27 Apr
2004
27 Apr
'04
11:25 p.m.
> I'm not as familiar with the 6809 as I am with
the 6502... what differences
> in architecture make it better? Larger stack frame? More (and larger)
> registers?
It also supports more addressing modes than the 6502
like position
independant (indexed off PC) and indirect addressing (in C, this would be
dereferencing a pointer). It also has a fairly orthogonal instruction set
which also helps in compiler writing (in my opinion). Of all the 8bit CPUs,
this happens to be my favorite.
One of the biggest advantages of the 6809 to a compiler is the fact that it does
all kinds of stack relative addressing. The 6502 can't even push/pop all of it's
registers let alone perform relative or indirect stack addressing with them...
This ability makes it very easy to deal with C's stack based local variables.
For example, you can access memory on the stack (or through any index register)
with a 0, 5, 8 or 16 bit offset, and you can use an 8 or 16 bit accumulator to
calculate an offset. You can also perform indirect operations using stack based
pointers, all in a single instruction. Plus, all index and stack registers on the
09 are 16 bits. The 6502's 8-bit index registers are somewhat at odds with C's
notion that a pointer can point 'anywhere'. So, on the 6502, you end up using
zero page 16 bit "registers" a lot, and these involve more overhead to
manipulate.
The 6809 is also my favorite 8-bitter (perhaps that's why there's 6 CoCo's in
my
collecion). The very first edition of my C compiler was targeted to the 6809. At
one point I even designed my own portable 6809 computer (hardware and software),
you can see it (and even try it out with a simulator) at my "old computers"
page:
http://www.parse.com/~ddunfield/museum/index.html
Look for entry called "D6809". In addition to the OS, the system has a
"bunch" of
utilities, several editors, an assembler and a simple implementation of APL. there's
even an 8080 emulator which allowed me to easily bring over code from my Altair which
was my main computing platform prior to this one.
It's disappointing that the 6809 never received as much acceptance or use as it
should
have. It was truly in a class by itself. Motorola documents show two circles, one
containing the words "8-bit" and one containing the words "16-bit". A
third circle,
linking the other two contains the word "M6809", and this is a good description
of the
part. It was an 8-bit CPU with a great deal of 16 bit capability.
Btw, I have scanned the "Motorola M6809-M6809E Microprocessor Programming
Manual" and
a fair bit of other 6809 reference material, which is all available on my site.
Regards,
--
dave04a (at) Dave Dunfield
dunfield (dot) Firmware development services & tools: www.dunfield.com
com Vintage computing equipment collector.
http://www.parse.com/~ddunfield/museum/index.html
28 Apr
28 Apr
5:30 a.m.
Dave Dunfield wrote:
It's disappointing that the 6809 never received as
much acceptance or use as it should
have. It was truly in a class by itself. Motorola documents show two circles, one
containing the words "8-bit" and one containing the words "16-bit". A
third circle,
linking the other two contains the word "M6809", and this is a good description
of the
part. It was an 8-bit CPU with a great deal of 16 bit capability.
In hindsight, the 6809 while the best 8 processer it only has 64k of
memory space. Time has proved the 64k data and 64k program space and
about 1 meg of memory ( something the PC-XT had or PDP11/unix ) was
needed for useful programs with a resonable OS. Also since the only
common 6809 system was by Radio Shack as games machine you never got the
good I/O like lower case letters and a real serial and floppy drives.
Ben.
6:08 a.m.
Rumor has it that ben franchuk may have mentioned these words:
Dave Dunfield wrote:
>It's disappointing that the 6809 never received as much acceptance or use
>as it should
>have. It was truly in a class by itself. Motorola documents show two
>circles, one
>containing the words "8-bit" and one containing the words
"16-bit". A
>third circle,
>linking the other two contains the word "M6809", and this is a good
>description of the
>part. It was an 8-bit CPU with a great deal of 16 bit capability.
I have a couple as well, and I thought it was quite descriptive of the
processor's capabilities. I wonder what kind of neat logo we could come up
with for the Hitachi 63C09... ;-)
In hindsight, the 6809 while the best 8 processer it
only has 64k of
memory space.
So? Most (all?) 8-bit CPUs have that "limitation" -> which can be easily
overcome with bankswitching and other techniques...
Time has proved the 64k data and 64k program space
and
about 1 meg of memory ( something the PC-XT had or PDP11/unix ) was needed
for useful programs with a resonable OS.
1) My CoCo3 has that -- except the program space was limited to about 40K,
due to OS hooks... 512K Memory (rather easily upgraded to 2M, altho one guy
stuffed 64Meg in his -- yes, it's possible!) and with 512K and OS-9, was
(and is) infinitely more useful than *any* 386 I've seen, no matter how
much storage it's stuffed with. My first 386 was a glorified Nintendo, my
CoCo was still my main work machine.
2) Show me a "reasonable OS" for anything Intel based. Then show me the
ones that existed back in '86 or before. Then show me the ones that will
run on anything <'386. Then see if any are left that cost <$150. The list
gets *real* small, *real* quick. OS-9 is a reasonable OS, was first written
in '76, didn't need thousands of patches to support reasonable hardware (4G
Hard Drive support in the OS, back in '82. Show me that on an Intel box...)
Also since the only
common 6809 system was by Radio Shack as games machine you never got the
good I/O like lower case letters and a real serial and floppy drives.
Ben.
Game machine? Hardly... the Commodore & Atari were *much* better with games
than the CoCo ever was. Floppy drives? RS drives were the best available
for that class of market at the time.
Machine Interface Storage
Commodore 19200bps ser. 120K? thereabouts, please correct me.
Atari 19200bps ser. 88K (single) 120K (enhanced)
CoCo 250000bps par. 156K (standard)
720K (OS-9, stock controller)
Heck, the CoCo tape unit is almost as fast as the Commie disk drive, and
could be tweaked to go faster! Also, Later CoCo2s & all CoCo3s had
lowercase builtin... and RS had a cartridge that would give you a full
80x25 mono screen w/lower case (and full decenders, too!) if you needed to
"get down to business"... and my CoCo3's max screen was 106x28 chars...
great for spreadsheets!
How does [[ Crappy graphics, crappy sound, great I/O speeds & Unix-class
multitasking OSs ]] == "Game Machine"???
Hmmm... it seems that your recollection of history & mine are a wee bit
different...
Laterz,
Roger "Merch" Merchberger
--
Roger "Merch" Merchberger | "Profile, don't speculate."
sysadmin, Iceberg Computers | Daniel J. Bernstein
zmerch(a)30below.com |
8:59 a.m.
Roger Merchberger wrote:
Rumor has it that ben franchuk may have mentioned
these words:
Forget this RUMOR crap! I DID SAY IT!
Game machine? Hardly... the Commodore & Atari were
*much* better with
games than the CoCo ever was. Floppy drives? RS drives were the best
available for that class of market at the time.
For a short time I did have a Coco 3, but 1) Could never
get the memory upgade, 2) The monitor for the Coco 3 died
and 3)I never did get OS9/II.
The real computer system for the 6809 was not the COCO Crap
but stuff I am finding out only now from the outdated systems.
The COCO was marketed as home computer, thus games and BASIC
programs were the main market.
Ben.
10:53 a.m.
So? Most (all?) 8-bit CPUs have that
"limitation" -> which can be easily
overcome with bankswitching and other techniques...
Motorola actually sold an MMU chip for the 6809 at one point (I have the
data sheets, but have never seen the chip). It was something of a kludge
-- you could load the new page number and a number-of-cycles into the
chip. After than number of clock cycles, the page changed. That let you
do a jump instruction on the 6809 at the same time, so you ended up
directly at the right address in the new page.
2) Show me a "reasonable OS" for anything
Intel based. Then show me the
ones that existed back in '86 or before. Then show me the ones that will
Dynix on a Sequnet Symmetry?
run on anything <'386. Then see if any are left
that cost <$150. The list
gets *real* small, *real* quick. OS-9 is a reasonable OS, was first written
OK, you win. FWIW, OS-9 was my first 'real OS', and I loved it. I still
think some features, particularly the I/O system, are really well designed.
Heck, the CoCo tape unit is almost as fast as the
Commie disk drive, and
could be tweaked to go faster! Also, Later CoCo2s & all CoCo3s had
What 'later CoCo2s'? That sounds like the 'Deluxe Colour Computer', and
AFAIK that was never officially sold (there are rumoured to be a couple
out there).
lowercase builtin... and RS had a cartridge that would
give you a full
80x25 mono screen w/lower case (and full decenders, too!) if you needed to
I rmember such cartridges being sold in 'The Rainbow' magazine, but I
wasn't aware Radio Shack ever sold them.
How does [[ Crappy graphics, crappy sound, great I/O
speeds & Unix-class
multitasking OSs ]] == "Game Machine"???
'Crappy Sound'? Not if you added the Speech/Sound cartridge. That
contained a microcontoller, AY-3-891x sound chip and SPO256 speech
synthesiser. It was as good as the sound on other machines of the time.
I guess what I disliked most about the CoCo was the physical design. You
rpetty much had to have the Multi-pack if you were doing anything serious
with ti, then you had the disk cotnroller, serial port cartridge, a
couple of homebrew things (look, this is _my_ machine :-)) plugged into
that, etc. Not a very convenient layout. I would hae prefered a PC-style
case (Yes, I know people repackaged their CoCos, but...).
-tony
2:13 p.m.
On Wed, Apr 28, 2004 at 02:08:31PM -0400, Roger Merchberger wrote:
Game machine? Hardly... the Commodore & Atari were
*much* better with games
than the CoCo ever was. Floppy drives? RS drives were the best available
for that class of market at the time.
Machine Interface Storage
Commodore 19200bps ser. 120K? thereabouts, please correct me.
Atari 19200bps ser. 88K (single) 120K (enhanced)
CoCo 250000bps par. 156K (standard)
720K (OS-9, stock controller)
The 1541, Commodore's disk drive for the C-64, was serial, but it was
sync serial based on bit-banging the I/O port at $0000/$0001. It was
not a UART-friendly speed like the Atari drives were. The capacity was
the same as the older IEEE-488 4040 - about 170K (the drive processor
varied the bit rate amongst 4 zones, the lowest of which was identical
to the Apple II drives, packing between 17 and 22? sectors per track,
depending on the track).
-ethan
--
Ethan Dicks, A-130-S Current South Pole Weather at 29-Apr-2004 02:00 Z
South Pole Station
PSC 468 Box 400 Temp -80.5 F (-62.6 C) Windchill -116 F (-82.2 C)
APO AP 96598 Wind 7.8 kts Grid 059 Barometer 679.5 mb (10643. ft)
Ethan.Dicks(a)amanda.spole.gov http://penguincentral.com/penguincentral.html
5:22 p.m.
The 1541, Commodore's disk drive for the C-64, was
serial, but it was
sync serial based on bit-banging the I/O port at $0000/$0001.
No, no -- the IEC serial bus goes through the CIAs. The *cassette* is done
by bitbanging $0000/1.
It was
not a UART-friendly speed like the Atari drives were. The capacity was
the same as the older IEEE-488 4040 - about 170K (the drive processor
varied the bit rate amongst 4 zones, the lowest of which was identical
to the Apple II drives, packing between 17 and 22? sectors per track,
depending on the track).
Hai. :)
--
---------------------------------- personal: http://www.armory.com/~spectre/ --
Cameron Kaiser, Floodgap Systems Ltd * So. Calif., USA * ckaiser(a)floodgap.com
-- "My inner geek can beat up your inner geek." -------------------------------
7:17 a.m.
On Wed, 28 Apr 2004, ben franchuk wrote:
In hindsight, the 6809 while the best 8 processer it
only has 64k of
memory space. Time has proved the 64k data and 64k program space and
about 1 meg of memory ( something the PC-XT had or PDP11/unix ) was
needed for useful programs with a resonable OS. Also since the only
Moore's law observes that size doubles every 18 months.
Boyle's law observes that software expands to fit all available space.
My corollary observes that Boyle's law leads Moore's.
But "needed"???!?
If the meg, or more, of memory is there, then it will be used.
That doesn't mean that it is necessary.
common 6809 system was by Radio Shack as games machine
you never got the
good I/O like lower case letters and a real serial and floppy drives.
I agree that Radio Shack crippled certain aspects of the Coco.
I always assumed that that was to avoid it outperforming their Model II
"business" computer.
But the floppy drives of the Coco were industry standard units,
with an industry standard (SA400) interface, with an industry
standard (WD179x) controller, to produce an industry standard MFM encoding.
9:06 a.m.
Fred Cisin wrote:
On Wed, 28 Apr 2004, ben franchuk wrote:
standard (WD179x) controller, to produce an industry standard MFM encoding.
And a cheap dirty WD179x interface that hogged the processor
when doing I/O. Serial I/O was bit banged ... etc ... etc ...
It just cost cut too many corners.
Ben.
10:44 a.m.
In hindsight, the 6809 while the best 8 processer it only has 64k of
memory space. Time has proved the 64k data and 64k program space and
I am trying to think of a common 8-bitter with more than 64K of memory
space. OK, some microcontrollers have 64K program space and 64K data
space, I guess...
about 1 meg of memory ( something the PC-XT had or
PDP11/unix ) was
needed for useful programs with a resonable OS. Also since the only
common 6809 system was by Radio Shack as games machine you never got the
And its close relative the Dragon 32. Both were based on the same
Motorola application note IIRC.
good I/O like lower case letters and a real serial and
floppy drives.
Err, we are talking about the same CoCo, I assume....
It most certainly does have real floppies. Standard Western Digial
controller chip (1773 in my system) and drives. The buit-in serial port
is, agreed, a bit-banger (and best used as a printer output only), but
Radio Shack sold a proper serial cartridge based on the 6551 chip. I have
a homebrew version in my CoCo, with 3 6850s in it (3 serial ports...)
The video display was a problem on the CoCo 1 and CoCo 2. Of course under
OS-9 you could hang a serial terminal off one of hte serial ports. And
the CoCo 3 had a much better video system (pity it's almost unheard-of in
the UK).
-tony
11:46 a.m.
Probably OT here, but there is a guy (John Kent?)
who put a 6809 in a FPGA.
So, there is a chance for all the guys who missed a MMU,FPU,...
on a 6809 ;-)
His webpage is at:
http://members.optushome.com.au/jekent/system09/index.html
Look for System09
7325
days inactive
7326
days old
10 comments
8 participants
participants (8)
-
ard＠p850ug1.demon.co.uk -
bfranchuk＠jetnet.ab.ca -
cisin＠xenosoft.com -
dave04a＠dunfield.com -
dickset＠amanda.spole.gov -
emu＠ecubics.com -
spectre＠floodgap.com -
zmerch＠30below.com