ND-Satellite/9

The ND-SATELLITE/9 is a compact minicomputer, from the NORD-100 series made by Norsk Data. The basic system consists of (from a sales document):

  • CPU and memory management system
  • 512 KBytes of MOS memory with error correction
  • 20 MByte fixed Winchester disk
  • 1.2 MByte floppy drive
  • 9 terminal ports
  • Cabinet, including power supply etc.
  • SINTRAN III/VS operating System
  • Spooling for up to 2 printers
  • NOTIS-WP editor and text-formatter
  • SINTRAN Utility programs
  • Subsystem Package
  • Resource usage Accounting System
  • SINTRAN Backup System

Options

  • One or two communications interfaces (HDLC). Each may be run as an ND-NET interface, or against a mainframe.

Technical specifications

Cabinet height ........ 635 mm
Cabinet depth ......... 660 mm
Cabinet width ......... 480 mm
Gross weight .......... circa 75 kg
Power .............. circa 500 watt, 230 Volt A/C, 50 Hz +- 1%
Operating temperature ... +10 degrees C to about + 30 degrees C, max temperature gradient per hour: 3 degrees C
start peak current ...... 50 amps - 30 millisecs
Main fuse ........... 10 amps, "G" characteristic

Specifics for this machine

Serial number: 900-135

card crate with 8 positions:

  1. 3002 - (no edge label) N100. CPU 3002E
  2. 3012 - (no edge label) 3012M Memory Management
  3. 3023 - 322623 PCB 3023 ND100 Megalink IF. Print: D Eco: G
  4. (empty)
  5. 3038 - (no edge label) 3038E 8" Disk Controller
  6. 3027 - 322627 PCB 3027 ND100 Fl. Disc Ctr. Print: C Eco: J
  7. 3013 - 322613 PCB 3013 ND100 8 Term IF. Print: L Eco: M
  8. 3034 - (no edge label) 3034B Dynamic Mem.

Hard drive: MPI 9410-24, 8 inch 24 MBytes, aka ND-593 22 MBytes 8 inch fixed Winchester drive

This machine runs SINTRAN III - VSE J

Documentation: see my ND documentation list,

Software: SINTRAN III Utility programs, Test Programs for ND-100,

Links

NDWiki, NODAFs guide til bruk av SINTRAN, NDemulator, Kermit: optimize transfer speed,

Operating system testing: SINTRAN III,

History

2012-10-16: serial access - luckily, it was just user error. port 5 works:

MINI>@who

          1  SYSTEM
===>     39  SYSTEM
        670  SYSTEM
        672  SYSTEM

Remember to use ssh -e none flb when connecting remotely, so that ssh's default escape char ('~') doesn't interfere with cu or something else.

2012-10-07: serial access. I connected up a machine (flb) and connected a serial cable (with adapter - I used a breakout box for now) to port 4 (aka terminal 38) and by using

root@kg-flb# cd /tmp
root@kg-flb# cu -e -l /dev/cuau0 -s9600

and I can access the machine via a serial terminal.

Hmm, after a while, both port 4 and port 5 stopped working. Not good. Problem with the terminal card, or the power supply?

2012-10-07: I tried starting SINTRAN on the Satellite again today, but the hard drive wasn't cooperative. No MCL, LOAD sequences could get it to boot. So I loaded DISC-TEMA (Test floppy number 2, 1560&, DISC-TEMA) and ran a parity-check on DISC-21MB-1, unit 0. And got lots of timeout errors. Not good. I must look for the error. Opened up the machine, it doesn't look like the hard drive is spinning at all. Took out the hard drive (use a 9/64 hex bit on the screws), looked at it, and put it back in. It seems like that was it, the machine now boots again. Yay! I also cleaned the filter in front of the front fans while i had the machine open. It wasn't. very dusty, but still.

2010-12-03: I had to open the machine again, to look at the 8 terminal interface card. The lines are divided into two groups: group A and group B. How these relates to the connectors on the back of the machine, I do not know yet. Anyway, selectors: dev. number selector - group A is set to 2, group B is set to 1. baud rate selector is set to 7 (9600 bps) for both groups. RS-232 / current loop: when holding the card with the connectors to the right, and the selectors to the left, component side facing you, group A switches are "up", and group B switches are "down". The only indication on the print is that "V24" is printed below the center on the switches, perhaps "down" is the same as "1" (and means V24 / RS-232)? Info on the NDWiki (3013) seems to confirm this. Testing the terminal ports:

port 2 - RS-232 - doesn't work - terminal 36 ?
port 3 - RS-232 - doesn't work - terminal 37 ?
port 4 - RS-232 - even parity - terminal 38 (decimal)
port 5 - RS-232 - even parity - terminal 39 (decimal)
port 6 - current loop - terminal 48 (decimal)
port 7 - current loop - terminal 49 (decimal)
port 8 - current loop - terminal 50 (decimal)
port 9 - current loop - terminal 51 (decimal)

I used my console terminal to test the current loop interfaces. The communication switches on the terminal are set for 9600 bps, 7-E-1 (7 data bits, even parity, one stop bit). To test the RS-232 interfaces, I used a usb-to-serial adapter connected to my FreeBSD workstation, and the program cu, like this: cu -e -l /dev/cuaU1 -s9600, which is even parity, 9600 bps, and use line /dev/cuaU1 (the name of the serial port of the usb adapter). The serial cable (actually I used a break-out box) has pin 2 and 3 crossed, DTR connected to DCD and DSR, and RTS connected to CTS.

2010-12-02: cards - I opened up the machine again, took the cover off the card cage, and wrote up which cards are in the machine. Most of the cards had a paper label on the card edge, for those missing that one, I read the text on the printed board itself. Here is the list:

  1. (no edge label) N100. CPU 3002E
  2. (no edge label) 3012M Memory Management
  3. 322623 PCB 3023 ND100 Megalink IF. Print: D Eco: G
  4. (empty)
  5. (no edge label) 3038E 8" Disk Controller
  6. 322627 PCB 3027 ND100 Fl. Disc Ctr. Print: C Eco: J
  7. 322613 PCB 3013 ND100 8 Term IF. Print: L Eco: M
  8. (no edge label) 3034B Dynamic Mem.

2010-11-29: Testing the fixed disk drive with DISC-TEMA (from ND-100 test programs, floppy number 2). I inserted the floppy, closed the door and did this:

MACL
1560&
DISC-TEMA

on loading, the program wanted to know the name of the disc, by using the help command, I figured out I had just a few possible matches:

DISC-21MB-1
DISC-21MB-2
DISC-23MB-1
DISC-23MB-2

I tried DISC-23MB-1; and got the error message: "The controller does not correspond with the specified disc-type.". So I tried DISC-21MB-2; the error message was "No such Device. 1)".

On entering DISC-21MB-1, I got this:

Data way to disc system 1 tested.
Memory address register on disc system 1 tested.
Memory buffers initialized.

And then I got the prompt (">"). Cool!

I tried the command PARITY-CHECK (it only reads from the disc, no risk of destroying it). It asks for a unit (0 -1). PARITY and unit 0 goes on for a while, then the console beeps, and I am returned to the prompt again. PARITY and unit 1 starts printing error messages almost at once, and goes on and on, I had to use the Escape button. Hmm, does this indicate that the disk drive is ok? The only thing I have done is to disconnect the connectors (which I had tried before) take it out of the cabinet, and put it back in again. Anyway, this is too good to miss, so I'll risk it:

MACL
&

and the system prints this:

BATCH NUMBER= 1
*** 00.01.14  TERMINAL 670:
PLEASE UPDATE THE CLOCK IF NECESSARY!!

Yay! The machine lives again! More output on the SINTRAN III page.

2010-11-29: I took out the fixed disk drive, to read the label and figure out exactly which model it is. The drive is secured by 4 hex nut bolts (9/64 inch, luckily I had that size bit). The drive is made by Magnetic Peripherals Inc., and has "FINCH IV 9410" stamped in a label. Partno: 77733501, serial no: 168. Handwritten on a label: "IV 168 - 24 M BYTES". Based on this, I think this is a CDC 9410-24 Finch drive; media size: 8 inch, capacity: 24 Mbytes, interface: SMD. It is also a ND label on it, which reads "129 334" and "84-04-05". Nowhere is "ND 593" to be seen on the drive, but that is probably what it is (ND 593 - 22 MBytes 8" fixed Winchester disk). I mounted the drive back in the machine again.

2010-10-28: More testing with the ND-100 test programs. For each test program, I load it from the floppy, and then I issue the RUN command, and note results. When the program says "THE TESTS ARE NOW LOOPING" I abort the testing with the ESCape key. Instruction test (INSTR, or instruction-b:bpun from floppy 1) - result: OK. Memory test (MEMORY, or memory-c:bpun from floppy 3) - result: OK. Paging test (PAGING, or paging-b:bpun from floppy 3) - result: OK. Cache test (CACHE, or cache-2063d:bpun from floppy 3) - this program says (when it starts): "THIS IS A NORD-100. PHYSICAL EXAMINE CANNOT BE DONE DURING THIS PROGRAM!" - result: it fails (15 errors in first test). Perhaps the machine doesn't have cache?

2010-10-27: I'm getting good help from very knowledgeable people. I was advised to try and run a program (config) from the [Test Programs for ND-100] floppy set. It was supposed to be on floppy number one in the set, so I inserted it into the drive and did:

MACL
1560&

which give me a '' prompt. However, trying config just said "no such file". The monitor supports the list-files command (or li-fi) and it was easy to verify that the config file wasn't on that floppy. So I just swapped floppies and listed the files on the other floppies in the set. and config* (or configuration-c:bpun) was on floppy number two. so here we go:

config
run

which listed out the configuration, memory map, the devices it found on some simple tests. The configuration is like this:

computer: ND-100 std.
floating   : 48 bits
paging    : MM-1
cache     : NO
ALD       : 20500
memory  : 512 KB

All tests passed, with the exception of terminal interface 6 (devno 350 - 357), which gave errors:

***ERROR***; DEVNO: 350
                      TRANSMITTER NEVER READY FOR TRANSFER,
                      BEFORE TRANSMISSION OF TEST-DATA.
***ERROR***; NO IDENTCODE FOUND ON LEVEL 10 (DEC.).
                      EXPECTED IDENTCODE: 45
***ERROR***; NO IDENTCODE FOUND ON LEVEL 12 (DEC.).
                      EXPECTED IDENTCODE: 45

there was also an error on the same device in the interrupt priority sequence test:

***ERROR***; DEVNO: 350
             DEVICE NEVER READY FOR TRANSFER

the output finished with

=== NO ERRORS ===

which is good, I suppose.

2010-10-24: I tried to load SINTRAN III (10,0$) from another floppy, but this turned out just like the first one. I wonder, when you load SINTRAN III from the floppy, is it loaded into memory, or onto the disk (winchester)?

2010-10-23: I connected the ND 320 terminal to the console port on the machine with a current loop cable (ND machines uses 20 mA current loop for console / terminals). Connected power to the terminal and the machine, powered up the terminal, and then the machine. The machine started, and the lamps POWER and OPCOM lit up. I also got the OPCOM prompt (#) on the console terminal, but nothing more happened. Pressing the buttons OPERATE + MCL (master clear) on the front panel gave me another "#" on the console, but pressing OPERATE + LOAD had no effect that I could detect. I also tried writing a "&" on the console (this is the same as using the LOAD button), but that didn't help either. I found the SINTRAN III floppies, inserted floppy number 1 in the floppy drive, closed the door and wrote 1560& (without pressing ENTER) on the console. This loaded MACM from the floppy, and gave this output:

MACM-1718-M
SEPT. 4, 1984

INITALIZED FOR:
 DISC-21MB-1, DISC-14MB-1

REMEMBER THE MACM COMMANDS:

)REDEF => REDEFINE DISC TYPE
)HENT   => GET SINTRAN FROM SAVE-AREA
22!        => START SINTRAN
10,0$    => LOAD SINTRAN FROM DISKETTE

TYPE ANY MACM COMMAND:

Unfortunately, trying the command )HENT just gave me an error:

% **** ERROR AT: 000001 **** WRITE ERROR
STATUS= 010131 - BLOCK ADR.= 000002

It seems like the winchester disk (or the controller?) is dead. Let's see if we can load SINTRAN from floppy then, by using the command 10,0$:

Nope, I got a lot of lines with

% **** ERROR AT: 000000 **** ILL. CHARACTER

Followed by this final line:

% **** ERROR AT: 000000 **** I/O ERROR

Not sure if this is a problem with the floppy drive, floppy controller or the floppy itself.

2010-10-22: The ND-246 (aka Tandberg TDV 2200) terminal is dead, doesn't power up. I must fix it, but later. The ND 320 (aka Tandberg TDV 2200/9 S) terminal powers up, and works (I can navigate the menus and so on).

2010-10-22: With the help of a friend, I picked up the computer, two terminals (ND-246 and ND-320), some documentation, and two boxes of floppies (8 inch) at the previous owner's place. My friend helped me, and we got everything into my apartment.