I went to a surplus store last week and found an odd Atari box. It says
it's an Automatic ROM/RAM Tester. It has two rows of address switches, one
switches on one row are labeled 10 through 0 and have LEDs above them. The
second row of switches only has five switches and they marked 11 through 15
and don't have corrosponding LEDS. Switch 11 is directly below switch 10,
switch 12 is below switch 9, and so on. It looks like they may be
multiplexed somehow. The box also has 8 Data switches and LEDs. The are
four low insertion force sockets on the box, two are for 25 pin DIPs and 2
are for 20 pin DIPs. There are also various other switchs marked 1K/2K,
RUN/STOP, Read/R/W*, and GAME/BOX. There is only one connector on the box,
it goes to a 50 pin card edge connector via a ribbon cable. There are no
power connectors or other connectors on it so it has to part of a larger
unit. Any ideas? Anyone?
I had just read the interview oyu mention yesterday. Could someone
give me some idea of what chips were involved in this thing? Did it
use PROMs, microcontrollers, or what? How can you make a game with
44 chips (not an expression of disbelief)?
>> It's reasonably obvious from looking at the design of all Apple ][
>> machines and their peripherals that they were designed to use the
>> simplest hardware possible. Things like the 'minimalist' disk
>> the bit-banger serial port and the Apple ][ I/O structure support
>> Why therefore did all Apple ][s use a hardware encoded keyboard with
>> fairly complex (and expensive) encoder IC? Software scanning would
>> been simpler in the hardware, not added much to the software, and
>> have been more versatile in many ways (the single-wire shift key mod
>> wouldn't have been necessary).
>At the time (mid-70's), surplus keyboards with ASCII encoded parallel
>outputs were readily available from many sources. Check, for example,
>the ads in the back of a '75 or '76 _Radio-Electronics_, or see
>many of the articles in the early _BYTE_'s. According to interviews
>with Wozniak (for example, _BYTE 1984:12 p.167), it was no big
>deal to find such a keyboard, put the right plug on the end, and
>plug it in (he says, of the Apple I, "You also had
>to get a keyboard and wire it into a 16-pin DIP connector".)
>Apple II motherboards were also available without a case or keyboard
>or power supply, and many hobbyists bought these as a way of saving
>a couple of bucks.
>Why use an encoder IC? Yes, it is cheaper if you skip this and
>have the CPU scan the keyboard. But encoder IC's weren't expensive
>at the time, and were an extremely common industry-standard
>solution; see, for example, Don Lancaster's _TTL Cookbook_ to get
>an idea at how readily available such encoders were in the mid-70's.
>Wozniak often did go for non-industry-standard solutions to
>peripheral interfacing, but generally only when it reduced the
>chip count. It seems to me that the encoder-IC approach to a
>keyboard has a lower part count than if you use SSI TTL IC's to let
>the CPU scan the keyboard.
>Chip count was an extremely important thing to Wozniak; for example,
>in the same interview I referenced before, he says
>was really annoyed because all their new games were coming out at 150-
>170 chips. He wanted low chip counts to reduce costs,
>and he had seen a version of Pong that I had done, that only used
>about 30 chips. He appreciated that. So he said if we could design
>a hardware Breakout in under 50 chips, we'd get 700 bucks; and if
>it was under 40 chips, we'd get $1000. ... We gave them a working
>breadboard for it. My first design was 42 chips. By the
>time we got it working it was 44, but we were so tired we
>couldn't cut it down. So we only got 700 bucks for
Get Your Private, Free Email at http://www.hotmail.com
I first discovered digital logic and TTL IC's about 1973, and wanted to
build something not trivial with them. The project I chose was John Conway's
game of "life" as described in Martin Gardner's Mathematical Games in
I ending using a 32x32 element grid using 2 (then expensive!) 2102 rams (No
A or other suffix then) I would calculate one generation from one ram and
write to the other. I used about a 60kHz clock and 74193 counters to
generate the addresses. The clock went to a decade counter and the first 8
states generated +/- x and y clocks to visit the 8 neighbors of a given
cell. If the cell was a "1", another counter was incremented, number of
neighbors. The 9'th state was at the given cell.
The rule was then used: Neighbors =2 and cell active, or Neighbors = 3, then
the new cell was active, otherwise it was not. The 10'th clock wrote the
cell to the new generation (the other ram). Thus 10 clock cycles were needed
for each cell, or 10240 for each generation. This was much faster than a
later 4MHz Z-80 program!
I displayed the results on a 5 inch oscilloscope, using 555 timers with PNP
transistor current sources to charge the timing capacitor, for x and y ramp
or deflection waveforms. This was straight out of a National Semiconductor
databook. The refresh rate was about 60Hz. The "load" signals for the
74193's was used to write data into locations from address switches. Finally
3 7490's and 7447's displayed the generation number on 7 segment displays.
It was fun to watch the generations flash by. The total number of IC's was
about 35 + 6 for the generation counter.
I recently found a Convergent computer in a surplus store here. It has the
display section, CPU section and hard drive in blocks that attach to each
other and looks like it uses some kind of external power supply. Each
block had a model number on it but the only one that I remember was CP-001
on the CPU. Anyone know what this thing is and wheather it's worth picking
At 10:36 PM 5/29/98 -0400, you wrote:
>I recently picked up some more Convergent machines at a thrift shop. I was
>pretty amazed; my 186 model was the only one I'd ever seen. Now I've got
>286 and 386 models. Wow!
>Anyway, I also picked up a few disk modules. Two 85 MB modules, I think.
>(compare to my existing pair of 10 MB modules...) One is an 'expansion'
>and needs to be connected to another disk module to work.
>My problem is this: the one disk that has CTOS installed was set up as a
>'Cluster Controller' and actually has accounts and passwords set up. I can
>just press 'Go' at the login on my CM001 and use the system. This doesn't
>work on the new system.
>Is there a way I can recover the files on this disk (meaning the OS and
>application software)? I have OS disks for the standalone version, but I'd
>like to preserve the cluster controller so starting over is a
>less-than-ideal option. I do have other working systems and parts
>available to press into service for the cause.
>Thoughts, hints, suggestions?
<> Found a DELL 316(386sx/16), has 5.25 360k, 1.44 floppy, 170mb IDE with
<> QUESTION: is there a CMOS and how do I access it?
<Should be CTL-ALT-ENTER if I remember correctly. Once you're at DOS you c
That was the winner, thanks. I should have specifed that it was a Pheonix
FYI it's a great Minix machine! For those that are interested there is
a site on the net (http://minix1.hampshire.edu). Bins and sources are
available on the net. It can be run on anything from an XT to P6.
There is also a book Operating Systems design and Implmentation 2nd
Edition with CDrom of sources and bins.
This is fairly recent so please reply offlist.
Found a DELL 316(386sx/16), has 5.25 360k, 1.44 floppy, 170mb IDE with
a DEC color tube and keyboard. It runs so it's going to become a minix or
QUESTION: is there a CMOS and how do I access it?
<Tandy had little input into the design really), it was a fine machine.
<OS-9 on the CoCo was the first real OS I ever used - it taught me a lot
<about writing device drivers, multitasking, etc. And the CoCo3 (alas not
<at all common in the UK) was a very interesting box.
OS9, I've heard a little about it but never seen it. Is it available?
I have a COCO (I think a III, has a few ASICs and 128k), no disk
controller but that can't be much majik.
<And don't forget the 3rd party software for the Z-80 machines. LDOS was
<arguably the finest Z80 operating system at the time - it was a lot
What ever happend to LDOS?
Many thanks to the Listmembers who have offered suggestions and
assistance and info.. I have discovered (and rectified) the problem
with my RK05 talking to it's host.
I could get into a long shaggy-dog story here, but basically I had
the *drive* end of the interface cable in bass-ackward.
>>> DOH! <<<
(Lame excuse: it's the end you can't see)
And now the terminal sez:
And, even though the problem turned out to be a forehead-slapper, I
must say that I have learned great gobs of stuff about how the drive
works, how it interfaces, and what makes it tick. I would *never*
have had the opportunity to get such (rare) understanding, had it
not been for the long diagnostic journey. And that is one of my
goals in collecting these things: to know them well enough to keep
them on the air, and to be able to share that knowledge with others
Thanks to Tony Duell, Tim Shoppa, Ward Griffiths, and Huw Davies for
their generous assistance.
Now: get the RL02s on-line, find formatter for the Kennedy 9trk,
find an interface for the System Industries 470M drive, get Unix
running, &tc; &tc; &tc.
But.. TRW Ham/Electronic/Computer swapmeet is tomorrow, the 28th
(Sat) here in SoCal... and I'm heading that way... to find all the
goodies before Marvin has a chance at them.
Another trio of questions. I am reading Steven Levy's Insanely Great,
about the making of the Apple Mac.
1)I finally have learned the purpose of Quickdraw through this book.
Do all windowing systems do this, or are there ones that render all
windows even if they are obscured. The name of the system I want to
know about ought not be mentioned for obvious reasons.
2)It mentioned a "new input device" that was before the Xerox machine,
to be used alongside the mouse, was flexible but required training.
What was it?
3)Should I keep my reference cards attached to the manuals or tear
Get Your Private, Free Email at http://www.hotmail.com