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

Re: [cosmacelf] Re: +Viking,+Voyager+firmware+anyone+?

Expand Messages
  • arthurok
    from what i remember the viking lander was fairly large and had radioactive thermal voltaic power generators no solar cells required ...
    Message 1 of 12 , Dec 25, 2009
      from what i remember the viking lander was fairly large
      "and had radioactive thermal voltaic power generators"
      "no solar cells required"
      ----- Original Message -----
      From: Trevor Nunes
      To: cosmacelf@yahoogroups.com
      Sent: Friday, December 25, 2009 9:05 AM
      Subject: [cosmacelf] Re: +Viking,+Voyager+firmware+anyone+?



      Plated wire memory and a modified 516 compatible ... Which would fit
      on a desk this I was told by the guy who coded the diagnostic and
      other aspects.

      I agree it doesn't make sense to cram a mini computer in but it seems
      that's what they did.

      On 12/25/09, wa9hsl <wa9hsl@...> wrote:
      > Trevor,
      >
      > I won't belabor the point here about what was used onboard the spacecraft
      > except to say that the URL that you gave points to the DDP-516 which is a
      > fairly large TTL mini-computer with CORE memory. The quoted instruction
      > times match core memory cycle times. The computer is also comparatively
      > heavy and power hungry. Even if you converted the memory configuration to
      > solid state (not specified in the sales folder) the TTL alone in such a
      > machine would have used heavy amounts of current.
      >
      > I really doubt that conventional core memory would have survived the launch
      > much less the radiation. i also really doubt that it was used in flight
      > based on weight and power requirements. It's possible that someone along the
      > line may mixed up the ground control systems with the flight configurations.
      >
      > al - wa9hsl
      > ....
      >
      >
      >
      > --- In cosmacelf@yahoogroups.com, Trevor Nunes <trevor.nunes@...> wrote:
      >>
      >> Hi,
      >>
      >> Yes the later missions used 1802's but for Viking it's a modified
      >> Honeywell 516 with 24bit addressing which if you look at the instruction
      >> set is a different beast all togethor than the COSMAC. I've found a link
      >> to JPL archives that lists archived microfiche identifying 'source
      >> listings' so there's some progress.
      >>
      >> http://archive.computerhistory.org/resources/text/Honeywell/Honeywell.u-COMP_DDP-516.102646115.pdf
      >>
      >> Originally I was convinced it was an 1802 running but it's clear the
      >> Lander used the 516 ... and some much slower processor for the Orbiter
      >> perhaps the same but clocked slower. Having Voyager CCS software is
      >> another goal...
      >> Trev.
      >>
      >
      >
      >

      --
      Sent from my mobile device




      [Non-text portions of this message have been removed]
    • jdripper
      ... Core memory was used in the Apollo Guidance Computer and the Saturn Guidance Computer, as well as in various missile guidance computers. It is quite
      Message 2 of 12 , Dec 25, 2009
        > I really doubt that conventional core memory would have
        > survived the launch

        Core memory was used in the Apollo Guidance Computer and the
        Saturn Guidance Computer, as well as in various missile guidance
        computers. It is quite robust and works very well in space
        applications.

        > much less the radiation.

        Core memory is one of the MOST robust forms of digital memory with
        regard to radiation.

        If plated wire memory was used in the Viking lander, it was probably
        chosen for reasons other than core being insufficiently robust. Plated wire memory was probably of slightly higher density than R/W core memory (though of less density than the "core rope" or "wire braid" form of core memory used for the read-only store of the AGC).

        Eric
      • Trevor Nunes
        Would plated or core memory hold charge if power was lost? ... -- Sent from my mobile device
        Message 3 of 12 , Dec 26, 2009
          Would plated or core memory hold charge if power was lost?


          On 12/26/09, jdripper <eric@...> wrote:
          >> I really doubt that conventional core memory would have
          >> survived the launch
          >
          > Core memory was used in the Apollo Guidance Computer and the
          > Saturn Guidance Computer, as well as in various missile guidance
          > computers. It is quite robust and works very well in space
          > applications.
          >
          >> much less the radiation.
          >
          > Core memory is one of the MOST robust forms of digital memory with
          > regard to radiation.
          >
          > If plated wire memory was used in the Viking lander, it was probably
          > chosen for reasons other than core being insufficiently robust. Plated wire
          > memory was probably of slightly higher density than R/W core memory (though
          > of less density than the "core rope" or "wire braid" form of core memory
          > used for the read-only store of the AGC).
          >
          > Eric
          >
          >
          >

          --
          Sent from my mobile device
        • ajparent1
          ... It was magnetic and yes they hold well without power. Rope core or braided rope core was the ROM of the day. I have a DEC PDP-8f and a PDP-11 both with
          Message 4 of 12 , Dec 26, 2009
            --- In cosmacelf@yahoogroups.com, Trevor Nunes <trevor.nunes@...> wrote:
            >
            > Would plated or core memory hold charge if power was lost?
            >


            It was magnetic and yes they hold well without power.

            Rope core or braided rope core was the "ROM" of the day.
            I have a DEC PDP-8f and a PDP-11 both with core memory and the cool thing is that programs from years or decades ago will still run without reentering. The uncool thing was big, power hungry and
            heat producing for read write core.

            The alternate tech of the day was low density drum or disk
            (typically under 128K characters). As solid state memory was
            still pretty small in 1973 with the 1101 (256bits by 1 static) and
            the 1103 (1024bits by 1 dynamic) plus needing two to three voltages and and some elaborate interfacing. In 1974 the 2102 (1024x1 static)
            was a mere 16 dollars each and the fast part was 1000 ns.


            Allison

            >
            > On 12/26/09, jdripper <eric@...> wrote:
            > >> I really doubt that conventional core memory would have
            > >> survived the launch
            > >
            > > Core memory was used in the Apollo Guidance Computer and the
            > > Saturn Guidance Computer, as well as in various missile guidance
            > > computers. It is quite robust and works very well in space
            > > applications.
            > >
            > >> much less the radiation.
            > >
            > > Core memory is one of the MOST robust forms of digital memory with
            > > regard to radiation.
            > >
            > > If plated wire memory was used in the Viking lander, it was probably
            > > chosen for reasons other than core being insufficiently robust. Plated wire
            > > memory was probably of slightly higher density than R/W core memory (though
            > > of less density than the "core rope" or "wire braid" form of core memory
            > > used for the read-only store of the AGC).
            > >
            > > Eric
            > >
            > >
            > >
            >
            > --
            > Sent from my mobile device
            >
          • Ray Sills
            HI Gang: Not thinking about the song... :)... but the reference to core memory holding it s content in a power failure. Not too many years ago, the CBS
            Message 5 of 12 , Dec 26, 2009
              HI Gang:

              Not thinking about the song... :)... but the reference to core
              memory holding it's content in a power failure. Not too many years
              ago, the CBS Television Network's master assignment switcher (called
              "MAX") used core memory for the very reason that after a power
              failure, it would be -vastly- easier and faster to reconstitute the
              various signal distributions around the Broadcast Center if the prior
              state (before power interruption) could be restored quickly.

              Over time, there were a few power failures of varying scope, and it
              turned out that the engineering decision of the time was a good one.
              The switcher, BTW, used mechanical stepping relays to send video,
              audio and control signals to various locations and devices in the
              plant. Because of the mechanical nature of the relays, the room
              which held the racks containing them, was a "room within a room" and
              under positive air pressure from a well-filtered source. To enter
              the room, you had to go through an airlock.

              The wonder of it all is that it ran so well for so long. There was a
              lot of inherent redundancy, so that if a particular source was
              selected and routed to an output, in most locations in the building,
              you could simply select an alternate source or destination, if the
              one you originally selected was not usable for whatever reason.

              The whole place could probably have been run from a single 1802, as
              far as the switching logic is concerned. :)


              73 de Ray

              On Dec 26, 2009, at 12:39 PM, ajparent1 wrote:

              >
              >
              > --- In cosmacelf@yahoogroups.com, Trevor Nunes <trevor.nunes@...>
              > wrote:
              >>
              >> Would plated or core memory hold charge if power was lost?
              >>
              >
              >
              > It was magnetic and yes they hold well without power.
              >
              > Rope core or braided rope core was the "ROM" of the day.
              > I have a DEC PDP-8f and a PDP-11 both with core memory and the cool
              > thing is that programs from years or decades ago will still run
              > without reentering. The uncool thing was big, power hungry and
              > heat producing for read write core.
              >
              > The alternate tech of the day was low density drum or disk
              > (typically under 128K characters). As solid state memory was
              > still pretty small in 1973 with the 1101 (256bits by 1 static) and
              > the 1103 (1024bits by 1 dynamic) plus needing two to three voltages
              > and and some elaborate interfacing. In 1974 the 2102 (1024x1 static)
              > was a mere 16 dollars each and the fast part was 1000 ns.
              >
              >
              > Allison
              >
              >>
              >> On 12/26/09, jdripper <eric@...> wrote:
              >>>> I really doubt that conventional core memory would have
              >>>> survived the launch
              >>>
              >>> Core memory was used in the Apollo Guidance Computer and the
              >>> Saturn Guidance Computer, as well as in various missile guidance
              >>> computers. It is quite robust and works very well in space
              >>> applications.
              >>>
              >>>> much less the radiation.
              >>>
              >>> Core memory is one of the MOST robust forms of digital memory with
              >>> regard to radiation.
              >>>
              >>> If plated wire memory was used in the Viking lander, it was probably
              >>> chosen for reasons other than core being insufficiently robust.
              >>> Plated wire
              >>> memory was probably of slightly higher density than R/W core
              >>> memory (though
              >>> of less density than the "core rope" or "wire braid" form of core
              >>> memory
              >>> used for the read-only store of the AGC).
              >>>
              >>> Eric
              >>>
              >>>
              >>>
              >>
              >> --
              >> Sent from my mobile device
              >>
              >
              >
              >
              >
              > ------------------------------------
              >
              > ========================================================
              > Visit the COSMAC ELF website at http://www.cosmacelf.comYahoo!
              > Groups Links
              >
              >
              >
            • wa9hsl
              Meaningful core data retention didn t happen overnight but was painstakingly developed. I worked on a recovery solution in 1970 for NSA who would often have
              Message 6 of 12 , Dec 28, 2009
                Meaningful core data retention didn't happen overnight but was painstakingly developed. I worked on a recovery solution in 1970 for NSA who would often have problem programs that ran literally for weeks on end and were vulnerable to power line failures. Nothing was offered by industry at the time and faultless restarts were highly desired. Core drivers and sensing logic experienced random firings as the voltage levels drop below a threshold value and this often dirties up core states. A lot of effort and electronics and a real rom-based interrupt program were required to make it all work together. The electronics was new and a little complicated at the era of time.

                Here are some other facts about core for your consideration:

                Conventional core was very noisy at higher speeds especially since hi-drive current levels were required for core select and inhibit drivers. Remember that you have to switch the core state to read it and then re-write the original state which only adds to the noise level.

                Core was a real high current user, all core that I worked on required constant forced-air cooling and some designs needed oil cooling. The alternative was to run it at very reduced speeds.

                Core sense amps were very vulnerable to noise and radiation since common signal levels were sub milli-volt levels.

                Core power supplies were almost always quite heavy...even nuclear ones.

                Core memory failure (or should I say driving logic) was very high for land based computer systems.

                Many enhancements to core technology came later but that was also about the time that CMOS logic, solid state memory and LSI and even RCA 1800 processors began to hit the scene.

                I sometimes shudder at the various electronic and electromechanical components used in spacecraft and conventional aircraft of the 1950s-1960s and 1970s. The miracle of the era was that some of this stuff actually worked long enough to cause a success to be declared and yet no more human life was lost than was. A lot of it failed too. Take a peak at the Smithsonian space displays someday and you'll shake your head too! I really don't see much "robustness" in all that but rather a lot of sweat and luck to make it all work for a time.

                al
                ....






                --- In cosmacelf@yahoogroups.com, "ajparent1" <kb1gmx@...> wrote:
                >
                >
                >
                > --- In cosmacelf@yahoogroups.com, Trevor Nunes <trevor.nunes@> wrote:
                > >
                > > Would plated or core memory hold charge if power was lost?
                > >
                >
                >
                > It was magnetic and yes they hold well without power.
                >
                > Rope core or braided rope core was the "ROM" of the day.
                > I have a DEC PDP-8f and a PDP-11 both with core memory and the cool thing is that programs from years or decades ago will still run without reentering. The uncool thing was big, power hungry and
                > heat producing for read write core.
                >
                > The alternate tech of the day was low density drum or disk
                > (typically under 128K characters). As solid state memory was
                > still pretty small in 1973 with the 1101 (256bits by 1 static) and
                > the 1103 (1024bits by 1 dynamic) plus needing two to three voltages and and some elaborate interfacing. In 1974 the 2102 (1024x1 static)
                > was a mere 16 dollars each and the fast part was 1000 ns.
                >
                >
                > Allison
                >
                > >
                > > On 12/26/09, jdripper <eric@> wrote:
                > > >> I really doubt that conventional core memory would have
                > > >> survived the launch
                > > >
                > > > Core memory was used in the Apollo Guidance Computer and the
                > > > Saturn Guidance Computer, as well as in various missile guidance
                > > > computers. It is quite robust and works very well in space
                > > > applications.
                > > >
                > > >> much less the radiation.
                > > >
                > > > Core memory is one of the MOST robust forms of digital memory with
                > > > regard to radiation.
                > > >
                > > > If plated wire memory was used in the Viking lander, it was probably
                > > > chosen for reasons other than core being insufficiently robust. Plated wire
                > > > memory was probably of slightly higher density than R/W core memory (though
                > > > of less density than the "core rope" or "wire braid" form of core memory
                > > > used for the read-only store of the AGC).
                > > >
                > > > Eric
                > > >
                > > >
                > > >
                > >
                > > --
                > > Sent from my mobile device
                > >
                >
              • Mark Graybill
                ... The first time I got a chance to look over an old Titan I booster in the late 80s, I laughed. The thing was slapped together like a bench test prototype
                Message 7 of 12 , Dec 28, 2009
                  On Dec 28, 2009, at 4:02 PM, wa9hsl wrote:
                  > I sometimes shudder at the various electronic and electromechanical
                  > components used in spacecraft and conventional aircraft of the
                  > 1950s-1960s and 1970s. The miracle of the era was that some of this
                  > stuff actually worked long enough to cause a success to be declared
                  > and yet no more human life was lost than was. A lot of it failed
                  > too. Take a peak at the Smithsonian space displays someday and
                  > you'll shake your head too! I really don't see much "robustness" in
                  > all that but rather a lot of sweat and luck to make it all work for
                  > a time.
                  >

                  The first time I got a chance to look over an old Titan I booster in
                  the late '80s, I laughed. The thing was slapped together like a bench
                  test prototype (this vehicle had been in service back in the early
                  60's.) There were terminal blocks RTVed to the inside of the airframe,
                  cables run willy-nilly with a few tie-downs placed where ever, what a
                  mess!

                  When I got my first look at electronic and electrical systems from a
                  Mercury capsule I about screamed. By the 80's we wouldn't have allowed
                  stuff built to that standard in a ground-based control room, much less
                  in flight hardware--never mind whether it's manned or not.

                  I've also had a chance to see the difference in the same boxes built
                  before and after Apollo 204. The difference was amazing. The boxes
                  from before looked like cheap commercial home radio equipment in
                  construction quality, with crud left in the enclosure, loose blobs of
                  solder, and the joints were of the "Slap on loads of solder" variety.
                  The stuff from after was very different. Proper prep work, proper
                  trimming of leads, well cleaned, etc. There were still some things I
                  didn't like, such as potential arc sites, metal labels inside the can,
                  connectors that could crush pins without the user knowing it, and so
                  on, but it was far better than the earlier stuff.

                  I can't believe as little life was lost in the early space program as
                  was. That Mercury stuff was incredible, when they said "fire switch"
                  back then, they _meant_ it.

                  -Mark G.
                • thinkpast
                  ... As has been posted, I encourage you to preserve your knowledge and remarks. I DON T encourage you to start a blog. Blogs are unorganized and essentially
                  Message 8 of 12 , Dec 29, 2009
                    --- In cosmacelf@yahoogroups.com, "wa9hsl" <wa9hsl@...> wrote:
                    >
                    >
                    > Meaningful core data retention didn't happen overnight but was painstakingly developed. I worked on a recovery solution in 1970 for NSA who would often have problem programs that ran literally for weeks on end and were vulnerable to power line failures. Nothing was offered by industry at the time and faultless restarts were highly desired. Core drivers and sensing logic experienced random firings as the voltage levels drop below a threshold value and this often dirties up core states. A lot of effort and electronics and a real rom-based interrupt program were required to make it all work together. The electronics was new and a little complicated at the era of time.
                    >
                    > Here are some other facts about core for your consideration:

                    As has been posted, I encourage you to preserve your knowledge and remarks. I DON'T encourage you to start a blog. Blogs are unorganized and essentially sequential discussions without editing. I encourage you instead to start or contribute to a Web page or site. Web pages can be edited, updated, and corrected; and are more coherent. They are not small bits of conversation, which apparently some people today find preferable to boring reading.

                    I'd be glad to provide a page at my domain, retrotechnology.com, on core technology. Contact me if you wish. Otherwise there are a number of sites which discuss core memory, and you can always start your own or use your own if you have a site.


                    > I sometimes shudder at the various electronic and
                    >electromechanical components used in spacecraft and
                    > conventional aircraft of the 1950s-1960s and 1970s.
                    > The miracle of the era was that some of this stuff actually
                    > worked long enough to cause a success to be declared
                    >and yet no more human life was lost than was. A lot of
                    > it failed too.


                    My Web site covers 1970's digital engineering of microcomputers. A lot of that stuff was well built (Multibus for example) but some of it was not. THere was a lot of new technology in the era and price and performace were under a lot of pressures, both high costs and low costs, new chips with new features versus TTL logic to do it all. Early 1970's floppy controller are a horrow-show of design.

                    I know enough about the "space race" to know it was a real race of technology, they used the highest-end stuff they could at times. But with lead times of several years that sometimes meant older tech or odd tech was put to use. Or, tech from previous missions and craft.

                    Much of this was new stuff in any event, there were no standard designs or implementations. That gets forgotten today, where a few designs and processors dominate personal computing. That's less so in micro controllers and microprocessors, but there's whole families of such chips and one works in a "family" to get things done.

                    The 1802 in that context is still amazing. It's still a low power champ. It's still a relatively straightforward structure, and the technology is mostly still familiar to anyone working with 8-bit processors today that are not "superintegrated". So many embedded computers today, have chips in them where half the logic and features are unused, and fractions of the instruction sets are unused. So there's a kind of "elegance" in using a decades-old processor in simple applications, or just for fun.

                    Herb Johnson
                    retrotechnology.com
                    contact me c/o my Web site, please
                  Your message has been successfully submitted and would be delivered to recipients shortly.