berklix.com

BSD-PIE icon

berklix.net

BSD icon
Gnu icon
Linux icon

flag_uk_de_icon_v.gif

Disclaimer

Symmetric 375

http://www.berklix.com/~jhs/symmetric/

by Julian Stacey

Preamble

I owned two of the last surviving Symmetric 375s (one went to "Bernd Kopriva" <bernd @ Kopriva _ERASE_ .de> Nov. 2003, & 2nd one too, 2006.05.10) (He also has what was my PC532.)

Occasionally people discover spare machines, & want to rebuild them, so I'm putting virtually all info on the net, including distribution tape & other binaries such as gcc & gdb, + scs reconfigurable kernel kit etc. If you discover one of these extremely rare boxes, don't throw it out - you can now reinstall it instead.

My FTP Archive: ftp://ftp.berklix.com/pub/symmetric/ I stopped merging & reducing & re ordering it around 2006.05.11 there's still duplication of some components. Several index files exist there in capital letter names. Somebody else is welcome to use tools like find & md5 & my own cmpd to further reduce it please.

Symmetric mail list subscribable via majordomo robot

Index Within This Page

Index Of Extra Files & Directories

  • via ftp: docs/ Scanned paper manual
  • 4.2 Tape jhs bits: makefile & notes
  • 4.2 Tape: SCS files on tape - to write a tape or install operating system on salvaged/ recovered/ found hardware
  • 4.3 Tape:
  • Private - offline stuff for JHS only.
  • via ftp: scs: boot.flp How to rebuild from floppy and tape
  • jhs: disc.labels
  • via ftp: scs: qa/ - Quality Assurance scripts
  • jhs: letters/
  • jhs: history
  • Pictures

Summary

Here's a modified copy of an /etc/motd I once had:
        S Y M M E T R I C   3 7 5
 Hardware: Symmetric 375, Processor NSC 32016, 11.0592 MHz, 2Mbyte RAM,
    WD1002-05 + 1-3 ST506 ~75M Discs, 4 x Serial, Ethernet, 1 Parallel,
     3.5 & 5.25" Flop.
 Timezone Kernel: PDT=GMT-7.  Not year 2000 compliant clock,
        so rdist will fail in direction this host to modern host.
make:
     28672 /bin/make
     29696 /usr/bin/local/make
    105472 /usr/gnu/bin/make
    /usr2/local/include/make:   bsd msdos system_5 unix
    /usr2/local/make:       make.c makefile
Coms Program Binaries:
    /usr/local/kermit   works ok (to guug)
        If Symmetric/modem link is at 38400, transfers fail, 9600 is OK.
    /usr/bin/local/kermit-5a has been compiled, but not tested.
    /usr/bin/local/sz   works ok (to pc532)
    /usr/bin/local/xmodem   untested
    gnu/taylor uucp & ftp binaries (slip included) exist.
SCS Binaries:   C,Pascal,2 Basics,F77,APL,Lisp,Prolog,Ingres,Gdb,Emacs,Kermit.
By SCS CC:  /usr/src/local/*.  Little Smalltalk 3.04.  Zmodem (sz,rz),
        Egrep, Gas, Gas-1.38.1, Gdb-3.5, Grep, Install, Tar, Texi2index,
        Texi2roff Objdump, Patch-2.0.12u4.
By GCC:     Bash-1.05, Bash-1.12, Bison-1.11, Cpio-1.4, Fileutils-1.4,
        Find-2.2, GCC-1.40, Ghostscript-2.3, Make-3.60, Zoo-2.1.
        Kermit-5a, gzip-1.2.2, vile
In gnu/bin: cat chmod cmp cp cut d dd dir du egrep gsnd head install ln ls
        mkdir mkfifo mt mv objdump paste rm rmdir tac tail touch v vdir 
Timezone:   PDT - (as reconfigurable kit to permit MET has bad tape driver).
   

Hardware

  • NSC 32016 11.0592 MHz.
  • 2M RAM, soldered, not socketed, no expansion room.
  • Discs: Seperate 5.25" form factor controller card with WD1002-05 for 3 ST506 Discs. 1 standard 5.25" internal disc, + SCS chassis socket for 2nd disc. I jhs extended both chassis for 3rd (2nd external) disc. (my internals are 135 M (1024 bytes/sector, 9 sectors/track, 15 tracks/cylinder, 1024 cylinders) & 80 M, externals are 75M),
  • 4 Serial, (top clock speed 38400 is limited pair wise, ie no 38400 on tty0 & 19200 on tty1), supports SLIP.
  • 1 Parallel Switchable (ie you can select the strobe polarity) Centronics or .....
  • Floppy external socket supports 2 floppy drives, 5.25" or 3.5", 720k (9 * 512 * 80 * 2) or 800k ( 5 * 1024 * 80 * 2 ) (& none of that evil PC slit & twisted ribbon stuff!)
  • Ethernet: 1 with (AUI), 1 without. I used this with an AUI to thin co-ax transceiver, IEEE 802.3 10Base2 Manufacturer Lantronix, Model LTX-2, Switch setting: HBE=On, Rcv=On, this allowed transmission at 50 K Bytes/sec to a faster receiving machine. Without RCV turned on rlogin still worked, but FTP locked up after 2K. When I used CentreCOM 210 Twisted pair transceivers, IEEE 802/3 10Base T (MAU), they only had an SQ (Heartbeat) switch, rlogin & ftp transmission worked with that turned off. In both cases en0 was ifconf'ed by the /etc/rc* stuff with with "-trailers" I believe
  • Tape Drive: SCSI Cassette (not cartridge), 60M drive, SCS driver uses it as 50M). SCSI external connector.
    works in single user mode, drops occasional single bytes multi user, to recover tape data, see slice.c
    • Top Of Page
      CONTENTS OF SYMMETRICS by JHS (interesting bits only, no boring glue logic)
      
      && Means same on both Symmetrics.
      
      && PCB Rev 32 S375
          Both have same single patch on board reverse side:
          DR Pin 3 to near a big hole.
      My 33c95 is socketed, Chucks isn't.
      
      The 2 adjacent DIP headers appear to be for 19200/38400 baud rate switching.
      Open = 38400, Connected=19200,
      presumably 1 DIP for /dev/console & tty1, other for tty2 & tty3.
      
      AS WRITTEN ON COMPONENTS.           MY NOTES
      && NS32016D-10 R S8640                      CPU
         NS32082D-10 N (S8632,S8636=Chuck)                MMU
      && NS32201D-10   SS8620                     TCU
         NS32081D-10 F (S8652,S8632=Chuck)                FPU
            | |  |  ^^^^     Year and month (more likely year & week).
            | |  ^       Flow code (ie which production line & factory).
            | ^^         MHz
            ^        D means ceramic
      && CrystalA 22.11840 MHz  MP221 CTS JAPAN
      && RAM  4 * 16 *  KM41256-15 6382 koreaA    No Sockets, No room for more.
      && Eproms Type 2732,    1 even & 1 odd P/N B8 & B9 Rev 1.11
      SCN2681A, Signetics (2 * K8701=jhs,K8449 & K8621=Chuck) C1N40 E RS 232
      WD 33C93-PL 00-02 8648 PROTO C.WDC'85(socketed) SCSI bus adapter module
      R6522AP R6522-33 8438 (Rockwell?)       Parallel Port
      MMI PAL 20L10CNS 652120
      NEC D4016C-2 8634B9206 <-- is this also on scs board, or just on the wd board?
      --------------------------------------------------------------------------------
      Ethernet (SCS) Board:
      Silk Screen Name: SCS ENET REV2 COPYR 1986 WFJ/lgj
      Crystal 20.008643 MHz
      AM7990DC 60 8541GP C. '84 AMD
      AM 7992 BDC 8618 DP
      --------------------------------------------------------------------------------
      Western Digital Controller Board 1002-05    (40 pin ribbon to cpu)
      Assy 61-031094-00        (only uses +5V )
      WD1014-CL  00-02 8635 C. WDC '82        ECC Buffer Controller
      WD1010A-PL 05-05 8640 C. WDC '82        Winchester Controller
      WD2797A-PL 02-04 8609 C. WDC '82 (socketed) Floppy Controller
      WD 62-002010-010 (socketed)
      NEC D4016-2
      --------------------------------------------------------------------------------
      Power ONE DC Power Supplies     Observed by JHS
          80W written on PCB
          PIN DESCRIPTION
          1   Live
          2   neutral
          3   Ground
          4   -12 0.6A
          5   +12 1.2 A
          6   +12 2.5 A / 5 Pk Orange Wire
          7   Common
          8   +5V 8A
          Max IP 125 W AC Cont, OP 80W DC Cont
      --------------------------------------------------------------------------------
      Streamer Cassette Tape Drive: TEAC MT-2ST/45S-24-U
          Herr Heiber of nbn confirmed it is a 60M drive 901106
          even though SCS driver only use 50M.
          he sent me a manual.
          Serial 714550, FCC ID:DX983C
          Head Assembly:   TEAC 14769360-00 7819 C, Seiko Epson Corp.
          Motor on front spindle (source) teac 14733870-50 dm-4500b No 19x7b
          Motor on back spindle (take up) teac 14733870-00 dm-4500a No 09x7c
      --------------------------------------------------------------------------------
      Winchester: Miniscribe model 6085 p/n 040006092C-077 Serial 117222
          Information for Model 6085 from Western Digital via pc bulletin boards:
              Phone:              +1 303 651 6000
              Type:               ST-506  MFM
              Size:               5.25" Full height form factor
              Capacity:           71 Mbyte (formatted)
              Cylinders:          1024
              Data Heads:         8
              Start Reduced Write Current:    1025
              Start Write Precomp:        512
              Encoding Method:        MFM (Modified Frequency. Modulation)
              Transfer Rate:          5 MEGABITS/SEC
              Average Access Time:        28 MS
              Sectors/Track X 512 bytes:  17
          Information from SCS disktab entry:
              miniscribe85|Miniscribe 6085 85Mb:\
                  :ty=st506:  Type of disk (e.g. st506, floppy)
                  :se#1024:   Sector size in bytes
                  :nt#8:      Number of tracks per cylinder
                  :ns#9:      Number of sectors per track
                  :nc#1024:   Total number of cylinders on the disk
                  JHS NOTE Capacity multiplies to 75497472 = 72M
                      (M=1024^2). There was no :wp#512: entry found
                      (Write pre-compensation cyl. (ST506 only)).
                  --------------------------------------------------------
                          JHS Note 136 cylinders  Root
                  :pa#9792:   Size of partition `a' in sectors
                  :oa#0:      Offset of partition `a' in cylinders
                  --------------------------------------------------------
                          JHS Note 60 cylinders   Paging
                  :pb#4320:   Size of partition `b' in sectors
                  :ob#136:    Offset of partition `b' in cylinders
                  --------------------------------------------------------
                          JHS Note 1024 cylinders,    Pack
                          copying (inc. bad sector replacement)
                  :pc#73728:  Size of partition `c' in sectors
                  :oc#0:      Offset of partition `c' in cylinders
                  --------------------------------------------------------
                  :pd#0:od#0:pe#0:oe#0:pf#0:of#0:pg#0:og#0:
                  --------------------------------------------------------
                          JHS Note 825 cylinders  /usr
                  :ph#59400:  Size of partition `h' in sectors
                  :oh#196: Offset of partition `h' in cylinders
          Information visually observable on drive:
              12v 1.8a 5v 0.9A     Total flaws 53 Surface 4 has 18 flaws.
              CYL HD  OFFSET
              283 0   4288
              26  1   9120
              37  1   176
              38  1   176
              49  2   9584
              50  2   9584
              130 2   2592
              198 2   6944
              200 2   4576
              201 2   4576
              215 2   864
              215 2   5904
              229 2   8941
              282 2   10032
              389 2   4144
              440 2   8672
              441 2   8496
              589 2   5104
              932 2   1296
              999 2   240
              486 3   4672
              506 3   8352
              779 3   9648
              847 3   4288
              80  4   208
              115 4   9184
              130 4   9424
              151 4   1923
              199 4   2864
              227 4   2304
              235 4   3040
              254 4   3104
              352 4   8720
              359 4   2022
              401 4   7552
              451 4   2512
              471 4   8144
              485 4   8704
              486 4   8704
              813 4   9333
              813 4   9568
              814 4   9568
              170 5   8272
              286 6   3312
              294 6   7968
              389 6   2784
              419 6   2054
              419 6   2144
              420 6   2054
              420 6   2128
              421 6   2054
              421 6   2144
              436 6   6544
              end of list
      
      Notes on biker (host from chuck) pals (seen with eproms above pals):
              EPROMS
      
              H3  H4
              H1  H2
      
          H3      H4
          PAL 20R4ACNS    PAL20L8ACNS
          8642        8626
          MMI
      
          H1      H2
          PAL 20R6ACNS    PAL20R6ACNS
          706296      706296
          MMI
      
rs232 From Manual: Serials are configured like DCEs (Modems). Internally DCD 8 is wired to DSR 6. Console DCD is wired active. Available: 2 TXD In, 3 RXD Out, 4 RTS In, 5 CTS Out, 6 DSR Out, 8 DCD Out, 20 DTR In.

Kermit Observed: When started, & connected just to led box, raises 5,6,8

I noticed lowered performance when hanging on a long 30 to 50 metre serial cable which sometimes didn't have a laptop connected at far end: ps then showed gettys with random noise as login name, simple solution was to put a terminating plug on far end of serial cable when not connected to a laptop. This leads to question of if Symmetric used a cheap 7400 series 5V chip, or a more expensive 12 differential chip ? I don't know.

  • Top Of Page
    DEVICE NAMES MENTIONED WHEN BOOTING:
        en is ether
        wd is winch drive
        sb = scsi
        cst is cassette tape
        cpu is scsi address 1
        cst 2 is scsi address 2
    

  • Top Of Page
    CONTENTS OF /usr/include/sys/reboot.h
        /* * @(#)reboot.h   3.1 7/29/84
         * @(#)Copyright (C) 1984 by National Semiconductor Corp.  */
        /* * Arguments to reboot system call.
         * These are passed to boot program in r11,
         * and on to init.  */
        #define RB_AUTOBOOT 0   /* flags for system auto-booting itself */
        #define RB_ASKNAME  0x01    /* ask for file name to reboot from */
        #define RB_SINGLE   0x02    /* reboot to single user only */
        #define RB_NOSYNC   0x04    /* don't sync before reboot */
        #define RB_HALT 0x08    /* don't reboot, just halt */
        #define RB_INITNAME 0x10    /* name given for /etc/init */
        #define RB_DFLTROOT 0x20    /* don't use bootdev for rootdev */
        #define RB_PANIC    0   /* reboot due to panic */
        #define RB_BOOT 1   /* reboot due to boot() */
    

  • Top Of Page
    BOOTING SYNTAX
    Stop unix (& sync discs)
        halt
    Boot from ?
        b0
    Load /boot resulting in prompt Boot:
        b1
    Boot multi user :
        b   or  b 0
    Boot /vmunix single user:
        b2
    Single to Multi user
        ^D
    Load boot from winchester, resulting in Boot: prompt
        b3  or  b 3 or  b 3 0
    Stand alone formatter from Boot:
        wd0a:/stand/format
            wd0h:           would access usr partition
            fd0         probably would access floppy disc
            cst2f           To format winchester
    Boot from cst2b:
            (wd0a:/boot was unusable (copied from miniroot))
            (unfortunately this boot doesn't recognise fd0: addresses,
            wdstrategy complains I/O out of filesystem boundaries)
        b3 1 2
    Boot Single User From Floppy (manual page 104) - this does not read
            wd0a:/boot, but instead reads wd3a:/boot (the floppy boot)
        b2 0 3
    Boot From Tape
        b3 1 2
    Boot From Winchester Swap Partition (after installing miniroot in swap
            partition with initroot program
        b3
        wd0b:vmunix
    
Tape Capacity
  • Top Of Page
    My drive is a TEAC MT-2ST/45S-24-U:w
    Maxell CS 600HD
        I think the Maxell CS600HX are 150M tapes.
        Tapes considered by Inmac & Seemuller to be 60M,
        But Symmetric considers them 50M, whether `testblock' blocking set to
        4k or 20k, Presumably what would stand a better chance of changing
        the capacity is to change the block size in the driver.
        Seemuller    Oct 1990
            49DM 1 off
            40DM 10 off
        SOLUTIONS Ltd +44 959 61204  May 90
            1 box of 10 @ 5% discount
            150 off price retail at #170 with 30% discount
        nbn Elektronik GmbH,         April 90
            Herr Kulartz, Gewerbegebeit, Herrsching. +49 8152 1017
            50M 1 cassette @ 45 DM
            60M 1 cassette @ 55 DM
            60M 10 cassettes @ 38 DM
            60M 50 cassettes @ 33 DM
            all + mwst + 1% + post (post=10dm for 10)
    SCS Inc (Defunct)
        $22.5   tapes found to be 40M
    -   -   -   -   -   -   -   -   -   -
    
Power To Main board
  • Top Of Page
    power converter, standard pc to symmetric
    ( as made ~910918, dismantled 9301 )
    
    standard pc         symmetric
    1    + 5 --+          0 1 black
    2    + 5 --+----------+--   + 5 2 red
    3    + 5 --+    +-----#--   +12 3 yellow
    4    - 5   |    |     |   0 4 black
    5      0   |    |     +--   + 5 5 red
    6      0   |    |  +-----   -12 6 violet
    7      0   |    |  |
    8      0   |    |  |
    9    -12 --#----#--+
    10   +12 --#----+
    11   + 5 --+
    12  (+ 5)
    
    power consumptions estimate @ 9301
        tape 12v 1.6A max steady (presumably when streaming), more at start.
        tape  5v 1  A max steady (presumably when streaming), more at start.
        micropolis 5" disc 12 2A average 4.3 @ start.
        micropolis 5" disc  5 2A average.
        cpu (by observation) 5V 70mA, 12V 50mA, -12V 45mA.
    
Floppy Wiring
  • Top Of Page
        FLOPPY INTERFACE PIN OUTS
    --------------------------------------------------------------------------------
    JULIAN'S COMBINED 5.25" + 3.5" FLOPPY DRIVE BOX
        (3.5" Chinon drive added to original TEAC 5.25" drive)
            TEAC    Part No     FD-55FV-13-U    19307200-13
            Chinon Model No    FB357
        Originally from Symmetric, usable by Unix & to be tried by Toshiba)
        This is the wiring inside the steel box, not the wiring on the external
        34 ribbon to 25 D converter (though this can be expected to normally
        be the same).
           D 25 PIN   RIBBON WIRE NUMBER    SIGNAL NAME
         1      34      Disk Change (Ready)
         2      8       Index
         3      26      Track 0
         4      28      Write Protect Detected
         5      30      Read Data
         6      4       In Use (External LED)
         7      6       Drive Select 3
         8      14      Drive Select 2
         9      12      Drive Select 1
        10      10      Drive Select 0
        11      16      Motor On
        12      22      Write Data
        13      24      Write Gate
        14       2      Mode Select (Disk Change)
        15      32      Side 1 Select
        16      18      Step Direction
        17      20      Step Pulse
        18-25       GROUND      All odd pins (1-33)
        TOGGLE SWITCH   Alternates Select 0 & 1 lines before presentation
                to floppies.
        Signals on flop interface are 0=active.
        Observation of HD line on Chinon 3.5" drive:
            5v= Hi Dens (hole), or No Disc, 0V=720K
        Although the 3.25" is a 1.4M drive, the Symmetric cant use it,
            because WD controller needs to double frequency.
    --------------------------------------------------------------------------------
    34 WAY RIBBON CABLE FLOPPY SIGNAL NAMES
    Pin In |Chinon 3.5" (jhs)   | NEC 3.5" (sja)      | NEC 5.25" (sja)    | Sym.
        Out|FB-357 (135 tpi)     | FD1137H         | FD1157C        | Port
     - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
     2 I *Mode Slct *DskChange   | Density (Hi=0,Lo=1) | <--        | N/C
     4 I In Use LED          | N/C         | Head Load/In Use   | N/C
     6 I *Drive Select 3         | Drive Select 3      | <--        | Sel 0
     8 O Index           | <--         | <--        | <--
    10 I Drive Select 0      | <--         | <--        | Sel 1
    12 I Drive Select 1      | <--         | <--        | Sel 2
    14 I *Drive Select 2         | Drive Select 2      | <--        | Sel 3
    16 I Motor On            | <--         | <--        | <--
    18 I Direction Select (In)   | <--         | <--        | <--
    20 I Step            | <--         | <--        | <--
    22 I Write Data          | <--         | <--        | <--
    24 I Write Gate          | <--         | <--        | <--
    26 O Track 0             | <--         | <--        | <--
    28 O Write Protect Detected  | <--         | <--        | <--
    30 O Read Data           | <--         | <--        | <--
    32 I Side Select         | <--         | <--        | <--
    34 O Disk Change *Ready      | Disc Change     | Disk Change*Ready  | N/C
        *Bla means text following asterisk is an option
        <--  means same as entry to left.     N/C means not connected
        Input/output direction is viewed from drive (not computer)
        Symmetric pin outs are taken from page 187 of SCS manual
        these do not conform to normal select conventions,
        Symmetric manual is wrong - probably applies to how floppy port looks
        if an internal floppy is fitted.
    --------------------------------------------------------------------------------
            Toshiba T1100+ FLOPPY PIN OUTS
    From: rpw3@ amdcad _ERASE_ AMD.COM (Rob Warnock) Systems Architecture Consultant
    Date: 7 Jul 87 05:44:43 GMT
    UUCP:     {amdcad,fortune,sun,attmail}!redwood!rpw3
    ATTmail:  !rpw3
    DDD:      (415)572-2607
    USPS:     627 26th Ave, San Mateo, CA  94403
    Toshiba doesn't document the pin outs on the 25-pin connector
    on the back, the one that's shared between the external floppy (can be either
    disk "A" or "B") and the parallel printer.
    So I engaged in a little electrical
    sleuthing, and figured it out the hard way (with the help of a voltmeter and
    many LEDs and an oscilloscope).  I believe I've succeeded [Minix boots!], and
    present the results herein.  May it be of some use to someone else.
    DISCLAIMER: There are no guarantees that this information is correct,
    reliable, or safe. While I am indeed using the connection described in
    my own system, I cannot guarantee that it will function in yours, or
    that it will not damage your system. There are at least two ambiguous
    pairs of signals in the pin outs [which I think don't matter, see below],
    and other pins which I show as "unused", even though they appear to have
    termination voltages on them. [Anyone having better information as to
    the "official" pin outs, please correct me.]
    I used a half-height Mitsubishi drive , using a small power pack,
    I'm a bit worried that 300ma of +12v might not be enough),
    T1100+ DB-25 pin    Signal name Floppy edge-connector pin
    ----------------    ----------- -------------------------
        01      Ready           34  [Note#1]
        02      Index           08  [Note#1]
        03      Track0          26
        04      Write-Protect       28
        05      Read-Data       30
        06-09       (n/u)               [Note#3]
        10      Select          10  [Note#2]
        11      Motor-On        16  [Note#2]
        12      Write-Data      22
        13      Write-Enable        24
        14      (n/u)               [Note#3]
        15      Head-Select     32
        16      Step-Direction      18
        17      Step-Pulse      20
        18-25       GROUND          All odd pins (1-33)
    Notes:
    #1. It appears that Ready and Index are AND'ed together inside the T1100+
        controller, since the disk worked equally well whichever signal was
        connected to pins 01 or 02. The assignment given above was chosen
        arbitrarily.
    #2. It appears that Select and Motor-On are are driven by the same signal
        inside the T1100+ controller, since the disk worked equally well whichever
        signal was connected to pins 10 or 16, and I did not see any difference
        on a 'scope. The assignment given above was chosen arbitrarily.
    #3. Pins 6-9 and 14 seem "unused", though they do have +5 "pull ups" on them.
        (This may be due to those signals being used when the port is selected
        as the parallel printer.)
    #4. All signals are "active-low", and are pulled up to +5 by the T1100+ when
        inactive.
    As I said above, it seems to work fine: MS-DOS "format" works, MS-DOS floppies
    can be read and written, and Minix boots and reads /usr.
    --------------------------------------------------------------------------------
    FLOPPY DISC RIBBON INTERFACE CABLE WITH TWIST As purchased by JHS 9-90.
    With 3 sets of [ 3.5" floppy twist connectors + 3 5.25" floppy connector ],
    and a twist between one set and the other 2 sets.
    
    First 9 wires are straight through connection,
    next 7 are twisted in a group, rest are straight.
             1   straight   1
             9   straight   9
         Select 0   10    twist    16   Motor On
         ground 11    twist    15   ground
         Select 1   12    twist    14   (Select 2)
         ground 13    twist    13   ground
         (Select 2) 14    twist    12   Select 1
         ground 15    twist    11   ground
         Motor On   16    twist    10   Select 0
            17   straight  17
            34   straight  34
    --------------------------------------------------------------------------------
    TOSHIBA LAPTOP MANUAL : as read by Joe McBeth (deceased)
        (female on floppy & computer)
    T1100+ DB-25 PIN    TOSH SIGNAL NAME
        01      Extern Drive Ready
        02      Index
        03      Track0
        04      Write-Protect
        05      Read-Data
        06      Disc Change
        07-09       Not Used
        10      Drive Select 1
        11      Motor-On
        12      Write-Data
        13      Write-Gate
        14      Low Density
        15      Side 1 Select
        16      Step-Direction
        17      Step-Pulse
        18-25       GROUND
    --------------------------------------------------------------------------------
    NEC LAPTOP `MULTISPEED' MANUAL DESCRIBING EXTERNAL FDC SOCKET
        As read by Julian           (15 PIN)
            SIGNAL      SIGNALBEZEICHNUNG   I/O
         1  Index       Inhaltsverzeichnis  0
         2  Motor On C  Motoreinschaltung C 1
         3  Select D    Laufwerkswahl D     1
         4  Select C    Laufwerkswahl C     1
         5  Motor On D  Motoreinschaltung D 1
         6  Step Direction  Laufrichtung        1
         7  Step        Schrittimpuls       1
         8  Write Data  Dateneinschreibung  1
         9  Write Enable    Schreibfreigabe     1
        10  Track 0     Spur 0          0
        11  Write protect   Schreibschutz       0
        12  Data        Dateileistung       0
        13  Head 1      Wahl Schreibkopf    1
        14&15   Ground      Erde
    --------------------------------------------------------------------------------
        NOTES ON FLOPPY DRIVERS - By Julian Stacey
    To accompany valid, dd, symmetric, unix msdos debug scenarios.
    
    The Symmetric Unix device driver source was read, this
    reads both head 0 & 1 of track 0, before progressing to track 1.
    
    MSDOS BOOT SECTOR ANALYSIS
        From head 0 sector 1 track 0 (note high byte is 2nd byte, ie 0C for
        first high byte) Reference: Advanced Msdos by Ray Duncan, page 166.
        OFFSET  KEY:
        -   1   size as reported by chkdsk ("bytes total disc space")
        -   2   media size 5.25" / 3.5" as 525 or 350
        0B & 0C 3   bytes per sector
        0D  4   sectors per cluster
        0E & 0F 5   reserved sectors
        10  6   number of FATs (File Allocation Tables)
        11 & 12 7   root entries
        13 & 14 8   total sectors
        15  9   media descriptor
        16 & 17 10  sectors per FAT
        18 & 19 11  sectors per track
        1A & 1B 12  heads
        1C & 1D 13  hidden sectors
        -   14  tracks per head (derived, not explicit in boot block)
        -   15  comment
    1   2   3   4 5 6 7   8    9   10 11 12 13 14 15
    1457664 350 512 1 1 2 224 2880 240  9 18  2  0 80 IBM PS/30 Msdos 3.3
    730112  350 512 2 1 2 112 1440 249  3  9  2  0 80 Toshiba  Msdos 3.2
    362496  525 512 2 1 2 112  720 253  2  9  2  0 40 IBM  3.1
    368640  525 512 2 1 2 112  720 253  2  9  2  0 40 IBM 3.1 IP's
    160256  525 512 2      640         8  1    40 Sanyo 1.? from boot sector
        Boot sector from the 160K disc does not make sense
    
    Comcen (ukc) wrote a disc readable by G0
    IP's PC-AT clone wrote a 360K floppy readable with /dev/rfdE0.
    

Various

  • Unix UCB BSD 4.2.
  • Kernel size: 204,800 bytes (inc. ethernet).
  • Standard Features: Lisp (Franz Lisp, Opus 38.79), Prolog (Advanced A. I. Systems' Prolog. Version 1.1, 6 August 1986) , Emacs
  • Extra Features from jhs: Kermit, gcc, gnu/taylor uucp, kermit-5a-386bsd, various other own C programs, + moved all bins from /etc/to /sbin, leaving just a few essential symbolic links.
  • Top Of Page

    SCS Binaries

    C,Pascal,2 Basics,F77,APL,Lisp,Prolog,Ingres,Gdb,Emacs,Kermit. Rdist (I think that's SCS), but no NFS. Top Of Page

    By SCS CC

    /usr/src/local/*. Little Smalltalk 3.04. Zmodem (sz,rz), Egrep, Gas, Gas-1.38.1, Gdb-3.5, Grep, Install, Tar, Texi2index, Texi2roff Objdump, Patch-2.0.12u4.
  • Top Of Page

    By GCC

    Bash-1.05, Bash-1.12, Bison-1.11, Cpio-1.4, Fileutils-1.4, Find-2.2, GCC-1.40, Ghostscript-2.3, Make-3.60, Zoo-2.1. Kermit-5a, gzip-1.2.2, vile
  • Top Of Page

    In gnu/bin

    cat chmod cmp cp cut d dd dir du egrep gsnd head install ln ls mkdir mkfifo mt mv objdump paste rm rmdir tac tail touch v vdir
  • Top Of Page

    Time zone

    PDT - (as reconfigurable kit to permit MET has bad tape driver). rtc was set to GMT Sunday 930926, when central European winter time started

    Time zone Kernel: PDT=GMT-7

    /sbin contains all the system binaries jhs removed from /etc.

  • Top Of Page

    make

    • 28672 /bin/make
    • 29696 /usr/bin/local/make
    • 105472 /usr/gnu/bin/make
    • /usr/local/include/make: bsd msdos system_5 unix
    • /usr/local/make: make.c makefile
  • Top Of Page

    Coms Program Binaries

    • /usr/local/kermit works ok (to GUUG) If Symmetric/modem link is at 38400, transfers fail, 9600 is OK.
    • /usr/bin/local/kermit-5a has been compiled, but not tested.
    • /usr/bin/local/sz works ok (to pc532)
    • /usr/bin/local/xmodem untested
    • gnu/taylor uucp & ftp binaries (slip included) exist.
  • Top Of Page

    Coms Program Sources

    : kermit-4c.scs.tar.Z kermit-5a-386bsd.tar.Z kermit-ms-doc-2.26.Z kermit-ms2.29-ukc3.2.tar.Z kermit.jhs.Z kermit.siemens.tar.Z umodem1.1.tar.Z xmodem-3.6.tar.Z zmodem.tar.Z Top Of Page

    usr/vsl/bin/

    Directory of my programs such as timesheet, valid etc.

    I have:

    • Additional FSF GCC (1.40 (& maybe 1.42)) compiler for it, that's much better than the native compiler. (I can now easily put this on an via ftp site if needed).
    • SCS Inc. reconfigurable kernel kit for it.
  • Software Signals
    • Top Of Page
          Provided by 386BSD & BSD4.2, Not used by uucp-1.04 code,
          Possibly safely re-assignable:
              SIGCONT     continue after stop
              SIGIO       I/O is possible on a descriptor (see fcntl(2))
              SIGURG      urgent condition present on socket
          Provided by 386BSD & BSD4.2, Not used by uucp-1.04 code:
              SIGEMT      emulate instruction executed
              SIGTTIN     background read attempted from control terminal
              SIGTTOU     background write attempted to control terminal
              SIGVTALRM   virtual time alarm (see setitimer(2))
              SIGPROF     profiling timer alarm (see setitimer(2))
              SIGXCPU     cpu time limit exceeded (see setrlimit(2))
              SIGXFSZ     file size limit exceeded (see setrlimit(2))
          Provided by 386BSD & BSD4.2, Not used by uucp-1.04 code,
          but probably too dangerous to re-assign:
              SIGBUS      bus error
              SIGFPE      floating-point exception
              SIGSTOP     stop (cannot be caught or ignored)
              SIGTRAP     trace trap
              SIGSYS      system call given invalid argument
          Provided by 386BSD & BSD4.2, Used by uucp-1.04 code.
              SIGCHLD     child status has changed
              SIGHUP      terminal line hangup
              SIGILL      illegal instruction
              SIGINT      interrupt program
              SIGABRT     abort(2) call (formerly SIGIOT)
              SIGKILL     kill program
              SIGPIPE     write on a pipe with no reader
              SIGQUIT     quit program
              SIGSEGV     segmentation violation
              SIGTERM     software termination signal
              SIGTSTP     stop signal generated from keyboard
              SIGUSR1     User defined signal 1
          Provided by 386BSD, but not by BSD4.2, referenced by uucp-1.04:
              SIGUSR2     User defined signal 2
          Provided by 386BSD, but not by BSD4.2, not referenced by uucp-1.04:
              SIGWINCH    Window size change
              SIGINFO     status request from keyboard
          Not Provided by 386BSD or BSD4.2, but referenced by uucp-1.04
          (compiles on BSD4.2 so presumably #ifdef'ed out):
              SIGACTION SIGALRM SIGBLOCK SIGCLD SIGHOLD
              SIGIN SIGPROCMASK SIGSET SIGSETJMP SIGVEC
      
    Kernel Comment
    • Top Of Page
      Comment about SCS kernel taken from TOP source by JHS:
      
      (preprocessor constant: scs).
      
      February, 1987, by Paul Vixie (dual!ptsfa!vixie!paul @ ucbvax.Berkeley.EDU).
      
      The SCS kernel is basically 4.2, but with a number of differences -- some I
      like, others I don't.  The list:
      
          all page counts are in kilobytes (all the ones I had to use, anyway);
          PGSHIFT is defined as 9 in {machine/param.h}, this may be an error;
          the physical page size is 512, but the kernel pages are 1024. sigh...
      
          in the proc structure, there is no p_rssize; p_tsize and p_dsize are
          also missing, though combined in p_tdsize (whose value is in BYTES,
          NOT PAGES).  p_ssize has either 0 or 16MB as its value, and is there-
          fore useless.  p_addr is missing...
      
          ...however, there is a field p_spti which indexes into the page table
          table.  each spt has a value telling how many pages are used (total
          reference code), and how many are actually in memory.  these were used
          for the "SIZE" and "RES" fields on the display...
      
          ...the lack of p_addr made it hard to find the U area for a proc, but
          since SCS put the process name and CPU time into the proc structure
          for use when ^T (info) is pressed, the U area wasn't needed anyway...
      
    Tape Drive - Specification
    First item of multiple things for sale from someone I never met, Note he reckons that model drive works OK, so it must be the SCS driver that loses the odd byte. - JHS @
    • Top Of Page
      Path: pcsbst!unido!mcsun!uunet!cis.ohio-state.edu!ucbvax!cas.org!dlb25
      From: dlb25@ cas.org (david besselievre)
      Newsgroups: misc.forsale.computers
      Subject: Tape Drives for Sale
      Message-ID: <9112111807.AA08220@ cas.org>
      Date: 11 Dec 91 18:07:20 GMT
      Sender: daemon @ ucbvax.BERKELEY.EDU
      Lines: 94
      
      
      I have the following Disk and Tape drives for Sale.  These used items are
      sold with the warrantee that I have tested and will refund or replace any
      drive that does not function correctly.  The WORM drive will be tested if
      requested, otherwise, I will not open the factory seal on the drive.
      
      Description                      Cost
      ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~        ~~~~~~~~
      
      External Teac QIC-2 60MB Cassette Tape Backup system.      225.00
      Includes case, power supply, drive, Interface card
      (Everex QIC-02 controller), cables, and software with
      manuals.  The software will backup 5MB per Minute both file
      by file and image backup.  The Software reliably restores
      single or multiple files from an image backup tapes.
      
      Dave Besselievre
      5441 Richlanne Dr.
      Hilliard, Ohio   43026
      
      (614) 771-9629  After 6PM
      
      osu-cis!chemabs!dlb25
      dlb25@ cas.org
      
       _________________________________________________
      |  ____   ___    ____|                |
      | / \  __ \ /    | David L. Besselievre   |
      ||   /     |\ ___| Chemical Abstracts Service |
      ||  |      |     | Columbus, OH 43210     |
      ||  |      |     |                |
      | \ ____ \____/|\____| osu-cis!chemabs!dlb25      |
      |            | dlb25@ cas.bitnet       |
      |            | dlb25@ cas.org          |
       -------------------------------------------------
      
Owners I'm interested to hear of other owners, Please contact me Other Owners include:
  • Matt Neely ne99mb70@_ERASE_ACS.WOOSTER.EDU (no such user @ 2003.12) There's a few other owners somewhere in my archives.
  • Salam of www.vintage.org has a 3rd in the collection of his museum associate, as of 2001.04.
  • Brenda Walker walker @ volcano _ERASE_ net wrote 24 Jun 2001: As sister to Bill Jollitz the designer of the Symmetric 375, I too have the machine and know of one other.
  • Christopher Cox ccox @ endlessnow _ERASE_ com wrote me June 2014, with URL to picture with Serial 375-8507-00024 with floppy drive
Other People
  • Hans-Peter Nilsson hp @ bitrange _ERASE_ com Maybe interested in gcc or something, if I can put it on Internet.
  • Bill Studenmund wrstuden @ sun510.nas.nasa _ERASE_ gov for ld.old ?
  • Bill McKie NASA Ames mckie @ sky.arc _ERASE_ nasa.gov
People
  • Top Of Page
    SCS User group contacts from Jordan 
    Bill Blue (enet bdale) was head of SCS user group
        bdale @ 2003.12 has just the PCBs, & will sub. the mail list.
        Bill Blue was also a big symmetric (dealer/user?)
    Paul Vixie: Much experience of SCS, but now sold his machine to chuck.
    Chuck:
        Jordan said:
            when winchester aged, the 12V 3.5 A power supply was
            not enough, needed 4A, so he upgraded to a 5A supply.
        Jordan says he has Paul Leaskys machine, & bill blue's,
        he said he only has 1.
        he has 3 drives, both size floppies, and tape., no ether
    
Odds
  • The manufacturer, SCS (Symmetric Computer Systems) Inc. is long defunct.
  • It was named the 375, in honour of the VAX 750, & SCS Inc. adverts promoted it as "Half A Vax"
  • New price was: $8000 / $9000 with Ethernet + built in tape, + VT220 clone + cheap nasty matrix printer, or $5000 with built in floppy, but no ether, no vt220, no matrix.
  • Designer was Bill Jollitz, whose initials WFJ appear in block capitals on the circuit board. (william @ _ERASE_okeeffe.Berkeley.EDU ) That's the same Bill Jollitz who wrote the Byte articles, translated into German by CT mag too, about porting BSD Unix to the Intel PC, which port he called 386BSD, which was the launch point for such great things as FreeBSD, NetBSD, & OpenBSD

  • They were exhibited at Vintage Computer Fest Europa, 1st May 2000, & 28&29 April 2001, & 27 & 28th April 2002 & May 3 & 4 2003. The ethernet machine was exhibited again in 2006.
  • Top Of Page

    Floppies - Chucks Labels on Chuck's floppies received from Paul 930108 (field order varies in accordance with varied floppy label order)

    Contents:
    Format:
    Vols:   1
    Date:   1/28/86
    SN: 8504-00049
    System: 375
    Config: 4.2BSD
        FIXIT DISK REV 2.1 wfj/scs
    ------------------
    Contents:   Root FS
    Format: Dump
    Vols:   6
    Date:   4/1/86
    SN:
    System: s/375
    Config: 4.2
    ------------------
    Contents:   Usr FS
    Format: Dump
    Vols:   23
    Date:   4/1/86
    SN:
    System: s/375
    Config: 4.2
    ------------------
    Contents:   4.2BSD FIXIT Floppy Bootable Floppy UNIX System
    Format:     800K Floppy Filesystem
    Date:   3/87
    Vols:   1
    System: 375 4.2BSD
    SN: na
    ------------------
    Contents:   375 Electronic Magazine Dump (Winter/Spring 87) Edition
    Format: Mountable Floppy Filesy
    Date:   6/87
    Vols:   1
    System: 375 4.2BSD
    SN:
    ------------------
    Contents:   375 Electronic Magazine Dump (Winter/Spring 87) Edition
    Format: 800K Floppy Dump
    Date:   6/87
    Vols:   11
    System: 375 4.2BSD
    SN: 8 on Vol-1, 9 On all others
    ------------------
    Contents:   VMUNIX with SCSI    (pencilled scs label)
    Format: tar
    Date:   8/87
    Vols:   1
    System: 375
    SN:
    ------------------
    Contents:   TAR VMUNIX scsird scs...    (pencilled scs label)
    Format:
    Date:
    Vols:
    System:
    SN:
    ------------------
    Hand Written non SCS label
    Symmetric S375
    tar /VMUNIX + tar -xf /dev/rfd0
    ------------------
    Hand Written non SCS label
    SYMMETRIC
    CST0,CST2
    RTC0
    1-19-90 CH (Presumably Chuck Hein )
    Fix it floppy
    ------------------
    Hand Written non SCS label
    SYMMETRIC
    CST2
    CST0    1/19-90 CH
    RTC0
    Fix it floppy
    ------------------
    Hand Written non SCS label
    UU43N
    ------------------
    Hand Written non SCS label
    
Finance
  • Top Of Page
    26-3-87 i phoned symmetric, spoke to phuong Nguyen
    $5k +
    475 cartridge
    500 scsi card
    ~300 delivery
    ~850 if want floppy as well as cartridge
    1200 if 85 rather than 50m disc
    shipping 2 weeks after receipt of cash
    they will apply for an export licence for me.
    terry reckons flights to states cost #500 return (12-87)
    
    comparison of buying a basic symmetric, then upgrading, or buying bigger m/c
    small           big
    4995            8095
    1000    50m>85m     -
    -               + floppy
    600 scsi card
    x   ether
    675 vdu
    ~350    printer
    850 cassette
    
    8470    TOTAL       8375
    implies get ether free if all prices are fair, & want lpr + vdu.
    --------------------------------------------------------------------------------
    87-12-17 Lynne quoted me complete system price (based on the July 1st price list
    8095    Complete 375 Software Developers Package
     100    Configurable Kernel
      20    Ingres
      20    EMACS
      15    Spice
      15    TOP
      20    GDB
     330    Transport (2nd day air)
     280    Floppy
    8895    SUB TOTAL
     112.5  5 Tapes @ 22.5
    9007.5  TOTAL
    
    this system does include the scsi, st506, flop & interface
    comes with the schematics for the correct revision
    4.3 still beta test, they are doing scsi research to get a more general driver.
    
    Bank Of America
    Account Number 14184-9-01338
    Palo Alto Corporate Banking Group (CBG)
    Palo Alto California
    1484-94301
    --------------------------------------------------------------------------------
    American Siemens programmer with a Symmetric
    Dave Mysan
    Berlin 011 030 802 5947 (home)
    Potsdamerstra~e 8,
    1000 Berlin 37.
    900629 I tried the number above, no good,
        but answer-phone said it was a berlin number, i left a message asking
        him (whoever) to ring back.
    he has a 90M Seagate,
    had the system 2 years,
    a year ago it died shortly after returning to berlin from a round trip to the states,
    he sent the whole box back to the states,
    they fixed the dead drive
    he knows a chap called tom kern who `sells programmers' in m}nchen
    he said there is no manual documenting the particular pascal dialect used
    franz lisp manual is on machine
    fortran 77 has no manual
    he has a connection to bitnet (post don't know)
    he mentioned the 875,
    --------------------------------------------------------------------------------
    i phoned Lynne to ask what the 875 was,
    the 875 will be a pedestal mounting box, sold to VARs, clipper based, available
    2 1 88, later in the year a clipper upgrade for the 375 will be released
    $5000 8MIPs
    --------------------------------------------------------------------------------
    87-12-19 exchange rate approx 1.63 dm = $1
    Lynne Mesner 9.40 am wed 12 dec. 12.30 ie 30th dec.
    + 1 415 651 6090
    no money in her bank on 29th no mention of receiving my letter
    2nd call, transfer had gone through will ship within 1 to 1.5 weeks
    phoned Lynne, gave her a hard time, she put me on to Bill Jollitz
    88-01-15 I phoned they are waiting for a better tape drive Mon./tues, then will ship,
    I get 4.3 free later, (other customers pay $1K) my machine on test, ok except tape drive.
    cant have 4.3 yet as there is a problem with signals & fp
    --------------------------------------------------------------------------------
    88-01-28 Symmetric delivered.
    --------------------------------------------------------------------------------
    88-01-After receipt teac +49 8152 39101 Mr Okada phoned me
    --------------------------------------------------------------------------------
    900629 IDQ phoned USA, there is no firm called Symmetric in fremont,
    900629 00 1 415 651 6090 no answer @ 18:50(mu)=09:48(usa)
    --------------------------------------------------------------------------------
    90-10-11 Jordan confirmed Symmetric had gone belly up!, but there is a Sym. users group
    --------------------------------------------------------------------------------
    

Apache: Web Server FreeBSD: Operating System