Packet Messages from the San Francisco CommuniTree



A seminar series on Packet Radio, as it now exists (and where it will probably go), will be held by Hank Magnuski, KA6M, who operates the Menlo Park packet repeater. He says: "The high level of interest in packet radio concepts and technology, plus the need to get some design groups going to work on establishing new networking protocols and hardware has prompted me to arrange a series of meetings. Anybody interested in helping...or giving a presentation please contact me (Hank, 415-854-1927 home)."

Times & Place

Tymnet Training Room, Tymnet, 20665 Valley Green Dr., Cupertino. I-280 to DeAnza...go South...Right at Valley Green to 1st 2-story bldg...Enter at East end.....All meetings at 8:p0 PM.

7 July --- VADCG protocol discussion

4 August --- problems and issues

1 September --- interfacing to packet radio

6 October --- AMICON/AMINET

3 November --- Gateway/Repeater design

1 December --- Advanced hardware

Call Hank 415-854-1927 for more info; Wm. Putney 408-446-7190 for better di-
rections to meeting if needed.



Last night your fairwitness went to the second, of a series of seminars that are being held by Hank Magnuski, KA6M, on Packet Radio. (see parent msg) My first </++_ deals with the directions. If you are coming from the NORTH there is no sign saying DeAnza or Cupertino ----> (this way). So if you are coming from the north the turnoff is AFTER hwy85, but BEFORE the Lawrence Exp. If you are coming from the SOUTH the directions work fine.

I was about an hour late for the meeting, so I can't give you a complete report, but I met a few users of the Conference Tree there, who might want to elaborate a bit more. (Please feel free to addto these branches)

Upon arriving, the topic of discussion was on how various stations would communicate to one another through various methods, including, satellites and repeaters, and the various protocols that would allow this communication. The basic flow was drawn out as follows:::

Terminal Node --- Single User Station
Host Node --- Multi-user Station
Repeater --- Reproduces Packets
Station Node --- Global Ctrl. PR/net
Gateway --- Protocol/Media

A discussion then followed on into what I would call the "stacking" order of the data and the problems and solutions of addressing data to other stations. All in all, it was a pleasure attending this seminar, listening to the info that Hank and the other participants exchanged, and highly recommend the sem- minars to follow. (see parent msg)


What is the NRZI encoding that is mentioned everywhere in this packet radio conference? The only NRZI I know about is used in digital magnetic tape recording and is not self-clocking. I use a format called SDLC which is very similar to HDLC and I encode it with a scheme called Phase Modulation which works like this:

Starting bit stream: 0 0 0 1 0 1 1 1 0 1 1
Encoded bit stream: 0 1 0 1 0 0 1 1 0 0 1 0 1 0 1 1 0 0 1 0 1

Is this what is called NRZI in packet radio?

I think it is wonderful that people are working on packet radio and I would like to learn more about it, thanks for all the technical info in the preceding (I hope I put this message in the right place so they are preceding) messages.

Peter Ryan

*** ANSWER??? 15-AUG-81

Peter, if you have access to 'Reference Data For Radio Engineers' (c Howard W. Sams Co.,Inc.) look on page 40-24. There, NRZ, NRZI and PM are discussed. From what info I have on P/R, the protocol is SDLC.

NRZI is a transition whenever the data changes states, SDLC is a 0 (Zero) whenever 5 consecutive 1's (one's) are encountered. The receiver is tasked with the removal of the 0. I can understand why you don't see how to recover a clock from NRZI. Frankly, I don't either. Send a binary 63, and recover the clock from it (without the sync field, that is) My disk system, and I think a lot of others uses NRZI, with sync, so the clock is easily recovered. If you need more info, I'll type it in. (My kingdom for a flying spot scanner interfaced to this machine!) And Curtis, if you can get the March 1981 issue of 'the packet' it has what you just explained in your review of the seminar. Someone ought to get all of the newslettess, articles ect together, and print them up. Very interesting reading, for what I have. Enough computering for today, snooze time.


*** IF/NW 11-SEP-81

The third seminar in our series on packet radio was held last evening, and was attended by 25 people. The theme of the meeting was "Interfacing to the Network".

Announcements -

Howard Nurse, W6LLO, has his 20M packet station available for schedules. He would like to make some contacts to checkout the equipment, as no one else in this area is ready. The station will, in general, be connected to the mailbox system here, but it is currently not continously on. Please call me or Howard, or send us a message.

Hank Magnuski distributed a 555 based on his VADCG board to implement a downline loaded version of the protocol. volunteered to try mounting 3 MK4118N3 1kx8 RAMS on his VADCG board to implement a downline loaded version of the protocol.

The VADCG 1200 baud modem schematics, recently received, were inspected byCqthe group.

Brian Yee and Dave Altekruse described their recent meeting with Larry Kayser, VE3QB. There were several questions concerning the satellite channel, and more information is needed on the details of the Vancouver-Ottawa link.

Speakers -

The main topic of the meeting was a description of how some of our network users have connected their persnal computer systems to packet.

Bob Reiling, W6JHJ

CNTL - 17H = Connect request
07H = Connect acknowledge
53H = Disconnect request
43H = Disconnect acknowledge

The Poll/Final (P/F) bit, 10H, is used to force a response from the receiving station. Used here and in other frame types for this function.

FMCALL - Call of station originating the frame (6 characters)
TOCALL - Call of station receiving the frame (6 characters)

The callsign is left-justified in the field and padded with trailing blanks if the call is shorter than
6 characters.

Supervisory Frames:

ADDR - Address of sender

CNTL - 7 6 5 4 3 2 1 0
----------- Nr|P/F|0 0

also pointed out the extreme diversity of systems in use, and yet there is little trouble in getting these systems to communicate with the current packet setup.

Network Usage Survey -

To wrap up the meeting and to get an answer to |he question of "What are we going to use it for?" a matrix of in|erests vs. potential users was contructed by the group:

Topic Experimentation/Design/Technical -- Votes - 9

Comp of sender
CNTL - 7 6 5 4 3 2 1 0
-----------|Nr|P/F| Ns | 0 | I-frame
Nr = Sequence count of next expected I-frame
P/F = Poll/Final bit
Ns = Sequence count of this I-frame
TEXT - Text field, 128 bytes maximum, ASCII code


T1 - Receive timeout, 2-3 seconds
T1S - Frame timeout, time for frame of maximum length
Tr 0 - Delay time (random) prior to transmission of first frame of a sequence, 950 milliseconds to 1.2