Identify this mystery DEC control panel - DEC

This is a discussion on Identify this mystery DEC control panel - DEC ; http://bitsavers.org/mysteryPanel_Nov74.jpg This is part of the Computer History Museum collection that no one, including Gordon Bell, has been able to identify. It appears to be for a 32 bit computer with a 360-like architecture, but with Unibus control signals. It ...

+ Reply to Thread
Page 1 of 3 1 2 3 LastLast
Results 1 to 20 of 51

Thread: Identify this mystery DEC control panel

  1. Identify this mystery DEC control panel

    http://bitsavers.org/mysteryPanel_Nov74.jpg

    This is part of the Computer History Museum collection that no one, including
    Gordon Bell, has been able to identify. It appears to be for a 32 bit computer
    with a 360-like architecture, but with Unibus control signals. It would appear
    to be for a PDP-15 style front panel bezel (two rows of switches) but is using
    the PDP-11 color scheme.


    --
    Posted via a free Usenet account from http://www.teranews.com


  2. Re: Identify this mystery DEC control panel


    "Al Kossow" wrote in message
    news:46114284$0$16355$88260bb3@free.teranews.com.. .
    > http://bitsavers.org/mysteryPanel_Nov74.jpg
    >
    > This is part of the Computer History Museum collection that no one,
    > including
    > Gordon Bell, has been able to identify. It appears to be for a 32 bit
    > computer
    > with a 360-like architecture, but with Unibus control signals. It would
    > appear
    > to be for a PDP-15 style front panel bezel (two rows of switches) but is
    > using
    > the PDP-11 color scheme.
    >
    >
    > --
    > Posted via a free Usenet account from http://www.teranews.com


    Not a clue, but pretty darned neat to see a 32-bit machine from DEC that
    predates VAX by four years.

    Internal prototype?



  3. Re: Identify this mystery DEC control panel

    On Mon, 2 Apr 2007, Al Kossow wrote:
    > It appears to be for a 32 bit computer with a 360-like architecture, but
    > with Unibus control signals. It would appear to be for a PDP-15 style
    > front panel bezel (two rows of switches) but is using the PDP-11 color
    > scheme.


    It doesn't look very S/360 ish to me. The instruction word is fixed
    format and not at all byte oriented. There's a 1-bit I bit (perhaps an
    indirect), 7-bit opcode, 4-bit register, 3-bit index address, and 17-bit
    memory address. Instruction addresses are also 17-bit. There's another
    20-bit address but I can't read what it says.

    My guess is that it is for some customer-built custom CPU, albeit with
    heavy DEC influence. Another possibility is that it was a pre-VAX
    prototype of a 32-bit DEC machine, perhaps from the PDP-15 team?

    -- Mark --

    http://staff.washington.edu/mrc
    Science does not emerge from voting, party politics, or public debate.
    Si vis pacem, para bellum.

  4. Re: Identify this mystery DEC control panel

    On Mon, 02 Apr 2007 17:10:17 -0700, Mark Crispin
    wrote:

    > On Mon, 2 Apr 2007, Al Kossow wrote:
    >> It appears to be for a 32 bit computer with a 360-like architecture,
    >> but with Unibus control signals. It would appear to be for a PDP-15
    >> style front panel bezel (two rows of switches) but is using the PDP-11
    >> color scheme.

    >
    > It doesn't look very S/360 ish to me. The instruction word is fixed
    > format and not at all byte oriented. There's a 1-bit I bit (perhaps an
    > indirect), 7-bit opcode, 4-bit register, 3-bit index address, and 17-bit
    > memory address. Instruction addresses are also 17-bit. There's another
    > 20-bit address but I can't read what it says.


    Note bit 15 of the address in the second row, this suggest to me that it
    was
    likely a 16 bit address with bit 15 indicating a mode/privilege

    >
    > My guess is that it is for some customer-built custom CPU, albeit with
    > heavy DEC influence. Another possibility is that it was a pre-VAX
    > prototype of a 32-bit DEC machine, perhaps from the PDP-15 team?
    >
    > -- Mark --
    >
    > http://staff.washington.edu/mrc
    > Science does not emerge from voting, party politics, or public debate.
    > Si vis pacem, para bellum.




    --
    Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

  5. Re: Identify this mystery DEC control panel

    Mark Crispin wrote:
    > On Mon, 2 Apr 2007, Al Kossow wrote:
    >> It appears to be for a 32 bit computer with a 360-like architecture,
    >> but with Unibus control signals. It would appear to be for a PDP-15
    >> style front panel bezel (two rows of switches) but is using the PDP-11
    >> color scheme.

    >
    > It doesn't look very S/360 ish to me. The instruction word is fixed
    > format and not at all byte oriented. There's a 1-bit I bit (perhaps an
    > indirect), 7-bit opcode, 4-bit register, 3-bit index address, and 17-bit
    > memory address. Instruction addresses are also 17-bit. There's another
    > 20-bit address but I can't read what it says.
    >
    > My guess is that it is for some customer-built custom CPU, albeit with
    > heavy DEC influence. Another possibility is that it was a pre-VAX
    > prototype of a 32-bit DEC machine, perhaps from the PDP-15 team?


    What makes you think it is even a DEC panel at all? The only signals
    that are vaguely DEC UNIBUS-ish are MSYN/SSYN, but these are pretty
    generic. It is obviously for a hexadecimal oriented architecture with
    true big-endian IBM style bit ordering (ie, MSB is bit zero) because
    the switches/lights are shaded in groups of four.

    None of the labels/acronyms ring any bells for my recollection of any
    PDP-x internal signals, registers, etc.

    I would guess this is not DEC, but some other architecture.

  6. Re: Identify this mystery DEC control panel

    On Tue, 3 Apr 2007, Don North wrote:
    > What makes you think it is even a DEC panel at all?


    Because it looks exactly what a 36-bit or 18-bit person would have done if
    he was told to build a 32-bit machine. Note that the 18-bit, 12-bit, and
    36-bit product lines were all big-endian.

    > It is obviously for a hexadecimal oriented architecture with
    > true big-endian IBM style bit ordering (ie, MSB is bit zero) because
    > the switches/lights are shaded in groups of four.


    All of which is consistant with what I wrote above.

    It's only semi-hexadecimal oriented. Note the 17-bit address, 3-bit
    index, and 7-bit opcode fields. All very much in line with what someone
    in the 18-bit/36-bit community would do, having no respect at all for
    bytes and nybbles...

    -- Mark --

    http://staff.washington.edu/mrc
    Science does not emerge from voting, party politics, or public debate.
    Si vis pacem, para bellum.

  7. Re: Identify this mystery DEC control panel

    The label in the upper left hand corner states "Special Order PDP/15 Order
    11/20/74" and looks fairly well aged. I'm guessing that was a guess from
    20 years ago from someone who was wondering what the heck it was for.

    Though the accuracy of 11/20/74 is peculiar. Maybe it was stamped / written
    somewhere else on the piece?

  8. Re: Identify this mystery DEC control panel

    In article <46114284$0$16355$88260bb3@free.teranews.com>, Al Kossow writes:
    > http://bitsavers.org/mysteryPanel_Nov74.jpg
    >
    > This is part of the Computer History Museum collection that no one, including
    > Gordon Bell, has been able to identify. It appears to be for a 32 bit computer
    > with a 360-like architecture, but with Unibus control signals. It would appear
    > to be for a PDP-15 style front panel bezel (two rows of switches) but is using
    > the PDP-11 color scheme.


    I note that the tag in the upper left hand corner has "PDP/15"
    scrawled on it.

    What I see is a fixed instruction layout similar to the PDP-10 except
    32 bit. Is this "address and a half" layout what you mean by 360-like
    (opcode, register, and address)?

    I thought my green card showed multiple instruction layouts for 360.

    Maybe it was an internal investigation into a possible future 32 bit
    system using ideas other than PDP-11 heritage.

    The Smithsonian had a blue PDP-15 on display in the Museum of
    American History, but that's closed for massive renovation now so
    we can't just drop in and compare. Maybe someone has another shot
    of a PDP-15?



  9. Re: Identify this mystery DEC control panel

    On 3 Apr 2007 08:02:24 -0500, koehler@eisner.nospam.encompasserve.org
    (Bob Koehler) wrote:

    >In article <46114284$0$16355$88260bb3@free.teranews.com>, Al Kossow writes:
    >> http://bitsavers.org/mysteryPanel_Nov74.jpg
    >>
    >> This is part of the Computer History Museum collection that no one, including
    >> Gordon Bell, has been able to identify. It appears to be for a 32 bit computer
    >> with a 360-like architecture, but with Unibus control signals. It would appear
    >> to be for a PDP-15 style front panel bezel (two rows of switches) but is using
    >> the PDP-11 color scheme.

    >
    > I note that the tag in the upper left hand corner has "PDP/15"
    > scrawled on it.


    The tag also says, "special order" and in the lower right corner
    "S/S". There was a group at DEC called Special Systems that did
    one-offs and short run items. Perhaps it came from there.


    --
    jeverett3sbcglobalnet (John V. Everett)

  10. Re: Identify this mystery DEC control panel

    On Tue, 3 Apr 2007, Bob Koehler wrote:
    > What I see is a fixed instruction layout similar to the PDP-10 except
    > 32 bit. Is this "address and a half" layout what you mean by 360-like
    > (opcode, register, and address)?
    > I thought my green card showed multiple instruction layouts for 360.


    Yes. S/360 had multiple instruction layouts and lengths (in terms of
    number of bytes).

    > Maybe it was an internal investigation into a possible future 32 bit
    > system using ideas other than PDP-11 heritage.


    That's exactly what I am thinking. I wonder if the 1974 date is
    misleading, and in fact it is from several years earlier; perhaps even an
    Edson de Castro design (which would put it into the 1960s).

    -- Mark --

    http://staff.washington.edu/mrc
    Science does not emerge from voting, party politics, or public debate.
    Si vis pacem, para bellum.

  11. Re: Identify this mystery DEC control panel

    On Mon, 02 Apr 2007 11:44:25 -0700, Al Kossow wrote:

    >http://bitsavers.org/mysteryPanel_Nov74.jpg
    >
    >This is part of the Computer History Museum collection that no one, including
    >Gordon Bell, has been able to identify. It appears to be for a 32 bit computer
    >with a 360-like architecture, but with Unibus control signals. It would appear
    >to be for a PDP-15 style front panel bezel (two rows of switches) but is using
    >the PDP-11 color scheme.


    What's printed on the back of this panel?

  12. Re: Identify this mystery DEC control panel

    Mark Crispin wrote:
    > It doesn't look very S/360 ish to me. The instruction word is fixed
    > format and not at all byte oriented. There's a 1-bit I bit (perhaps an
    > indirect), 7-bit opcode, 4-bit register, 3-bit index address, and 17-bit
    > memory address. Instruction addresses are also 17-bit. There's another
    > 20-bit address but I can't read what it says.


    You're absolutely right. It is the instruction set of the SDS Sigma (looking
    at the manuals on bitsavers) Fixed 32 bit instructions made the light bulb
    go on.

    There were a few Sigma clone makers in the mid-70's, but it's odd that
    Special Systems would be involved. Some sort of weird Sigma/PDP-11 hybrid
    ala Unichannel-15?

    Oh well, maybe someone will see the posting and have some more info some
    day..



    --
    Posted via a free Usenet account from http://www.teranews.com


  13. Re: Identify this mystery DEC control panel

    On Tue, 03 Apr 2007 15:15:49 GMT, I
    wrote:

    >On 3 Apr 2007 08:02:24 -0500, koehler@eisner.nospam.encompasserve.org
    >(Bob Koehler) wrote:
    >
    >>In article <46114284$0$16355$88260bb3@free.teranews.com>, Al Kossow writes:
    >>> http://bitsavers.org/mysteryPanel_Nov74.jpg
    >>>
    >>> This is part of the Computer History Museum collection that no one, including
    >>> Gordon Bell, has been able to identify. It appears to be for a 32 bit computer
    >>> with a 360-like architecture, but with Unibus control signals. It would appear
    >>> to be for a PDP-15 style front panel bezel (two rows of switches) but is using
    >>> the PDP-11 color scheme.

    >>
    >> I note that the tag in the upper left hand corner has "PDP/15"
    >> scrawled on it.

    >
    >The tag also says, "special order" and in the lower right corner
    >"S/S". There was a group at DEC called Special Systems that did
    >one-offs and short run items. Perhaps it came from there.


    After posting the above I've had a look at the panel again and I'm
    wondering if it's from DEC at all. It has a display for Program Status
    Word, but I don't recall ever hearing that term at DEC. We usually
    used PC (Program Counter). I also recall SC (Sequence Counter) but I
    don't think that was a DEC term. PSW sounds like IBM to me.

    Also on the panel there's "Memory Port" and three abbreviations under
    it, PROC (Processor), DCH (Data Channel), and BKM (Block Multiplexor).
    These are also very IBMish terms.

    --
    jeverett3sbcglobalnet (John V. Everett)

  14. Re: Identify this mystery DEC control panel

    On Mon, 02 Apr 2007 23:26:51 -0700, Mark Crispin
    wrote:

    > On Tue, 3 Apr 2007, Don North wrote:
    >> What makes you think it is even a DEC panel at all?

    >
    > Because it looks exactly what a 36-bit or 18-bit person would have done
    > if he was told to build a 32-bit machine. Note that the 18-bit, 12-bit,
    > and 36-bit product lines were all big-endian.
    >
    >> It is obviously for a hexadecimal oriented architecture with
    >> true big-endian IBM style bit ordering (ie, MSB is bit zero) because
    >> the switches/lights are shaded in groups of four.

    >
    > All of which is consistant with what I wrote above.
    >
    > It's only semi-hexadecimal oriented. Note the 17-bit address, 3-bit
    > index, and 7-bit opcode fields. All very much in line with what someone
    > in the 18-bit/36-bit community would do, having no respect at all for
    > bytes and nybbles...


    EABX in upper right, would that be Extended Address Base Register?

    >
    > -- Mark --
    >
    > http://staff.washington.edu/mrc
    > Science does not emerge from voting, party politics, or public debate.
    > Si vis pacem, para bellum.




    --
    Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

  15. Re: Identify this mystery DEC control panel

    On Tue, 03 Apr 2007 08:56:59 -0700, Al Kossow wrote:

    > Mark Crispin wrote:
    >> It doesn't look very S/360 ish to me. The instruction word is fixed
    >> format and not at all byte oriented. There's a 1-bit I bit (perhaps an
    >> indirect), 7-bit opcode, 4-bit register, 3-bit index address, and
    >> 17-bit memory address. Instruction addresses are also 17-bit. There's
    >> another 20-bit address but I can't read what it says.

    >
    > You're absolutely right. It is the instruction set of the SDS Sigma
    > (looking
    > at the manuals on bitsavers) Fixed 32 bit instructions made the light
    > bulb
    > go on.


    I was also think about those, CII 10070 in France was a clone of the Sigma
    7,
    and they weree active in 1974 (Compagnie Informatique internationale later
    merged with Honeywell-Bull)
    >
    > There were a few Sigma clone makers in the mid-70's, but it's odd that
    > Special Systems would be involved. Some sort of weird Sigma/PDP-11 hybrid
    > ala Unichannel-15?
    >
    > Oh well, maybe someone will see the posting and have some more info some
    > day..
    >
    >
    >




    --
    Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

  16. Re: Identify this mystery DEC control panel

    On Tue, 03 Apr 2007 15:15:49 GMT,
    wrote:
    > (. . .)
    > After posting the above I've had a look at the panel again
    > and I'm wondering if it's from DEC at all.
    > It has a display for Program Status Word,
    > but I don't recall ever hearing that term at DEC.


    Mark Crispin wrote:
    > (. . .)
    > perhaps even an Edson de Castro design (which would put it into the 1960s).


    Interesting.
    The only computer other than the PDP-8 & -12 that I worked on (at the
    console)
    for any length of time was the Data General Nova series.
    And I know I used the term "Program Status Word" in some part of my career.
    I don't (think) I have any of my DG Nova books around anymore,
    but maybe someone else could look into that angle.

  17. Re: Identify this mystery DEC control panel

    In message
    scenic_man wrote:

    > On Tue, 03 Apr 2007 15:15:49 GMT,
    > wrote:
    >> (. . .)
    >> After posting the above I've had a look at the panel again
    >> and I'm wondering if it's from DEC at all.
    >> It has a display for Program Status Word,
    >> but I don't recall ever hearing that term at DEC.

    >
    > Mark Crispin wrote:
    >> (. . .)
    >> perhaps even an Edson de Castro design (which would put it into the 1960s).

    >
    > Interesting.
    > The only computer other than the PDP-8 & -12 that I worked on (at the
    > console)
    > for any length of time was the Data General Nova series.
    > And I know I used the term "Program Status Word" in some part of my career.
    > I don't (think) I have any of my DG Nova books around anymore,
    > but maybe someone else could look into that angle.


    Interesting, because when I looked at the panel it seemed slightly
    familiar - a CALMA graphics workstation using a DG Nova or Eclipse (we
    had both at different times). I'm not saying it IS one of those, but
    the colour scheme looked very like it.



    --
    Alan Adams, from Northamptonshire
    alan.adams@orchard-way.freeserve.co.uk
    http://www.nckc.org.uk/

  18. Re: Identify this mystery DEC control panel

    koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:

    > The Smithsonian had a blue PDP-15 on display in the Museum of American
    > History, but that's closed for massive renovation now so we can't just
    > drop in and compare. Maybe someone has another shot of a PDP-15?


    There is at least one PDP-15 (I remember seeing at least 3, back in the blimp
    hangar days) at the Computer History Museum, so Al and company certainly have
    access to that for comparison.

    --
    Rich Alderson | /"\ ASCII ribbon |
    news@alderson.users.panix.com | \ / campaign against |
    "You get what anybody gets. You get a lifetime." | x HTML mail and |
    --Death, of the Endless | / \ postings |

  19. Re: Identify this mystery DEC control panel

    On Apr 3, 11:00 am, "Tom Linden" wrote:
    > On Tue, 03 Apr 2007 08:56:59 -0700, Al Kossow wrote:
    > > Mark Crispin wrote:
    > >> It doesn't look very S/360 ish to me. The instruction word is fixed
    > >> format and not at all byte oriented. There's a 1-bit I bit (perhaps an
    > >> indirect), 7-bit opcode, 4-bit register, 3-bit index address, and
    > >> 17-bit memory address. Instruction addresses are also 17-bit. There's
    > >> another 20-bit address but I can't read what it says.

    >
    > > You're absolutely right. It is the instruction set of the SDS Sigma
    > > (looking
    > > at the manuals on bitsavers) Fixed 32 bit instructions made the light
    > > bulb
    > > go on.

    >
    > I was also think about those, CII 10070 in France was a clone of the Sigma
    > 7,
    > and they weree active in 1974 (Compagnie Informatique internationale later
    > merged with Honeywell-Bull)
    >
    >
    >
    > > There were a few Sigma clone makers in the mid-70's, but it's odd that
    > > Special Systems would be involved. Some sort of weird Sigma/PDP-11 hybrid
    > > ala Unichannel-15?

    >
    > > Oh well, maybe someone will see the posting and have some more info some
    > > day..

    >
    > --
    > Using Opera's revolutionary e-mail client:http://www.opera.com/mail/


    Could it be from a DECsystem-10 (KL or KI-10) processor...


  20. Re: Identify this mystery DEC control panel

    Al Kossow wrote:
    > Mark Crispin wrote:
    > > It doesn't look very S/360 ish to me. The instruction word is fixed
    > > format and not at all byte oriented. There's a 1-bit I bit (perhaps an
    > > indirect), 7-bit opcode, 4-bit register, 3-bit index address, and 17-bit
    > > memory address. Instruction addresses are also 17-bit. There's another
    > > 20-bit address but I can't read what it says.

    >
    > You're absolutely right. It is the instruction set of the SDS Sigma (looking
    > at the manuals on bitsavers) Fixed 32 bit instructions made the light bulb
    > go on.
    >
    > There were a few Sigma clone makers in the mid-70's, but it's odd that
    > Special Systems would be involved. Some sort of weird Sigma/PDP-11 hybrid
    > ala Unichannel-15?


    My only insights about Sigma clones:

    1. Telefile made Sigma compatible memory systems that essentially
    converted the Sigma wiring to Unibus and then slapped their DEC-
    compatible memory system in there. While the telefile cabinetry was
    pretty spiffy (I really liked the transparent cabinet fronts), I NEVER
    saw anything like a CPU-front-panel, it was all memory exercisor stuff
    from them.

    2. Belobox made Sigma clone CPU's, but the ones I know about were from
    the 80's and at least sometimes had 680x0's under the hoods. Googling
    for "Belobox" produces web pages that have very little to nothing to
    do with the company we did business with (although it was in LA too,
    and maybe a common thread is emphasizing business with aerospace/
    defense companies, which were big Sigma-compatible customers long
    after SDS/XDS was gone!)

    Tim.


+ Reply to Thread
Page 1 of 3 1 2 3 LastLast