garage; taking bits, peeling them back
>to bare iron, and then modding it to your heart's
>desire.
>
>I'll say this again: I believe that if the design is
>both simple and elegant, with specific I/O
>ports and/or mem. addy's reserved for specific
>functions, the 'black-box' approach should
>work for the glue logic. If you're a poor hick
>like me, you use TTL chips harvested from
>the junk-o-copia and spares bin (and R*dio
>Sh*ck, that 7-11 for your component needs,
>they're unloading all sorts of IC's for a
>comparative bargain). If you're an EE type,
>or whiz-bang college student, or a have
>a bigger hobby budget(not just $$, but Tau,
>which in some equations = $$), you do it
>in a CPLD, and use the rest of the macro-cells/
>gates/whatever for your own purpose, for
>example cascading to another device for video.
>
>>Memory lapse? What FDC issue
>>>As I stated, I'm no expert on FDC's- in fact,
>>>part of my motivation for this project is to
>>>learn about such things from the very basest
>>>level- and frmware just doesn't get much more
>
>>>basic. I say let the debate begin- 765 vs WD17xxx
>
>>>(what is 27xxx? upgrade/advanced features?).
>>>Pro's and Con's of each.I'm off to research the WD
>>>device(s).
>
>My main beefs w/ the 765 is this nugget from the
>NEC 765A spec sheet (the chip I have):
>"The µPD765B does not require synchronisation
>between the CLK and WCLK inputs."
>The CLK I've used successfully (on another project,
>not floppy-related) was a 3.59xxxMHz
>jobby, most likely intended for TV use. I was wondering
>how in blazes I was going to sync it w/ a 250KHz or
>500KHz!! Or is it as simple as just gettting a new integral
> (i.e. 4.00MHz) CLK.

Simple and it's not an issue. The internal processor is run off of
CLK and the serialiser (fancy shiftregister) runs off of WCLK.

The only real issue is the HLT and SRT times are derived off
the CLK so if you run them slower the times will be longer than spec
proportionately.

>>Booting is handeled with big (32kb or larger) EEproms
>
>That's what I did on the my other project- a 256-KB
>multi-sector FLASH jobbie scavenged from a dead
>mobo- worked great. Tied the upper unused bits
>lo for testing, eventually planned for some latching[1]
>scheme to use it as a sort of flash-ram-disk.

BTDT, very useful way to go.

>>A CF card is cheaper and easier to implement
>>once your past the socket.
>
>I've considered that. I also have a 32 or 64 MB
>disk-on-chip, and wondered if it didn't beg to
>be used on this sort of project. Would like
>to see some more data on the CF idea.

CF is basicly an IDE semiconductor disk (flash).
The biggest difference is unlike most IDE drives it supports
8bit transfer mode (no need to extend the databus). The only
down side is the connector is very small though there are IDE
to CF adaptors.

>I rigged my nano-system to boot from
>the FEEPROM; it copied the relevant
>code+data into the upper 32KB of RAM,
>which included copy-down code, jumped
>to the upper 32K, did the I/O voodoo to swap
>in the lower 32K of ram, and the copy-down
>code copied the actual intended executable
>code+data back down to 0000h, RST 0,
>and voila! Sticking with that cumbersome model
>has but one advantage: it doens't matter
>what's providding the data in the lower 32K (
>or whatever size rams one uses) on boot;
>just copy high, bank in the ram, copy low,
>RST 0, and presto!
>[1]:which was how I programmed it- using 3 latches,
>a 2-to-4 line decoder, and some inverters, and a
>PC parallel port. Not the simplest design, but
>it worked, and could program every bit. I learned
>a lot from that project.
>
>>Opinion:
>>If you have atleast one system that satisfies the floppy needs
>>then just wite or find some serial file transfer code. And forget
>>a floppy.
>
>I'll take that opinion and use it to make my case for
>tty/SIO/SCC as a necessity. Floppy seems like
>ars gratia ars, or just a plain ol' pain in the ars.

leaving out floppy allows for rapid assembly and test
without the need for floppy power, and space.

>> but mounting and programming
>>are issues most people will not want to tackle.
>
>Which also has prevented me from working
>w/ surface mount.
>
>Thanks for you're input Allison!!
>
>TTFN,
> Tarkin

On Thu, 27 Apr 2006 03:55:01 GMT, dgriffi@cs.csbuak.edu (Dave
Griffith) wrote:

>Allisonnospam@nouce.bellatlantic.net wrote:
>
>> I always thought the P112 was cool but in some repects
>> it was a closed system where adding IO was going to be
>> harder.
>
>Closed? Last time I checked, everything about it is freely
>redistributable. The OS (ZSDOS) is GPLed. I don't know how design
>files can be GPLed, but Dave Brooks told me he wants them distributed in
>a similar spirit.

You misunderstood. Closed as in single board most signals not
available at edge connectors and internal configuration is as is.

I know the idea of closed usually means software and prints
unavailable. However a small board like that is not as hardware
hackable.

Allison

Allisonnospam@nouce.bellatlantic.net wrote:

> On Thu, 27 Apr 2006 03:55:01 GMT, dgriffi@cs.csbuak.edu (Dave
> Griffith) wrote:
>
>>Allisonnospam@nouce.bellatlantic.net wrote:
>>
>>> I always thought the P112 was cool but in some repects
>>> it was a closed system where adding IO was going to be
>>> harder.
>>
>>Closed? Last time I checked, everything about it is freely
>>redistributable. The OS (ZSDOS) is GPLed. I don't know how design
>>files can be GPLed, but Dave Brooks told me he wants them distributed in
>>a similar spirit.
>
> You misunderstood. Closed as in single board most signals not
> available at edge connectors and internal configuration is as is.
>
> I know the idea of closed usually means software and prints
> unavailable. However a small board like that is not as hardware
> hackable.
>
> Allison


Aren't most, if not all, of those internal signals on the expansion
connector? The board layout isn't fixed either, it was changed to support a
different battery.

Full documentation is available.

I wouldn't consider it a closed system.

craigm

On Thu, 27 Apr 2006 06:39:47 -0500, craigm
wrote:

>Allisonnospam@nouce.bellatlantic.net wrote:
>
>> On Thu, 27 Apr 2006 03:55:01 GMT, dgriffi@cs.csbuak.edu (Dave
>> Griffith) wrote:
>>
>>>Allisonnospam@nouce.bellatlantic.net wrote:
>>>
>>>> I always thought the P112 was cool but in some repects
>>>> it was a closed system where adding IO was going to be
>>>> harder.
>>>
>>>Closed? Last time I checked, everything about it is freely
>>>redistributable. The OS (ZSDOS) is GPLed. I don't know how design
>>>files can be GPLed, but Dave Brooks told me he wants them distributed in
>>>a similar spirit.
>>
>> You misunderstood. Closed as in single board most signals not
>> available at edge connectors and internal configuration is as is.
>>
>> I know the idea of closed usually means software and prints
>> unavailable. However a small board like that is not as hardware
>> hackable.
>>
>> Allison
>
>
>Aren't most, if not all, of those internal signals on the expansion
>connector? The board layout isn't fixed either, it was changed to support a
>different battery.
>
>Full documentation is available.
>
>I wouldn't consider it a closed system.
>
>craigm

To a point. The problem of partial address decodes and ability to
disable selected devices were problematic. As a Z80 class single
board computer it was excellent. However my criteria is different as
I may wish to have things like a integrated video subsystem or more
parallel IO and those require a bit different design approach.

When your used to bus based systems (S100, Multibus, and others)
open is more liberal. Subsections can be confined to a board that is
optional. SBCs (single board computers) are often more of an
encapsulated design where it's details are open but more likely to
only be used "as is" to the limits of the design.

While the plans for P112 and it's kin are available and same for code
the design is a SBC system. Expandability is more incidental rather
than designed in and exploited. Yes it's possible to copy parts and
modify it but then you have to know the unique parts used
and their interactions with the design and software. Then again Dave
knows this from reengineering the original P112 for the second
production version. However, as a complete and very small SBC
it's an excellent design.



Allison

Tarkin wrote:
> 765A's can be had from JDR MicroDevices for
> 1.29 USD, in stock and available in 40 pin.
>
>
>
> As I stated, I'm no expert on FDC's- in fact,
> part of my motivation for this project is to
> learn about such things from the very basest
> level- and frmware just doesn't get much more
> basic. I say let the debate begin- 765 vs WD17xxx
> (what is 27xxx? upgrade/advanced features?).
> Pro's and Con's of each.I'm off to research the WD
> device(s).
>
> TTFN,
> Tarkin

Sorry for the long delay in answering the quaestion about the WD279x
series parts.

The WD 1770/1772 were the first generation SD/DD FDC chips by Western
Digital (WD). The WD1771 was a SD FDC chip. The next family of parts
was the WD179x second generation of SD/DD FDC chips. These chips
required an external data seperator when working at the higher data
rates. The last and final series was the WD279x series of parts. These
parts have a built in data seperator as well as a few other
enhancments. If you want to use WD parts then the 279x series is what
you want.

As pointed out the 765 is also available. B.G.Micro is a good source of
the these and several of the WD parts. B.G.Micro also carries the 9216
data seperator chip in an 8-pin dip. If you use the 765 or WD179x
series parts you will also need the 9216.

I've looked for some of the other FDC chips mentioned (National 8473
and SMC 37C665) but can't find any place that has them in stock.

One final note on standard Z80 parts. Again, B.G.Micro seems to have a
very good selection of 2.5, 4 and 6MHz parts.

One note about B.G.Micro. They are a good place to do business with but
they seem to be at the bottom of the components "food chain". In other
words, when places like DIGI-KEY, Jameco & JDR stop carrying an old
part. B.G.Micro starts picking them up. So their your last chance to
get an old part. Word to the wise, If you buy from B.G.Micro, buy a
lot!

Don

I'd like to make a proposal as well.

Goals:

This project needs to be easy to build AND finish, so that more people
will actually do so... something we can mount on a wooden base to get
started with, wire by hand, find parts to, and see fast results from.
Project also needs to be a little different than everything else thus
more interesting to see end results. I argue that doing yet another
Z80 project is just too predictable in outcome. So here's what I would
like to do.

- single board with expansion room (no bus is needed if we have the
main features from the outset)
- 6809 CPU (this will take us into lesser known territory)
- 2 x serial port -- one for terminal, one for data transfers.
- separate graphics output (TMS9928 is simple chip to design around)
please no IBM 6845 vid chip.
- no floppy chip (circuitry too complicated! I want to actually have
time to finish this project!)
- CF storage

Our tasks would then be to translate CPM -> 6809 unless it exists
already, or use a new operating system altogether. Allison already has
a lot of experience in this area.

Now, I know you all are talking Z80 project, but I thought this would
be a fun, new direction to go and maybe some of us can do something
like this, and others can go the Z80 route. But instead of
endlessly debating which is the typical outcome of "committee driven"
projects, some of us who are serious can quickly choose the components
that make sense, keep the project SIMPLE yet COOL, and get our asses in
gear towards having fun.

~ J

js@cimmeri.com wrote:
> I'd like to make a proposal as well.
>
> Goals:
>
> This project needs to be easy to build AND finish, so that more people
> will actually do so... something we can mount on a wooden base to get
> started with, wire by hand, find parts to, and see fast results from.
> Project also needs to be a little different than everything else thus
> more interesting to see end results. I argue that doing yet another
> Z80 project is just too predictable in outcome. So here's what I would
> like to do.
>
> - single board with expansion room (no bus is needed if we have the
> main features from the outset)
> - 6809 CPU (this will take us into lesser known territory)
> - 2 x serial port -- one for terminal, one for data transfers.
> - separate graphics output (TMS9928 is simple chip to design around)
> please no IBM 6845 vid chip.
> - no floppy chip (circuitry too complicated! I want to actually have
> time to finish this project!)
> - CF storage
>
> Our tasks would then be to translate CPM -> 6809 unless it exists
> already, or use a new operating system altogether. Allison already has
> a lot of experience in this area.
>
> Now, I know you all are talking Z80 project, but I thought this would
> be a fun, new direction to go and maybe some of us can do something
> like this, and others can go the Z80 route. But instead of
> endlessly debating which is the typical outcome of "committee driven"
> projects, some of us who are serious can quickly choose the components
> that make sense, keep the project SIMPLE yet COOL, and get our asses in
> gear towards having fun.
>
> ~ J

I can't wait to see where this goes. :-)

Don

On 30 Apr 2006 09:34:22 -0700, "js@cimmeri.com"
wrote:

>
>I'd like to make a proposal as well.
>
>Goals:
>
>This project needs to be easy to build AND finish, so that more people
>will actually do so... something we can mount on a wooden base to get
>started with, wire by hand, find parts to, and see fast results from.
>Project also needs to be a little different than everything else thus
>more interesting to see end results. I argue that doing yet another
>Z80 project is just too predictable in outcome. So here's what I would
>like to do.
>
>- single board with expansion room (no bus is needed if we have the
>main features from the outset)
>- 6809 CPU (this will take us into lesser known territory)
>- 2 x serial port -- one for terminal, one for data transfers.
>- separate graphics output (TMS9928 is simple chip to design around)
>please no IBM 6845 vid chip.
>- no floppy chip (circuitry too complicated! I want to actually have
>time to finish this project!)
>- CF storage
>
>Our tasks would then be to translate CPM -> 6809 unless it exists
>already, or use a new operating system altogether. Allison already has
>a lot of experience in this area.
>
>Now, I know you all are talking Z80 project, but I thought this would
>be a fun, new direction to go and maybe some of us can do something
>like this, and others can go the Z80 route. But instead of
>endlessly debating which is the typical outcome of "committee driven"
>projects, some of us who are serious can quickly choose the components
>that make sense, keep the project SIMPLE yet COOL, and get our asses in
>gear towards having fun.
>


Dave introduced me to 6809 and it's a very good part but it's use of
memory and CP?M while possible isn't totally compatable. Besides
CUBIX exists and for the 6809 environment it's a superior OS. He has
plans and software at his site and it's a very buildable machine by
present standards. Well worth doing and interesting in it's own
right but CP/M on 6809, the only point would be to share text files
and data though Dave does have a good 8080 emulator for it!

Z80 systems. The basic Z80 with a pair of 32kb rams and logic for
a 32K phantom boot rom is a trivial design and forms a core that is
useable for multiple applications. To run CP/M you need user IO
and mass storage. In this day and age that is almost anything
and therein lies the area of greatest debate and problems of
consturction. Where I see somthing interesting is breaking the
mold. The typical CP/M system is one of two forms Kaypro style
(CRT Keybord based box) or P112 where the user interface is
external terminal in some form. I've been playing in the realm
of LCD display and small keyboard with a PAD style input
(up/down/left/right and select buttons) and menues. The idea
is to develop a user interface that is still text based but one
traverses menues (pull downs?) to do routine tasks rather than
entering commands. Something to think about?

Remember CP/M does not demand a lot and it prevents
very little. Handy thing that it's mostly out of the way leaving it
up to the user.


Allison

On 30 Apr 2006 09:47:06 -0700, "Don" wrote:

>
>js@cimmeri.com wrote:
>> I'd like to make a proposal as well.
>>
>> Goals:
>>
>> This project needs to be easy to build AND finish, so that more people
>> will actually do so... something we can mount on a wooden base to get
>> started with, wire by hand, find parts to, and see fast results from.
>> Project also needs to be a little different than everything else thus
>> more interesting to see end results. I argue that doing yet another
>> Z80 project is just too predictable in outcome. So here's what I would
>> like to do.
>>
>> - single board with expansion room (no bus is needed if we have the
>> main features from the outset)
>> - 6809 CPU (this will take us into lesser known territory)
>> - 2 x serial port -- one for terminal, one for data transfers.
>> - separate graphics output (TMS9928 is simple chip to design around)
>> please no IBM 6845 vid chip.
>> - no floppy chip (circuitry too complicated! I want to actually have
>> time to finish this project!)
>> - CF storage
>>
>> Our tasks would then be to translate CPM -> 6809 unless it exists
>> already, or use a new operating system altogether. Allison already has
>> a lot of experience in this area.
>>
>> Now, I know you all are talking Z80 project, but I thought this would
>> be a fun, new direction to go and maybe some of us can do something
>> like this, and others can go the Z80 route. But instead of
>> endlessly debating which is the typical outcome of "committee driven"
>> projects, some of us who are serious can quickly choose the components
>> that make sense, keep the project SIMPLE yet COOL, and get our asses in
>> gear towards having fun.
>>
>> ~ J
>
>I can't wait to see where this goes. :-)
>
>Don

Brings to mind burrows proliferated by genius rodentia.


Allison

"js@cimmeri.com" wrote:
>
> I'd like to make a proposal as well.
>
> Goals:
>
> This project needs to be easy to build AND finish, so that more people
> will actually do so... something we can mount on a wooden base to get
> started with, wire by hand, find parts to, and see fast results from.
> Project also needs to be a little different than everything else thus
> more interesting to see end results. I argue that doing yet another
> Z80 project is just too predictable in outcome. So here's what I would
> like to do.
>
> - single board with expansion room (no bus is needed if we have the
> main features from the outset)
> - 6809 CPU (this will take us into lesser known territory)
> - 2 x serial port -- one for terminal, one for data transfers.
> - separate graphics output (TMS9928 is simple chip to design around)
> please no IBM 6845 vid chip.
> - no floppy chip (circuitry too complicated! I want to actually have
> time to finish this project!)
> - CF storage
>
> Our tasks would then be to translate CPM -> 6809 unless it exists
> already, or use a new operating system altogether. Allison already has
> a lot of experience in this area.
>
> Now, I know you all are talking Z80 project, but I thought this would
> be a fun, new direction to go and maybe some of us can do something
> like this, and others can go the Z80 route. But instead of
> endlessly debating which is the typical outcome of "committee driven"
> projects, some of us who are serious can quickly choose the components
> that make sense, keep the project SIMPLE yet COOL, and get our asses in
> gear towards having fun.
>
I thought that the MC6809 was a *difficult* part to obtain...


--
+----------------------------------------------------------------+
| Charles and Francis Richmond richmond at plano dot net |
+----------------------------------------------------------------+

On Sun, 30 Apr 2006 20:21:03 -0500, Charles Richmond
wrote:

>"js@cimmeri.com" wrote:
>>
>> I'd like to make a proposal as well.
>>
>> Goals:
>>
>> This project needs to be easy to build AND finish, so that more people
>> will actually do so... something we can mount on a wooden base to get
>> started with, wire by hand, find parts to, and see fast results from.
>> Project also needs to be a little different than everything else thus
>> more interesting to see end results. I argue that doing yet another
>> Z80 project is just too predictable in outcome. So here's what I would
>> like to do.
>>
>> - single board with expansion room (no bus is needed if we have the
>> main features from the outset)
>> - 6809 CPU (this will take us into lesser known territory)
>> - 2 x serial port -- one for terminal, one for data transfers.
>> - separate graphics output (TMS9928 is simple chip to design around)
>> please no IBM 6845 vid chip.
>> - no floppy chip (circuitry too complicated! I want to actually have
>> time to finish this project!)
>> - CF storage
>>
>> Our tasks would then be to translate CPM -> 6809 unless it exists
>> already, or use a new operating system altogether. Allison already has
>> a lot of experience in this area.
>>
>> Now, I know you all are talking Z80 project, but I thought this would
>> be a fun, new direction to go and maybe some of us can do something
>> like this, and others can go the Z80 route. But instead of
>> endlessly debating which is the typical outcome of "committee driven"
>> projects, some of us who are serious can quickly choose the components
>> that make sense, keep the project SIMPLE yet COOL, and get our asses in
>> gear towards having fun.
>>
>I thought that the MC6809 was a *difficult* part to obtain...

I'll let Dave comment if he notices.

FYI: there is also the Hitachi HD6309 so there are more of them than
most think and they will be found embedded in stuff. Apparently they
are not so scarce.

But then we hear of people saying that Z80s are hard to find.

Allison

>>> This project needs to be easy to build AND finish, so that more people
>>> will actually do so... something we can mount on a wooden base to get
>>> started with, wire by hand, find parts to, and see fast results from.
>>> Project also needs to be a little different than everything else thus
>>> more interesting to see end results. I argue that doing yet another
>>> Z80 project is just too predictable in outcome. So here's what I would
>>> like to do.
>>>
>>> - single board with expansion room (no bus is needed if we have the
>>> main features from the outset)
>>> - 6809 CPU (this will take us into lesser known territory)
>>> - 2 x serial port -- one for terminal, one for data transfers.
>>> - separate graphics output (TMS9928 is simple chip to design around)
>>> please no IBM 6845 vid chip.
>>> - no floppy chip (circuitry too complicated! I want to actually have
>>> time to finish this project!)
>>> - CF storage
>>>
>>> Our tasks would then be to translate CPM -> 6809 unless it exists
>>> already, or use a new operating system altogether. Allison already has
>>> a lot of experience in this area.
>>>
>>> Now, I know you all are talking Z80 project, but I thought this would
>>> be a fun, new direction to go and maybe some of us can do something
>>> like this, and others can go the Z80 route. But instead of
>>> endlessly debating which is the typical outcome of "committee driven"
>>> projects, some of us who are serious can quickly choose the components
>>> that make sense, keep the project SIMPLE yet COOL, and get our asses in
>>> gear towards having fun.

>I'll let Dave comment if he notices.

Ah another chance to post a plug for CUBIX !
(Hey Allison, hows your system coming?)


I have published a small 6809 design to run my CUBIX OS - this has been
getting a bit of interest lately and there are a few other people currently in
the process of building up systems.

Here is a bit more information:

In the mid 80's I designed a little 6809 based computer system, including the
hardware and all software. The operating system is called CUBIX, and may be
of interest to anyone who wants to build up small homebrew computer. Here is
some information to consider:

- The system is designed to give high functionality with minimal hardware - you
can build a completely operational CUBIX system with a fairly small handful of
parts. (I have built working CUBIX systems in a single afternoon).

- CUBIX is designed to be very portable - it can handle pretty much any I/O
devices and disk geometry, and you can even change to new devices or
redefine the disk geometry "on the fly". The system is so portable, that many
system ports can be done from nothing more than the ROM image (I provide
a detailed porting document).

- CUBIX itself occupies the better part of an 8K ROM, and provides a nice file
system, a command line "shell" (including a good selection of built in
commands and "script/batch" files), and over 100 system calls/services.

- There are a few dozen applications and utilities, which include
- All the standard file and system maintenance utilities
- Several editors, including a fairly decent one which operates line-
by line or full-screen visual.
(The system includes a TTY specification subsystem which allows you
to taylor the screen oriented programs to pretty much any terminal.)
- HELP : A vax/vms like help command, which provides interactive
documentation to all commands, applications and utlities.
- ASM : 6809 assembler, and RAID : a fairly powerful 6809 debugger.
- BASIC
- C A complete implementation of my Micro-C compiler (in fact the
6809 was the first target for the compiler).
- FORTH : A very fast FORTH which compiles to directly executable 6809
machine code instead of threaded lists.
- MAPL : A little APL system.
- SIM80 : An 8080 simulator - allowed me to run most of the code I had
written for my Altair on CUBIX
(I included on the disk as examples, an 8080 BASIC
interpreter, and an 8080 CHESS program)
- Lots more...

- The system is quite mature/stable - I used it as my main computer for several
years, and there has been a small base of other users. (At one time I sold
the software as a commercial product).

- The system is very well documented. The main documents are:
- CUBIX system users guide
- CUBIX system programming manual
- CUBIX porting guide
- Separate documents for all of the major applications.
- The distribution "documents" diskette, as a 360k 5.25" disk which contains
nothing but documentation, and it is completely full - not a single free
sector.

- All source code, documentation and bnaries are now freely available.
(Excluding the C compiler which my commercial compilers are still based
on).

- I have provided schematics for a simple CUBIX system using a serial console,
and matching sample drivers.

- I have done a PC based simulator for my original D6809 portable computer,
which allows you to try the system out by running it (full functionality) on
your PC. I also have a command to import/export individual files to/from the
simuilator disk images, and you can use my ImageDisk program to transfer
disk images back and forth between the simulator image files and physical
CUBIX compatible diskettes.

Although it would be considered a "toy system" by modern 32/64 bit standards, a
CUBIX system gives you a pretty powerful computer with minimal hardware, and can
be a very rewarding small project (as a number of people have told me over the
years). It also gives you a very unique system that you won't experience
anywhere else.

I invite/encourage anyone who wants to experiencing building a small/useful
computer "from the ground up" to check out the CUBIX system - Most of the
material mentioned above can be found on my web site (See URL in sig) - go
to the the "Dunfield 6809" computer under "Homebuilts".


>>I thought that the MC6809 was a *difficult* part to obtain...

>FYI: there is also the Hitachi HD6309 so there are more of them than
>most think and they will be found embedded in stuff. Apparently they
>are not so scarce.

6809s can still be found ... and yes, the 6309 is a very good substitute
(get the 63C09 - it runs at 3Mhz - bus speed - equivalent to 12Mhz or so
by other vendors numbers).

Btw -a floppy controller doesn't have to be all that hard ... the one I
include in my design is 765 based (easy to find), and IIRC it's about
10 ICs.

Regards,
Dave

--
dave06a@ Collector of classic pre-PC computer systems.
dunfield. If you have an old 8/16 bit non-PC system in need of a good
com home, please contact me at email address on the left, or
via contact link of this web site:
http://www.parse.com/~ddunfield/museum/index.html