Loading ...
Sorry, an error occurred while loading the content.
 

Re: [cosmacelf] Re: fun stuff-- and ACE bus question

Expand Messages
  • Mark Graybill
    ... There were a number of neat projects for the ISA bus, especially early on before companies got people trained to the idea that they were only supposed to
    Message 1 of 13 , Jul 12, 2003
      > I seem to recall a hobbyist magazine that had a "poor man's" spectrum
      > analyzer project that plugged into the PC's ISA bus. Considering how
      > expensive a *real* spectrum analyzer was/is, it was a considerable
      > achievement for the electronics hobbyist.

      There were a number of neat projects for the ISA bus, especially early on
      before companies got people trained to the idea that they were only supposed
      to go down to the local CompuMart and buy a prepackaged card for their
      computers. The thing was that someone who did a neat hardware hack on one of
      the old 8-bit systems had something they could then package and sell without
      a lot of fretting about whether it would run on the system it was plugged
      into because of timing differences, etc. Any hobbyist who tried to go pro
      with a homebrew ISA design was almost certain to get burned by support
      costs, customer returns, etc. The days of the cottage hardware industry were
      numbered already with that bus dominating the marketplace.

      >
      > > I was down at a Hughes Aircraft site a few years back and
      > > in the middle of a rack full of equipment for testing Maverick
      > > missile electronics I saw a good old Vic-20...
      >
      > I love hearing about stuff like that! :) It just cracks me up to no
      > end. I think that kind of thing epitomizes the term "Yankee
      > ingenuity".

      Me, too. It's gotten to the point in some circles where the younger techs
      don't even realize that what they're working with is a general purpose
      computer system. I ran across a Northstar Horizon being used to control a
      mix process for solid rocket propellants, and the techs all thought it was
      some specialized piece of equipment designed for the task. I may have made a
      mistake when I showed them how to start up MBASIC and play the games,
      though...

      >
      > Now I get to learn something- what is the ACE bus?
      >
      It's a 44 connector bus used as a standard by the folks who put out the old
      Ipso Facto newsletter. There's more info on it at cosmacelf.com. I like to
      stick with standards even when doing one-offs so long as the standard
      doesn't break the design. It's a habit that's paid off more than once.

      > I quite like the RCA Microboard bus.

      That sounds like an interesting bus, and it sounds like it's got more
      structure to it than the ACE bus, which basically just brings the CPU
      signals out to the card edge.

      > I
      > continue to marvel that the PC bunch still has a hard time sharing
      > interrupts (*reliably*), when even early 8-bit system developers have
      > known how to do it properly.

      The thing that frustrates me most about the computer industry is when
      well-known solutions to clear problems are ignored in new designs, whether
      it be in hardware or software. And then we have to go through the same
      problems all over again with yet another generation of people who've joined
      the industry since the last time we had this particular nightmare. And every
      time everybody thinks that nobody's ever seen this before, and that they
      have to come up with something new to solve it (thereby repeating all the
      iterations that happened in the course of the prior solution.)

      At least half of what people attribute to me in the way of "engineering
      genius" is the result of the fact that I always start working on a problem
      by assuming that my situation is not unique and that someone has already
      encountered this and learned how to deal with it a long time ago, and that
      the best thing I can do is find them and learn from them rather than flying
      by the seat of my pants into something I roll up by myself in a vacuum.

      I also get tired of hearing how the entire industry is going to go "poof"
      every time there's a slowdown in sales of the latest fad gizmo (followed by
      a mad scramble to find another fad gizmo), but that's another story.

      > Your project sounds like lots of fun. I'm feeling guilty since I've
      > been fighting my PC's for some time now and just finished (I hope)
      > restoring a family member's PC to stability.
      >
      > You know the drill:
      Boy I do know the drill. And I decided two years ago to get out of it. I was
      acting as tech support for my entire extended family, all their friends, and
      half the people at church. And I learned that no matter how hard I worked or
      how much I told them or what I did to the PeeCee it was just going to end up
      broken ten minutes after I walked out the door. Those things are tar babies,
      and I finally decided I wasn't going to take a swing at them any more.

      So now when someone says to me "You're good with computers, maybe you can
      help me with this problem I have with my PC..." I say "Sorry, I really don't
      know anything about PCs. The computers I work with are an entirely different
      sort of thing. If you don't believe me, come look at my PC. I need to use it
      because a lot of the design software I run will only run on a stinking PC,
      but I can't even get the thing to copy files properly. If you run into any
      problems with a computer that controls an environmental conditioning system,
      or automated test equipment, or your aircraft's avionics system, or
      something like that, I'm your man. But I'll warn you, <grin> my rates aren't
      cheap."

      I'm friendly about it, of course. But giving up a hopeless task of user
      support has recovered at least one and usually two evenings a week for me to
      spend with my family or on my own projects. And when people ask me what to
      buy for a new computer, I say "Buy a Mac. It's running Unix, and that's an
      operating system I _can_ help you with." Most of them balk at the price, but
      I ask them whether they'll end up paying the difference in pain killers for
      the headaches they'll get if they go buy another PC. (I should note that I
      think the new Macs are hardly a panacea, Apple has yet to work out some of
      the quirks in their approach to bringing Unix to the desktop, but I do
      consider them a significant improvement over the current crop of PCs. Just
      my two cents, no intent to start a holy war here.)

      >
      > Your project sounds like lots of fun. You'll post a picture when
      > you're done, right?

      I'm having a great time with it. And yes, there will be pictures.

      -Mark G.
    • J.C. Wren
      You have to understand that the PC bus was designed by a co-op. This is not a myth, nor urban lore. It s hard cold fact. I know a couple of the fellows
      Message 2 of 13 , Jul 12, 2003
        You have to understand that the PC bus was designed by a co-op. This is not
        a myth, nor urban lore. It's hard cold fact. I know a couple of the fellows
        (well, guys. They're not IBM-fellow type fellows, just guys) who worked on
        the PC, the PC-Jr, the original VGA card, and the XT HD. In fact, a friend
        of mine who was the lead design on the PC Jr has serial number #1 of both the
        PC Jr, the Speak'n'Spell card (I think it was called that), and XT HD.

        You have to remember that IBM planned to stay in the PC market for *one*
        year. IBM never expected PCs to catch on. Also, the PC was somewhat
        crippled from the outset to not compete against the DisplayWriter, which was
        IBMs premier WP system at the time. And the PC Jr, although technically
        superior in design to the PC (for instance, video memory could be located
        anywhere in the address space, not just at 0xb0000 or 0xb8000), it too was
        crippled to not compete too hard against the PC.

        As a result, the engineering that was put in to the DisplayWriter was not
        done on the PC. Why worry about interrupt sharing details when you were
        going to offer a half dozen cards, all from IBM. You didn't *need* to share
        interrupts. And when the AT came out, they weren't going to throw away the
        card base that now existed because of the PC. That would have been suicidal.
        (One of the reasons MCA never really caught on. It was technically a well
        crafted bus, although a little complicated to interface to. And the
        mechanical constraints were pretty tight for the time. Cards dimensions had
        to be accurate to within 1/100" of an inch. Not so good with clone cases,
        who STILL can't get it right).

        And yes, if you look at some of the bus timings for DMA (regular stuff isn't
        bad at all, but DMA...), you required a negative set up time, on the order of
        20ns *before* you knew DMA would happen. That's why DMA was such a bitch.

        So given the cirsumstances at the time, decisions they made (or didn't make)
        made sense. It wasn't until the PC caught on that everything became an
        engineering nightmare.

        Oh, and do you know why the PC is 8088 based instead of 6809? Primarily
        because they had about 6 Intel MDS development stations sitting around from
        the DisplayWriter project. Since the PC was originally an unsanctioned
        project, there was no budget to buy the Motorola systems. Which would have
        been the preferred choice, and probably would have not introduced a 10 year
        set back in computing due to segmented architecture.

        --John

        On Saturday 12 July 2003 15:46 pm, Mark Graybill wrote:

        >
        > > I
        > > continue to marvel that the PC bunch still has a hard time sharing
        > > interrupts (*reliably*), when even early 8-bit system developers have
        > > known how to do it properly.
        >
        > The thing that frustrates me most about the computer industry is when
        > well-known solutions to clear problems are ignored in new designs, whether
        > it be in hardware or software. And then we have to go through the same
        > problems all over again with yet another generation of people who've
        > joined the industry since the last time we had this particular nightmare.
        > And every time everybody thinks that nobody's ever seen this before, and
        > that they have to come up with something new to solve it (thereby repeating
        > all the iterations that happened in the course of the prior solution.)
      • Dave Ruske
        Hmm, good question. I m looking at the ACE CPU board schematic and it appears that I/O Select is an input to that card that gets exclusive NORed with some
        Message 3 of 13 , Jul 12, 2003
          Hmm, good question. I'm looking at the ACE CPU board schematic and it
          appears that I/O Select is an input to that card that gets exclusive
          NORed with some EPROM select logic and used as the chip select for the
          fourth RAM/EPROM socket on the card (chip 20). My guess is that the
          design allows for some memory mapped I/O of some sort on another card,
          but... well, I've been on the software side of things for too long, and
          quite frankly I've seen better schematics.

          I'll try to get these scanned over the next day or three and let you
          hardware wizards work it out. Cleaning the scans up to where they'll be
          readable could be tricky, though.

          I haven't yet located an article on the ACE bus itself, which surprises
          me a little. I'll post it if I can find any such thing.

          Dave

          On Friday, July 11, 2003, at 07:28 PM, Mark Graybill wrote:
          > Anyway, I've got an on-topic item, too. On the ACE bus there's a signal
          > called I/O select. Can anyone tell me what it is and how is it usually
          > derived? Is it something like an OR of N0, N1 and N2? I was looking at
          > the
          > ACE bus since we're building on boards with the standard 22/44 contact
          > card
          > edge on them.
        • Dave Ruske
          I want names. I want addresses. I want to hurt someone! (Just kidding of course, but good Lord, SEGMENTS! What a nightmare...) Dave
          Message 4 of 13 , Jul 12, 2003
            I want names. I want addresses. I want to hurt someone!

            (Just kidding of course, but good Lord, SEGMENTS! What a nightmare...)

            Dave

            On Saturday, July 12, 2003, at 02:59 PM, J.C. Wren wrote:
            > Since the PC was originally an unsanctioned
            > project, there was no budget to buy the Motorola systems. Which would
            > have
            > been the preferred choice, and probably would have not introduced a 10
            > year
            > set back in computing due to segmented architecture.
          • Mark Graybill
            Yes, and there s no telling how many prospective assembly programmers were put off by the segmented architecture alone. The 6809 was the leading contender for
            Message 5 of 13 , Jul 12, 2003
              Yes, and there's no telling how many prospective assembly programmers
              were put off by the segmented architecture alone.

              The 6809 was the leading contender for the CPU for the Macintosh, too,
              when Jef Raskin was still on the project. One of the folks I worked
              with who'd been on the original PC project mentioned to me that the
              68008 had been considered as well, and would have been preferred over
              the 8008, but Moto wasn't as willing to lie about ship dates for their
              chip as certain other companies, so it got dropped out of the running.

              -Mark G.

              On Saturday, July 12, 2003, at 06:16 PM, Dave Ruske wrote:

              > I want names. I want addresses. I want to hurt someone!
              >
              > (Just kidding of course, but good Lord, SEGMENTS! What a nightmare...)
              >
              > Dave
              >
              > On Saturday, July 12, 2003, at 02:59 PM, J.C. Wren wrote:
              >> Since the PC was originally an unsanctioned
              >> project, there was no budget to buy the Motorola systems. Which would
              >> have
              >> been the preferred choice, and probably would have not introduced a 10
              >> year
              >> set back in computing due to segmented architecture.
              >
            • Mark Graybill
              I finally got the palm stuff installed on the Mac, and gave your utility a try, Dave. It works great! I wonder--is there a technical reason for the 16K file
              Message 6 of 13 , Jul 12, 2003
                I finally got the palm stuff installed on the Mac, and gave your
                utility a try, Dave. It works great! I wonder--is there a technical
                reason for the 16K file size limit? There aren't many files larger than
                that that I'd consider moving around, but it would be nice to have the
                ability if it's not too big a deal.

                However, thanks a lot! This makes a great addition to TinyElf. What
                cross dev tools do you recommend for Mac OS X? I tried to get Crossbow
                2.70, but my system choked on the .sit file, and I wasn't sure if it
                was an X or Classic app.

                -Mark G.
              • J.C. Wren
                I dunno about the Mac aspect, but Molarota really shot themselves in the head when the announced the 68000 3 quarters before it shipped. Many companies
                Message 7 of 13 , Jul 12, 2003
                  I dunno about the Mac aspect, but Molarota really shot themselves in the head
                  when the announced the 68000 3 quarters before it shipped. Many companies
                  decided not to do a 6809 design (which was a really cool chip, at least, the
                  E version. The non-E was a bear), and wait for the 68000 to show up.

                  Then GM took something like the first 100,000 off the line to be used in the
                  engine control units, starving the rest of the market, and basically pushing
                  out delivery for another 2 quarters.

                  I still have a little 6809 board I point-to-pointed, running a friend of
                  mines monitor. He was big in the '09, and SWTPC systems. IIRC, he wrapped a
                  replacement CPU card using the 6809, and was running a copy of OS-9 on it.

                  I should pull mine out and see if it still powers up.

                  --John

                  On Saturday 12 July 2003 22:22 pm, Mark Graybill wrote:
                  > Yes, and there's no telling how many prospective assembly programmers
                  > were put off by the segmented architecture alone.
                  >
                  > The 6809 was the leading contender for the CPU for the Macintosh, too,
                  > when Jef Raskin was still on the project. One of the folks I worked
                  > with who'd been on the original PC project mentioned to me that the
                  > 68008 had been considered as well, and would have been preferred over
                  > the 8008, but Moto wasn't as willing to lie about ship dates for their
                  > chip as certain other companies, so it got dropped out of the running.
                  >
                  > -Mark G.
                  >
                  > On Saturday, July 12, 2003, at 06:16 PM, Dave Ruske wrote:
                  > > I want names. I want addresses. I want to hurt someone!
                  > >
                  > > (Just kidding of course, but good Lord, SEGMENTS! What a nightmare...)
                  > >
                  > > Dave
                  > >
                  > > On Saturday, July 12, 2003, at 02:59 PM, J.C. Wren wrote:
                  > >> Since the PC was originally an unsanctioned
                  > >> project, there was no budget to buy the Motorola systems. Which would
                  > >> have
                  > >> been the preferred choice, and probably would have not introduced a 10
                  > >> year
                  > >> set back in computing due to segmented architecture.
                  >
                  > Yahoo! Groups Sponsor
                  >
                  >
                  >
                  > ADVERTISEMENT
                  >
                  >
                  >
                  >
                  >
                  >
                  > ========================================================
                  > Visit the COSMAC ELF website at http://www.cosmacelf.com, or view the
                  > Wiki/FAQ at http://1802.bitting.com/
                  >
                  > To unsubscribe from this group, send an email to:
                  > cosmacelf-unsubscribe@yahoogroups.com
                  >
                  >
                  >
                  > Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
                • Dave Ruske
                  ... The current version of TinyELF is limited to 16K, for some legacy reason I had when I originally wrote it two years ago... for the life of me, I can t
                  Message 8 of 13 , Jul 13, 2003
                    On Saturday, July 12, 2003, at 09:26 PM, Mark Graybill wrote:
                    > I finally got the palm stuff installed on the Mac, and gave your
                    > utility a try, Dave. It works great! I wonder--is there a technical
                    > reason for the 16K file size limit? There aren't many files larger than
                    > that that I'd consider moving around, but it would be nice to have the
                    > ability if it's not too big a deal.

                    The current version of TinyELF is limited to 16K, for some legacy
                    reason I had when I originally wrote it two years ago... for the life
                    of me, I can't quite remember what it was. At the time, though, I
                    hadn't provided ASCII keyboard support, so there probably wasn't much
                    use for that much RAM anyway. I should reinvestigate this, Palm OS 2
                    compatibility isn't much of an issue any more.

                    > However, thanks a lot! This makes a great addition to TinyElf. What
                    > cross dev tools do you recommend for Mac OS X? I tried to get Crossbow
                    > 2.70, but my system choked on the .sit file, and I wasn't sure if it
                    > was an X or Classic app.

                    I don't know if there are any cross development tools on OS X yet, to
                    be honest. Assemblers are simple things to write, though, and I have an
                    emulator running but still in need of much work. So better OS X support
                    is just a matter of (spare) time... :)

                    Dave
                  • Stewart
                    CrossBow has been around for a long time and may not have been updated recently. I think I had problems with it even under OS 9.1, before moving to X/Classic.
                    Message 9 of 13 , Jul 13, 2003
                      CrossBow has been around for a long time and may not have been updated
                      recently. I think I had problems with it even under OS 9.1, before
                      moving to X/Classic. I keep it on an older PowerMac 9500 in the attic
                      now, running OS 8.1 and still equiped with the old style serial plugs
                      supporting my homebrew cables on my eprom burner and tape punch.

                      Stewart
                      ----------

                      --- In cosmacelf@yahoogroups.com, Dave Ruske <dave@r...> wrote:
                      > On Saturday, July 12, 2003, at 09:26 PM, Mark Graybill wrote:
                      > > I finally got the palm stuff installed on the Mac, and gave your
                      > > utility a try, Dave. It works great! I wonder--is there a technical
                      > > reason for the 16K file size limit? There aren't many files larger
                      than
                      > > that that I'd consider moving around, but it would be nice to have the
                      > > ability if it's not too big a deal.
                      >
                      > The current version of TinyELF is limited to 16K, for some legacy
                      > reason I had when I originally wrote it two years ago... for the life
                      > of me, I can't quite remember what it was. At the time, though, I
                      > hadn't provided ASCII keyboard support, so there probably wasn't much
                      > use for that much RAM anyway. I should reinvestigate this, Palm OS 2
                      > compatibility isn't much of an issue any more.
                      >
                      > > However, thanks a lot! This makes a great addition to TinyElf. What
                      > > cross dev tools do you recommend for Mac OS X? I tried to get Crossbow
                      > > 2.70, but my system choked on the .sit file, and I wasn't sure if it
                      > > was an X or Classic app.
                      >
                      > I don't know if there are any cross development tools on OS X yet, to
                      > be honest. Assemblers are simple things to write, though, and I have an
                      > emulator running but still in need of much work. So better OS X support
                      > is just a matter of (spare) time... :)
                      >
                      > Dave
                    • Lee Hart
                      ... Great line, Mark! -- Lee A. Hart Ring the bells that still can ring 814 8th Ave. N. Forget your perfect offering Sartell, MN
                      Message 10 of 13 , Jul 13, 2003
                        Mark Graybill wrote:
                        > [PCs] are tar babies, and I finally decided I wasn't going to take
                        > a swing at them any more.

                        Great line, Mark!
                        --
                        Lee A. Hart Ring the bells that still can ring
                        814 8th Ave. N. Forget your perfect offering
                        Sartell, MN 56377 USA There is a crack in everything
                        leeahart_at_earthlink.net That's how the light gets in - Leonard Cohen
                      • Dave Ruske
                        The schematics for the ACE CPU card are available through 3 links at the bottom of the cosmacelf.com documents page: http://www.cosmacelf.com/docs.htm Also
                        Message 11 of 13 , Jul 13, 2003
                          The schematics for the ACE CPU card are available through 3 links at
                          the bottom of the cosmacelf.com documents page:

                          http://www.cosmacelf.com/docs.htm

                          Also added to that page is a copy of an advertisement in the December
                          1977 Byte magazine for an 1802-based computer by Child Odyssey
                          Enterprises of New Mexico (thanks to Mark Graybill for providing the
                          scan of this ad).

                          Enjoy!

                          Dave
                        Your message has been successfully submitted and would be delivered to recipients shortly.