I happen to have a Siemens terminal, model S26361-K164-V1, and it seems its interface with the computer is not a more common RS-232, but something called SS97.

I couldn't find much about it. It seems to have been popular in embedded and process control, but, unfortunately, there are too many false positives for Google to yield anything interesting about its specs. It seems to be somewhat related to RS-422, but I couldn't find anything in that direction.

Is it an official de-jure standard that's documented somewhere?

An image of a seemingly identical unit's front:

photo by Alessandra Schellnegger

Image from a SINIX manual:

Diagram of the back. Top row of connectors is absent and has only covers


1 Answer 1


I happen to have a Siemens terminal,

Let me guess, a 97801? :))

and it seems its interface with the computer is not a more common RS-232, but something called SS97.

SS97 stands for Schnittstelle (interface) System 97 ... note the first two digits of the terminal number? It was introduced around 1978 to replace a prior proprietary interface (SS81) used with mainframe terminals to connect printers, floppies or other local extension like bar code readers, magnet card readers, special keyboards or OCR reader.

SS97 is implemented as a V.11 of the EIA-422 kind. So a quite standard asynchronous serial using differential transmission.

Differential transmission was chosen to allow cable distances common in an office setup. Siemens guarantees up to 60 meter in standard environment.

Is it an official de-jure standard that's documented somewhere?

It's plain serial, no hidden tripwires. Well, there are a few details on the host side, differing between console and user terminal as well as terminal and other peripherals, but none may really matter if you just want to connect a terminal to some serial.

As mentioned, speed is usually 38.4 kBit/s. Data is 7N1 or 8N1, depending on model - just use 8N1. Connector is a DE9. Pinout:

enter image description here

From a Comment

The only thing that looks like a model number is the S26231-K164-V1 on the back of the main unit

The S number is a Sachnummer (Item Number).

  • S26361 marks it as a PC product, while
  • K164 is the exact model and
  • V1 a version (seldom used)

(the 97801 seems to have the brains inside the monitor case)

Huh? All (standard) Sinix Terminals are 97801. Differences are noted in a trailing code, like 97801-301 or -305, etc.

The K164 is according to a list I have a 97801-370. A somewhat unusual one as it got no CRT, but is to be used with an external screen in a BTX setup. Usually as page edit workstation.

I ordered a converter (easier than building one)

You still may have to solder the connector and fix some of the other signals (see above). Since the differential drivers converters are dedicated chips it might as well be possible to pick up TTL signals right from the UART.

and I hope I didn't fry the interface by connecting an RS-232 to it.

Usually not.

  • 2
    38.4k is pretty good for that time. Terminals I used at the time (well, actually a few years later) were generally maximum 19.2k. I checked and the VT100 - also introduced in 1978 - was max. 19.2k. Commented Apr 5 at 3:19
  • 1
    @manassehkatz-Moving2Codidact well, yes and no.When introduced for the new 970 series mainframe terminals it was just local I/O. Host connection was done using BAM at 288 kBit/s (Hdx). Only with the later Unix (Sinix) based micros it also was used as host interface. So there was some pressure upwards :)) The first 97801 series was build using mostly 9780 (the mainframe terminal) components. Including its awesome keyboard. Still miss that feeling. I'd be even willing to canibalize some to finally build an acceptable PC keyboard. They are just way to hard to find.
    – Raffzahn
    Commented Apr 5 at 11:23
  • 1
    @manassehkatz-Moving2Codidact BTW, those micro terminals (97801) were introduced in 1983 or 84 (IIRC) and accept a superset of VT100 instruction including all that code switching and so on. So adapting one to use with other systems might not be hard at all. Internally the first series was 8085 based, the later (like the above) used some intel based micro controller (8035 possibly)
    – Raffzahn
    Commented Apr 5 at 12:15
  • This unit has the screen integrated on top of the brains, with a signal cable connecting it to the main board on the outside. Power is probably connected on the inside. I've been in contact with the only other person who seems to own something similar (cpu-ns32k.net/Siemens.html). It's one weird machine. I'll add more photos as I work on it.
    – rbanffy
    Commented Apr 6 at 12:21
  • 1
    @rbanffy Now I really need to see pictures, including connectors, as with the regular screen on top it shouldn't be a K164. I have some 97801 in storage (and two older ones in my garage). They were plenty, but since most were used in businesses, especially large companies/government, they tend to be scrapped when no longer used, not stored in some basement :)) Udo Möller is well known for his NS32xxx work (also lives just a few km from here). No, power is not separate, but part of the screen connector. That way it ges switched on by the CPU using lower voltages. Done for better handling.
    – Raffzahn
    Commented Apr 6 at 13:06

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .