>OK, hands up everyone who's done a real-time embedded app with an 8008!
Never actually programmed one, but did get to handle equipment that used
them, although it was obsolete and withdrawn from service by then.
However before I got to that stage I did have occasion to deal with a data
capture board that used the first 16 bit micro that National Semiconductor
made (cannot recall its name, but it might have been the Pace), or actually
it was a board that used the instruction set, but predated the micro itself.
It had a pair of national Semiconductor EPROMs on it for the program, and
they were very early technology, where the program leaked away after about 6
months. We had to reprogram every 6 months as they went AWOL, using a little
box with switches to enter the data. Took about a day for 2 people to enter
and verify a 256 word program this way, a job that no-one liked.
> >OK, hands up everyone who's done a real-time embedded app with an 8008!
Well, never really did an embedded app, but my first "PC" was an 8008 with
128 bytes of RAM and the total I/O complement consisted of 8 LEDs and 8
switches. With the limited I/O it was pretty boring, so I quickly upgraded
to a 6502 with a whopping 1K of RAM and a model 15 teletype.
Well, maybe not that quick ... I think it took me half a year to get the
thing to talk to the baudot teletype. Assembly was by hand and the initial
program load was accomplished by punching octal onto a paper tape to feed
the EEPROM burner. The EEPROM burner was owned by another department across
the research campus, so I could only get it once in a while late at night.
Pretty slow way to debug compared to ICSP! Especially since the only device
was the TTY that wouldn't talk! I ended up building a pretty elaborate
circuit with lots of LEDs to give me a clue as to what was happening on the
bus. (Didn't even have a scope back then).
It wouldn't wish that grief on my worst enemy, but I wouldn't trade that
experience for anything. I guess learning is always painful, and I learned
plenty!
>
>Well, maybe not that quick ... I think it took me half a year to get the
>thing to talk to the baudot teletype. Assembly was by hand and the initial
>program load was accomplished by punching octal onto a paper tape to feed
>the EEPROM burner. The EEPROM burner was owned by another department across
>the research campus, so I could only get it once in a while late at night.
>Pretty slow way to debug compared to ICSP! Especially since the only device
>was the TTY that wouldn't talk! I ended up building a pretty elaborate
>circuit with lots of LEDs to give me a clue as to what was happening on the
>bus. (Didn't even have a scope back then).
>
>It wouldn't wish that grief on my worst enemy, but I wouldn't trade that
>experience for anything. I guess learning is always painful, and I learned
>plenty!
Do they even teach that sort of thing anymore?
Always start with the output! Otherwise, how do you know what's going on?
My first computer (of any kind) was a PDP-10 timesharing
system with good old teletype terminals. This thing occupied
either five or seven (can't remember) six foor high racks with
all the wiring under the floor. We mostly ran BASIC on it,
though there were times when some of us actually got down
to the machine level. Uts owners (ITS dept) frowned on us
doing that.
First micro was the 6502 in an Atari 800, but that was my
home hobby machine. For work we used 8080s, and the
later upgrade from Zilog the Z-80.
Who remembers Zilog trying to claim patent rights to any
name starting with Z-? It was all an advertising thingm but
at the time we were quite amused by it. Got them a LOT of
attention.
We did a MIL project with the RCA 1802 once - used it
because it was CMOS and used really low currents, and
becasue it was apparently RAD-hard compard to the Pmos
and Nmos chips the Z80s and other used.
Here's one for you - why do all 1802 instructions take 8
clocks to execute? Because the chip has a 1-bit ALU and
every instruction is clocked past the ALU 1 bit at a time.
Saved power.
I did an 1802 job once. Very, very strange sort of prehistoric instruction
set, as I recall with a set of 16-bit orthogonal registers and a pretty
unique stack based interrupt architecture. I still have a cpm-based
assembler around here somewhere for it.... It was one of the few micros that
was rad-hard enough to be space qualified. There was this strange sort of
development board for it called the COSMAC (or something like that) that had
a bunch of seven-segment displays and a little hex keypad on it.
Ed
{Original Message removed}
Yep, the COSMAC ELF as a matter of fact. I used one of these in school
back in the late 70's early 80's. The one I had came with 256 bytes of
RAM, a Video Interface Chip, 4 or 5 expansion slots, a Hex Keypad, some
7 segment LED's (IIRC there were 4, all address and the 2 LSD's for data),
an LED for "Q" output, a toggle switch for Run-Program selection, and a
handful of documentation. My first program was about 249 bytes or so
that would display the starship "Enterprise" on a TV screen. But if you
lost power, you had to eneter it again by hand. I was also going to use
this machine for my senior project, but wound up using a VIC 20 instead.
I still have this computer somewhere in my garage. I ought to dig it out
and see if it still works.
Then there's the Radio Shack Model 4, the Model 100, the CoCo, the TI
99/4, the Adam, and a few others that I can't recall the model of at this
point in time. I ought to put these on eBay and see if anyone is
interested. Then again, maybe not.
Regards,
Jim
> I did an 1802 job once. Very, very strange sort of prehistoric
> instruction set, as I recall with a set of 16-bit orthogonal registers
> and a pretty unique stack based interrupt architecture. I still have a
> cpm-based assembler around here somewhere for it.... It was one of the
> few micros that was rad-hard enough to be space qualified. There was
> this strange sort of development board for it called the COSMAC (or
> something like that) that had a bunch of seven-segment displays and a
> little hex keypad on it.
> Ed
> {Original Message removed}
On Jan 19, 2005, at 11:36 AM, Roy J. Gromlich wrote:
> My first computer was a PDP-10 timesharing system
Yah, me too. Shared by several high-schools and such, which
resulted in a rather primitive form of "networking" (messaging
and email existed. Sort of.) In fact, I was sucked into
mainframes and timesharing for a long time. I spent some
bucks on an S100 system that I never quite got working, but
the first working personal computer I ended up owning was an
HP-150 (x86 running msdos, but not quite a PC clone.)
On Jan 19, 2005, at 1:43 AM, Paul James E. wrote:
> the COSMAC ELF as a matter of fact. I used one of these in school
> back in the late 70's early 80's. The one I had came with 256 bytes
> of
> RAM, a Video Interface Chip, 4 or 5 expansion slots, a Hex Keypad,
> some
> 7 segment LED's
Hard-to-find 5x7 dot matrix HEX LEDs (TIL311s?) IIRC. (does that make
us
back on topic? :-)
> the COSMAC ELF as a matter of fact. I used one of these in school
> back in the late 70's early 80's. The one I had came with 256 bytes
> of
> RAM, a Video Interface Chip, 4 or 5 expansion slots, a Hex Keypad,
> some
> 7 segment LED's
I think I still have a couple of blank boards to make some IIRC. :)) The
basic Elf did not have the video display, that came later. The other device
for which the video chip, that went with the Cosmac CPU chipset, was made
was the video game console that RCA did. As the NZ agents for RCA, the
company I worked for imported one to look at, and boy did it look old
fashioned. I think I still have the memory dumps I did of the ROMs in it.
I do still have the COSMAC development board somewhere, populated with 4k of
2101 256 byte memory chips which take up about half the board space. One of
these was the first micro I ever really did any programming on.
> I think I still have a couple of blank boards to make some IIRC. :))
The
> basic Elf did not have the video display, that came later. The other
device
> for which the video chip, that went with the Cosmac CPU chipset, was
made
> was the video game console that RCA did. As the NZ agents for RCA, the
> company I worked for imported one to look at, and boy did it look old
> fashioned. I think I still have the memory dumps I did of the ROMs in
it.
>
> I do still have the COSMAC development board somewhere, populated with
4k of
> 2101 256 byte memory chips which take up about half the board space.
One of
> these was the first micro I ever really did any programming on.
Pedanticism: 2101's are only 4-bits wide making them 256 nibble srams.
;-)
If you have a spare CDP1861, I'd pay good money for it. I need one to
fix my ELF. The RCA Studio II used it for video.
>
> Roy
>
> Yep, the COSMAC ELF as a matter of fact. I used one of these in
school
> back in the late 70's early 80's. The one I had came with 256 bytes
of
> RAM, a Video Interface Chip, 4 or 5 expansion slots, a Hex Keypad,
some
> 7 segment LED's (IIRC there were 4, all address and the 2 LSD's for
data),
Mine only has 2 7-segment LED's and the Q LED, no address indicator.
:-( My board was a second generation (hex keypad instead of 8 toggle
switches). You had to key in a short routine(if you were smart) that
would sequence memory by storing the address into the memory location.
For example, address 0x10 would contain the value 0x10. Then as you
were keying in a large program, you would occaisionally turn on the
memory protect switch and hit the input key to see if the next memory
location contained the value you expected, or if contact bounce on the
Input key had done you in somewhere. You would then reset the CPU
(leaving the memory protect switch turned on) and then pound away at the
input key until you got to the memory location where you left off. What
a great system. ;-)
> an LED for "Q" output, a toggle switch for Run-Program selection, and
a
> handful of documentation. My first program was about 249 bytes or so
> that would display the starship "Enterprise" on a TV screen. But if
you
> lost power, you had to eneter it again by hand. I was also going to
use
> this machine for my senior project, but wound up using a VIC 20
instead.
> I still have this computer somewhere in my garage. I ought to dig it
out
> and see if it still works.
Yes you should, then you should sell me your CDP1861. ;-) There's a
big following for this machine now, it's become sort of a cult classic.
Check this out: http://www.cosmacelf.com
Kids are building replicas from NOS parts (when they can be found) and
there are FPGA implementations as well. Remember Tom Pittman's "A Short
Course on Programming"? I still have a copy somewhere.
> Then there's the Radio Shack Model 4, the Model 100, the CoCo, the TI
> 99/4, the Adam, and a few others that I can't recall the model of at
this
> point in time. I ought to put these on eBay and see if anyone is
> interested. Then again, maybe not.
Let me know on that ELF. I think you live in the Houston area.
>> the COSMAC ELF as a matter of fact. I used one of these in
>> school
>> back in the late 70's early 80's. The one I had came with 256
>> bytes
>> of RAM, a Video Interface Chip, 4 or 5 expansion slots, a Hex
>> Keypad,
>> some7 segment LED's
...
> I do still have the COSMAC development board somewhere, populated
> with 4k of
> 2101 256 byte memory chips
Modern youth may not appreciate what it took to get that 4 kB.
The 2101 is 256 x 4, so you'd need 32 for your 4 kB of memory.
Or to make the 256 MB of an entry level modern PC you'd need. about 2
million of them :-).
Moore's law lives !!!! :-).
>Pedanticism: 2101's are only 4-bits wide making them 256 nibble srams.
>;-)
Yeah, sorry, no wonder there are so many chips on the board :)
>If you have a spare CDP1861, I'd pay good money for it. I need one to
>fix my ELF. The RCA Studio II used it for video.
No I don't, but I imagine that it would not be too hard to make a
replacement out of an FPGA. The internal logic was not that terribly
complicated as I remember.
There was also a CDP1862 for European scan rates, which had the timing
registers fiddled with the appropriate values for 625/50 instead of 525/60.
Yeah, the video game I talked about was the Studio II.
> Somebody wrote:
>> I do still have the COSMAC development board somewhere, populated
>> with 4k of
>> 2101 256 byte memory chips
>
> Modern youth may not appreciate what it took to get that 4 kB.
> The 2101 is 256 x 4, so you'd need 32 for your 4 kB of memory.
> Or to make the 256 MB of an entry level modern PC you'd need. about 2
> million of them :-).
> Moore's law lives !!!! :-).
IIRC, those chips were something like US$5.00 each in the late 70's,
making memory quite expensive by todays standards.
> Modern youth may not appreciate what it took to get that 4 kB.
> The 2101 is 256 x 4, so you'd need 32 for your 4 kB of memory.
Remember the 1802-based ETI660 and CHIP-8 ? Great fun to
learn assembly on. Output to colour TV and rudimentary sound.
Young 'uns in the family (me included) found the simple games
entered in hex from magazines just as challenging as any on a
PC. Still got mine kicking around
On Thu, 20 Jan 2005 04:54:47 -0600, michael brown wrote:
2101s...
> IIRC, those chips were something like US$5.00 each in the late 70's, making memory quite expensive by todays standards.
I used 2114s for my Z80 machine, and I seem to remember that a 6" x 8" card would hold 2kB - and the chips cost about UK£20 per kByte (and in the late seventies that was a lot of money :-) These were Static, of course - Dynamic RAM was much cheaper but you had all that messy refreshing to do (the Z80 would help with that, but it was much easier not to have to bother).
> -----Original Message-----
> From: .....piclist-bouncesKILLspam.....mit.edu
> [EraseMEpiclist-bouncesspam_OUTTakeThisOuTmit.edu] On Behalf Of Alan B. Pearce
> Sent: 2005 Jan 19, Wed 08:06
> To: Microcontroller discussion list - Public.
> Subject: Re: [OT] Old Time Micros, was LED Matrix
>
> >OK, hands up everyone who's done a real-time embedded app
> with an 8008!
>
> Never actually programmed one, but did get to handle
> equipment that used them, although it was obsolete and
> withdrawn from service by then.
>
> However before I got to that stage I did have occasion to
> deal with a data capture board that used the first 16 bit
> micro that National Semiconductor made (cannot recall its
> name, but it might have been the Pace), or actually it was a
> board that used the instruction set, but predated the micro itself.
> It had a pair of national Semiconductor EPROMs on it for the
> program, and they were very early technology, where the
> program leaked away after about 6 months. We had to reprogram
> every 6 months as they went AWOL, using a little box with
> switches to enter the data. Took about a day for 2 people to
> enter and verify a 256 word program this way, a job that no-one liked.
>
> We did a MIL project with the RCA 1802 once - used it
> because it was CMOS and used really low currents, and
> becasue it was apparently RAD-hard compard to the Pmos
> and Nmos chips the Z80s and other used.
"The 1801 was introduced in early 1975 and the 1802 in early 1976. The
1801 microprocessor and the COSMAC architecture were the first based on
CMOS chip technology. CMOS chips could be suspended to save battery
life, operated in extended temperature ranges, and were more tolerate of
electronic interference. It is no wonder that the COSMAC was the first
microprocessor in space. The 1801 flew on an OSCAR satellite launched in
1978. Other COSMAC spacecraft include UoSAT-1, UoSAT-2, Voyager (3
1802's), Viking, and Galileo. The COSMAC's that flew in space were
space/radiation hardened versions using a CMOS/SOS chip technology.
These radiation hardened versions were developed in conjunction with the
Sandia National Laboratories."
I don't work in the field, but it's my understanding that the venerable
1802 silicon-on-sapphire version is still used today in satellite
designs. Tried and true.
I was at the RCA labs in New Jersey (By then Harris had bought them)
in 1986/87 working on some software tools. There was a man who had
just passed retirement age but kept on until after the last major Voyager
encounters just in case he was need for some last minute system
emergency. He had personally hand taped out the 1801 processor
masks.
> Roy J. Gromlich wrote:
>
> > We did a MIL project with the RCA 1802 once - used it
> > because it was CMOS and used really low currents, and
> > becasue it was apparently RAD-hard compard to the Pmos
> > and Nmos chips the Z80s and other used.
>
> http://www.antiquetech.com/chips/RCA1801.htm
>
> "The 1801 was introduced in early 1975 and the 1802 in early 1976. The
> 1801 microprocessor and the COSMAC architecture were the first based on
> CMOS chip technology. CMOS chips could be suspended to save battery
> life, operated in extended temperature ranges, and were more tolerate of
> electronic interference. It is no wonder that the COSMAC was the first
> microprocessor in space. The 1801 flew on an OSCAR satellite launched in
> 1978. Other COSMAC spacecraft include UoSAT-1, UoSAT-2, Voyager (3
> 1802's), Viking, and Galileo. The COSMAC's that flew in space were
> space/radiation hardened versions using a CMOS/SOS chip technology.
> These radiation hardened versions were developed in conjunction with the
> Sandia National Laboratories."
>
> I don't work in the field, but it's my understanding that the venerable
> 1802 silicon-on-sapphire version is still used today in satellite
> designs. Tried and true.
>
> Nate
Moores law. I bought 11 PDP-11's at one point for a network simulation
lab in the early 70's the memory was $1,000 / K a new unheard of bargain.
It was core. Two or three years earlier PDP 8's were $20,000 with
4K of 12 bit memory. Computers for every home was just around the
corner :)
> Russell McMahon wrote:
>
> > Somebody wrote:
> >> I do still have the COSMAC development board somewhere, populated
> >> with 4k of
> >> 2101 256 byte memory chips
>
> IIRC, those chips were something like US$5.00 each in the late 70's,
> making memory quite expensive by todays standards.
rl.ac.uk>
