• 187-047 IBM SYSTEM 36/38 WORKSTATION EMULATION ADAPTER/A

    From Louis Ohland@21:1/5 to All on Sun Oct 31 19:19:46 2021
    System 36/38 Workstation Emulation Adapter/A 187-047

    https://www.ibm.com/common/ssi/ShowDoc.wss?docURL=/common/ssi/rep_ca/7/897/ENUS187-047/index.html

    DESCRIPTION
    The IBM System 36/38 Workstation Emulation Adapter/A and
    System 36/38 Attachment Cable allow IBM Personal System/2 Models 50,
    60, and 80 to emulate either 5250 Displays or 5250 Printers. The
    adapter is attached to the twinax cable used in System 36/38
    environments. The adapter, along with the IBM System 36/38
    Workstation Emulation Program, Version 1.0, allows access to the
    application programs on the System 36/38 as if the user were at a
    5250 workstation. As a printer, the user can print the output of
    System 36/38 application programs on a printer attached to the IBM
    Personal System/2. The IBM System 36/38 Workstation Emulation
    Program provides limited multitasking capabilities that allow the IBM
    Personal System/2 to interact with the System 36/38 as if it were
    four separate terminals or printers. The sessions can be any
    combination of printer, display, or host graphics sessions that do
    not exceed four sessions total. That is, the user can easily switch
    between up to four different System 36/38 tasks, as well as one DOS
    task.

    An 8088 processor is provided on the IBM System 36/38
    Workstation Emulation Adapter/A reducing the timing dependencies on
    the IBM Personal System/2.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Grant Taylor@21:1/5 to Louis Ohland on Sun Oct 31 22:39:30 2021
    On 10/31/21 6:19 PM, Louis Ohland wrote:
    Personal System/2 ... System 36/38

    It wasn't until I was reading this message that I realized that the
    smaller single user system was referred to as a /Personal/ System as
    opposed to just a System which is used by multiple people.

    Interesting twist. I'm not at all surprised by that. It's so subtle
    IBM that it seems to fit to me.



    --
    Grant. . . .
    unix || die

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From IBMMuseum@21:1/5 to All on Mon Nov 1 11:49:25 2021
    Personal System/2 ... System 36/38

    It wasn't until I was reading this message that I realized that the
    smaller single user system was referred to as a /Personal/ System as
    opposed to just a System which is used by multiple people.

    Interesting twist. I'm not at all surprised by that. It's so subtle
    IBM that it seems to fit to me.

    Neat determination - The System/23 ("Datamaster") has a common origin with the 5120 (following the 5110 to start that series, and released only a month before the 5150 IBM PC). Starting the "mid-range" category, the System/32 was only one user at a time,
    then the System/34 and later started having twin-ax terminals. And the later software to link a "personal" computer to an AS/400 or mainframe host was named "Personal Communications".

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to All on Mon Nov 1 13:22:21 2021
    http://ps-2.kev009.com/pcpartnerinfo/ctstips/9486.htm

    Dave, are these the ones you got?

    https://ibmmuseum.com/Adapters/5250/IBM/MCA/3638FFF8/

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From IBMMuseum@21:1/5 to Louis Ohland on Mon Nov 1 11:50:54 2021
    On Monday, November 1, 2021 at 12:23:16 PM UTC-6, Louis Ohland wrote:
    http://ps-2.kev009.com/pcpartnerinfo/ctstips/9486.htm

    Dave, are these the ones you got?

    https://ibmmuseum.com/Adapters/5250/IBM/MCA/3638FFF8/

    Despite the later date, it is still "Version 1.0", so I would assume so...

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to Louis Ohland on Mon Nov 1 20:26:11 2021
    https://ibmmuseum.com/Adapters/5250/IBM/MCA/3638FFF8/3638FFF8_Top.jpg

    Lower DIP night be an Intel 8259A

    https://ardent-tool.com/communications/5250.html

    Upper left chip 8088-2 ?

    Later 5250 uses 3 256KB DRAM, I >assume< two are for data, one parity.

    Both 36/38 and 5250 use the LSD soaked sugar cube, 7378945

    36/38 metal caps 59X3380, 595949, 5959949

    What does a workstation emulator have to do? Perhaps keep four sessions
    alive, sordid out the text display, and what? Trying to figure out the need/purpose of the big three. One is not like the others.

    Sure wish we had the 36/38 workstation emulator tech ref.


    On 11/1/2021 20:08, Louis Ohland wrote:
    I was Trolling the Great Moravian Hope on 8x 8088s. Not sure if they
    could control the bus, nor do I think they could juggle even 16MBs.

    Maybe V20s, right?

    Didn't notice much for special sauce controllers, other than three IBM
    silver caps, one sugar cube, nothing exotic.

    On 11/1/2021 19:54, IBMMuseum wrote:
    Yeah, can't read the two intel DIPs either.

    Dave's photos are usually ok, but I think the camera is doing some
    rather heavy post-processing on the photos. High contrast detail -
    especially text - often looks unnatural/wonky, and low contrast detail
    is smudged, more than it should be. Strange...

    And they were likely hurried - I don't think I posted them to Facebook
    and then re-downloaded them for my webspace, but who knows. That said,
    I'm trying to get a better sort of photo rig here and be more
    concerned about having the boards cleaned better.



    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to All on Mon Nov 1 21:13:18 2021
    36/38 metal caps 59X3380, 5959949, 5959949

    The two silver capped chips on the right are the same.

    ESD? Is that "Entry Systems Division" or the much more mundane "Electro
    Static Discharge"?

    36/38 metal caps 59X3380, 595949, 5959949

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From IBMMuseum@21:1/5 to Louis Ohland on Mon Nov 1 19:02:11 2021
    On Monday, November 1, 2021 at 7:27:06 PM UTC-6, Louis Ohland wrote:
    https://ibmmuseum.com/Adapters/5250/IBM/MCA/3638FFF8/3638FFF8_Top.jpg

    Lower DIP night be an Intel 8259A

    https://ardent-tool.com/communications/5250.html

    Upper left chip 8088-2 ?

    Later 5250 uses 3 256KB DRAM, I >assume< two are for data, one parity.

    Both 36/38 and 5250 use the LSD soaked sugar cube, 7378945

    36/38 metal caps 59X3380, 595949, 5959949

    What does a workstation emulator have to do? Perhaps keep four sessions alive, sordid out the text display, and what? Trying to figure out the need/purpose of the big three. One is not like the others.

    Sure wish we had the 36/38 workstation emulator tech ref.
    On 11/1/2021 20:08, Louis Ohland wrote:
    I was Trolling the Great Moravian Hope on 8x 8088s. Not sure if they
    could control the bus, nor do I think they could juggle even 16MBs.

    Maybe V20s, right?

    Didn't notice much for special sauce controllers, other than three IBM silver caps, one sugar cube, nothing exotic.

    On 11/1/2021 19:54, IBMMuseum wrote:
    Yeah, can't read the two intel DIPs either.

    Dave's photos are usually ok, but I think the camera is doing some
    rather heavy post-processing on the photos. High contrast detail -
    especially text - often looks unnatural/wonky, and low contrast detail >>> is smudged, more than it should be. Strange...

    And they were likely hurried - I don't think I posted them to Facebook
    and then re-downloaded them for my webspace, but who knows. That said,
    I'm trying to get a better sort of photo rig here and be more
    concerned about having the boards cleaned better.



    Correct, on the 36/38 WSEA it is an 8259A-2 below the 8088-2

    The 40-pin DIP on the "short" adapter is also an 8088-2

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to Louis Ohland on Mon Nov 1 21:15:50 2021
    Forgot, looks to be a VR near the DB15. Why do we need a voltage
    regulator by a Twinax port?

    Enough for one day.

    On 11/1/2021 21:13, Louis Ohland wrote:
    36/38 metal caps 59X3380, 5959949, 5959949

    The two silver capped chips on the right are the same.

    ESD? Is that "Entry Systems Division" or the much more mundane "Electro Static Discharge"?

    36/38 metal caps 59X3380, 595949, 5959949

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From IBMMuseum@21:1/5 to All on Mon Nov 1 19:29:56 2021
    Forgot, looks to be a VR near the DB15. Why do we need a voltage
    regulator by a Twinax port?

    Yes, all the twin-ax terminals and most of the emulation adapters have a voltage regulator there - there is a requirement to terminate a twin-ax chain.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to IBMMuseum on Tue Nov 2 09:15:20 2021
    So, thinking of SCSI termination -AND- TERMPWR, I'm curious.

    So, my low level of curiosity imagines the Twinax cable is Coax, and my
    DIMM memories of 10base2 have terminators.

    So how does this VR automagically terminate a workstation emulator at
    the end of the chain? Remember that "we can never break the chain"....

    I wendt through the Twinax cable guide, and it must >assume< the VR is
    working correctly, or not. In case of certain errors, replace the
    adapter... No explanation...

    On 11/1/2021 21:29, IBMMuseum wrote:
    Forgot, looks to be a VR near the DB15. Why do we need a voltage
    regulator by a Twinax port?

    Yes, all the twin-ax terminals and most of the emulation adapters have a voltage regulator there - there is a requirement to terminate a twin-ax chain.


    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to IBMMuseum on Tue Nov 2 09:26:01 2021
    https://web.archive.org/web/20110615181859/http://www.nalanda.nitc.ac.in/industry/appnotes/Natsemi/AN-516.pdf

    The IBM adapter does not use the Nat Semi chips, but instructive, somewhat.

    Termination in this document is with resistors to the cable sheath?

    Not an answer.

    On 11/1/2021 21:29, IBMMuseum wrote:
    Forgot, looks to be a VR near the DB15. Why do we need a voltage
    regulator by a Twinax port?

    Yes, all the twin-ax terminals and most of the emulation adapters have a voltage regulator there - there is a requirement to terminate a twin-ax chain.


    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From IBMMuseum@21:1/5 to All on Tue Nov 2 10:13:31 2021
    IBM incorporated "auto" termination in later twin-ax pigtails and had separate terminator caps that could be used. Maybe an IBM tech that installed the 8-bit emulation adapters could provide insight? Jose, can you see?

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to IBMMuseum on Tue Nov 2 14:49:25 2021
    Dave, what systems that used Twinax actually have technical
    documentation which -MIGHT- explain the actual card-level functions?

    Anyone know of a well written Twinax paper?

    On 11/2/2021 12:13, IBMMuseum wrote:
    IBM incorporated "auto" termination in later twin-ax pigtails and had separate terminator caps that could be used. Maybe an IBM tech that installed the 8-bit emulation adapters could provide insight? Jose, can you see?


    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Tomas Slavotinek@21:1/5 to IBMMuseum on Tue Nov 2 21:06:09 2021
    On 2.11.2021 18:13, IBMMuseum wrote:
    IBM incorporated "auto" termination in later twin-ax pigtails and had separate terminator caps that could be used. Maybe an IBM tech that installed the 8-bit emulation adapters could provide insight? Jose, can you see?

    Many of the later twinax terminal cables are indeed auto-terminating.
    Both ends of the cable are typically terminated by default (by shunting
    the line to the ground/shield via two resistors). These resistors get disconnected (mechanically) when you plug something into the connector...

    No additional power source is needed for the line termination. I also
    don't think that these later terminals and terminal emulator cards are
    capable of terminating the line. Not sure though.

    If the TO-220 part really is a regulator, it probably supplies power for
    the line interface chip 7378945(ESD). Negative voltage supply? What does
    the package say? Probulation needed...

    The four diodes on the adapter card are probably surge protection.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Tomas Slavotinek@21:1/5 to Tomas Slavotinek on Tue Nov 2 21:22:36 2021
    Here:

    https://priorart.ip.com/IPCOM/000099938

    This describes it a little bit better and with illustrations.

    What they show there is the kind of cable you would normally use with
    the later terminals/emulator boards.

    On 2.11.2021 21:06, Tomas Slavotinek wrote:
    On 2.11.2021 18:13, IBMMuseum wrote:
    IBM incorporated "auto" termination in later twin-ax pigtails and had
    separate terminator caps that could be used. Maybe an IBM tech that
    installed the 8-bit emulation adapters could provide insight? Jose,
    can you see?

    Many of the later twinax terminal cables are indeed auto-terminating.
    Both ends of the cable are typically terminated by default (by shunting
    the line to the ground/shield via two resistors). These resistors get disconnected (mechanically) when you plug something into the connector...

    No additional power source is needed for the line termination. I also
    don't think that these later terminals and terminal emulator cards are capable of terminating the line. Not sure though.

    If the TO-220 part really is a regulator, it probably supplies power for
    the line interface chip 7378945(ESD). Negative voltage supply? What does
    the package say? Probulation needed...

    The four diodes on the adapter card are probably surge protection.


    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Tomas Slavotinek@21:1/5 to Louis Ohland on Tue Nov 2 21:19:04 2021
    Yes, it should be possible to repurpose the 8088 and the interrupt
    controller to do some other tasks. Though, the subsystem is rather
    limited, especially when it comes to its I/O capabilities.

    Eight 5250 co-pro cards? Also possible but good luck finding some
    application for such a setup :).

    But realistically the card could be useful for debugging, triggering, etc.

    What does it do normally? Its primary task is to receive and send data/commands. On the emulator boards it also has to communicate with
    the host PS/2 system. In the dedicated terminal boxes it runs the main
    "control program" (processing user input/output and what not).

    On 2.11.2021 2:26, Louis Ohland wrote:
    https://ibmmuseum.com/Adapters/5250/IBM/MCA/3638FFF8/3638FFF8_Top.jpg

    Lower DIP night be an Intel 8259A

    https://ardent-tool.com/communications/5250.html

    Upper left chip 8088-2 ?

    Later 5250 uses 3 256KB DRAM, I >assume< two are for data, one parity.

    Both 36/38 and 5250 use the LSD soaked sugar cube, 7378945

    36/38 metal caps 59X3380, 595949, 5959949

    What does a workstation emulator have to do? Perhaps keep four sessions alive, sordid out the text display, and what? Trying to figure out the need/purpose of the big three. One is not like the others.

    Sure wish we had the 36/38 workstation emulator tech ref.


    On 11/1/2021 20:08, Louis Ohland wrote:
    I was Trolling the Great Moravian Hope on 8x 8088s. Not sure if they
    could control the bus, nor do I think they could juggle even 16MBs.

    Maybe V20s, right?

    Didn't notice much for special sauce controllers, other than three IBM
    silver caps, one sugar cube, nothing exotic.

    On 11/1/2021 19:54, IBMMuseum wrote:
    Yeah, can't read the two intel DIPs either.

    Dave's photos are usually ok, but I think the camera is doing some
    rather heavy post-processing on the photos. High contrast detail -
    especially text - often looks unnatural/wonky, and low contrast detail >>>> is smudged, more than it should be. Strange...

    And they were likely hurried - I don't think I posted them to
    Facebook and then re-downloaded them for my webspace, but who knows.
    That said, I'm trying to get a better sort of photo rig here and be
    more concerned about having the boards cleaned better.




    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From IBMMuseum@21:1/5 to Louis Ohland on Tue Nov 2 13:37:26 2021
    On Tuesday, November 2, 2021 at 1:50:20 PM UTC-6, Louis Ohland wrote:
    Dave, what systems that used Twinax actually have technical
    documentation which -MIGHT- explain the actual card-level functions?

    Anyone know of a well written Twinax paper?
    On 11/2/2021 12:13, IBMMuseum wrote:
    IBM incorporated "auto" termination in later twin-ax pigtails and had separate terminator caps that could be used. Maybe an IBM tech that installed the 8-bit emulation adapters could provide insight? Jose, can you see?


    The Italian company "Farabi" made an 8-bit adapter that is register-compatible with the IBM adapters - so it is documented somewhere. AST and IDEAssociates were also well-known for their 8-bit and microchannel adapters, many times there was a preference
    for using them instead of the IBM adapters. Twin-ax is a 1Mbps data link that has documented signaling and frame structure.

    More scans and images coming...

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to All on Tue Nov 2 15:47:40 2021
  • From Tomas Slavotinek@21:1/5 to Tomas Slavotinek on Tue Nov 2 22:04:05 2021
    https://priorart.ip.com/IPCOM/000107517

    A lil bit more advanced variant...

    On 2.11.2021 21:22, Tomas Slavotinek wrote:
    Here:

    https://priorart.ip.com/IPCOM/000099938

    This describes it a little bit better and with illustrations.

    What they show there is the kind of cable you would normally use with
    the later terminals/emulator boards.

    On 2.11.2021 21:06, Tomas Slavotinek wrote:
    On 2.11.2021 18:13, IBMMuseum wrote:
    IBM incorporated "auto" termination in later twin-ax pigtails and had
    separate terminator caps that could be used. Maybe an IBM tech that
    installed the 8-bit emulation adapters could provide insight? Jose,
    can you see?

    Many of the later twinax terminal cables are indeed auto-terminating.
    Both ends of the cable are typically terminated by default (by
    shunting the line to the ground/shield via two resistors). These
    resistors get disconnected (mechanically) when you plug something into
    the connector...

    No additional power source is needed for the line termination. I also
    don't think that these later terminals and terminal emulator cards are
    capable of terminating the line. Not sure though.

    If the TO-220 part really is a regulator, it probably supplies power
    for the line interface chip 7378945(ESD). Negative voltage supply?
    What does the package say? Probulation needed...

    The four diodes on the adapter card are probably surge protection.



    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Louis Ohland@21:1/5 to IBMMuseum on Tue Nov 2 16:06:47 2021
    Bombing out on hardware level stuff.

    On 11/2/2021 15:37, IBMMuseum wrote:
    The Italian company "Farabi" made an 8-bit adapter that is register-compatible with the IBM adapters - so it is documented somewhere. AST and IDEAssociates were also well-known for their 8-bit and microchannel adapters, many times there was a
    preference for using them instead of the IBM adapters. Twin-ax is a 1Mbps data link that has documented signaling and frame structure.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Tomas Slavotinek@21:1/5 to Louis Ohland on Tue Nov 2 21:52:57 2021
    On 2.11.2021 21:47, Louis Ohland wrote:
    https://priorart.ip.com/IPCOM/000058785/Using-the-IBM-System-36-38-Emulation-Adapter-Card-As-a-Time-Source


    Mhm, that's pretty much what I meant by triggering in my other post.
    Even IBM used for that purpose it seems... Interesting.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From IBMMuseum@21:1/5 to All on Tue Nov 2 14:27:57 2021
    Here:

    https://priorart.ip.com/IPCOM/000099938

    This describes it a little bit better and with illustrations.

    What they show there is the kind of cable you would normally use with
    the later terminals/emulator boards.

    Correct, that is the hydra for the InfoWindow terminals and later emulation adapters: http://ibmmuseum.com/Adapters/5250/IBM/Twin-Ax_Pigtails/Pigtail3.jpg

    Initially, IBM used termination caps attached to the last 'T' - this is for an emulation board because the twin-ax terminals of the time didn't use the DB-15 connection: http://ibmmuseum.com/Adapters/5250/IBM/Twin-Ax_Pigtails/Pigtail1.jpg

    AST and IDEAssociates just had a termination switch on their adapters: https://ibmmuseum.com/5250Emul/IDEAssociates/Back_Bracket.jpg

    Later IBM adapters had a breakout box with a plunger on the second connection to disconnect termination when a cable was in place - this is the first implementation of an "auto" termination: http://ibmmuseum.com/Adapters/5250/IBM/Twin-Ax_Pigtails/
    Pigtail2.jpg

    and

    http://ibmmuseum.com/Adapters/5250/IBM/MCA/3638FFF8/Breakout.jpg

    I'll show more detail of the DB-15 connection (which can have pins in either direction, depending on the InfoWindow terminal) soon, to add to the documents that Tom found.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)