Page 1 of 1
NASA: Bus Interface Unit for TMScommunications hardware
Posted: Sat Dec 30, 2023 4:34 pm
by tokafondo
A prototype coaxial cable bus communication system used in the Trend Monitoring System to interconnect intelligent graphics terminals to a host minicomputer is described. The terminals and host are connected to the bus through a microprocessor-based RF modem termed a Bus Interface Unit (BIU). The BIU hardware and the Carrier Sense Multiple Access Listen-While-Talk protocol used on the network are described.
What powers this system? Of course, a 6502.
https://ntrs.nasa.gov/citations/19800009062
Schematics and details in the document
Re: NASA: Bus Interface Unit for TMScommunications hardware
Posted: Sun Dec 31, 2023 3:23 pm
by BillO
That's really neat.
It would have been nice if the included the code. Some of those routines might be worth mining.
Apparently NASA use the 6502 quite a bit.
Re: NASA: Bus Interface Unit for TMScommunications hardware
Posted: Sun Dec 31, 2023 4:26 pm
by BigDumbDinosaur
The communications protocol was interesting and sort of embodied some of the principles of Ethernet, along with some principles of TCP/IP. I noted a 6522 was used as an I/O device, as was a 6850 ACIA. The fact that the 6850 was used, and not the 6551, is telling. Even though the 6850 lacks an on-chip bit rate generator, it is still a better ACIA that the 6551...and that’s not saying much.
Re: NASA: Bus Interface Unit for TMScommunications hardware
Posted: Sun Dec 31, 2023 7:45 pm
by tokafondo
I read two 6522 actually, to form a 16 bit parallel bus.
There is a document about the software this unit run, but I don't see listings there
Re: NASA: Bus Interface Unit for TMScommunications hardware
Posted: Sun Dec 31, 2023 11:34 pm
by barnacle
I too was struck by the resemblance to Ethernet, particularly the collision detection/avoidance protocol. Interesting though that they quote a nine-mile packet length, as opposed to the couple of hundred metres of Ethernet.
Neil
Re: NASA: Bus Interface Unit for TMScommunications hardware
Posted: Mon Jan 01, 2024 11:49 am
by BigDumbDinosaur
I too was struck by the resemblance to Ethernet, particularly the collision detection/avoidance protocol. Interesting though that they quote a nine-mile packet length, as opposed to the couple of hundred metres of Ethernet.
It appears from what I understood of what I read that the instantaneous data rate was actually slow. That would help to mitigate the transmission line effects inevitable with a long cable run.
An Ethernet segment running through UTP cable is limited to 100 meters, but can extended by using a repeater between segments to restore signal integrity. It’s possible to get carried away with that and end up with timing violations.
The old “thicknet” (10BASE-5) and “thinnet” (10BASE-2) implementations could go much further...500 meters per segment with thicknet—usually implemented with a low-loss version of RG-8/U coax, or 185 meters with thinnet—usually implemented with RG-58A/U cable. In the latter 1980s, I did some thinnet installations, but by then 10BASE-T was starting to gain acceptance. Field-assembling BNC connectors onto stiff coax was not my idea of fun—punching down UTP is relatively painless by comparison. 
Re: NASA: Bus Interface Unit for TMScommunications hardware
Posted: Mon Jan 01, 2024 12:01 pm
by barnacle
You'd be amazed how many BNCs I have stuck on the end of cables in my time at the BBC
Vampire taps, though...
Neil
Re: NASA: Bus Interface Unit for TMScommunications hardware
Posted: Mon Jan 01, 2024 12:12 pm
by BigDumbDinosaur
Fortunately, I didn’t have to deal with those, as I never did a thicknet installation. That is not to say I didn’t work with RG-8 coax...when I was in the U.S. Navy, I often had to butt heads with thick coax aboard ship. That stuff was everywhere, it seemed, even inside of the gun turrets.