[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

marex news may 21, Packet Race condition

ISS Amateur Radio Status: May 21, 2001

By Miles Mann WF1F,
MAREX-NA (Manned Amateur Radio Experiment, North American Division)

This weeks ISS packet tip.  The ISS Packet Race Condition:

Last week I received an email from an amateur radio station, that was
surprised to see me respond to this ISS Packet Chat room Ping.  The
sending station did NOT see his chat room ping ECHO back from ISS.  He
asked the question, why did others see this packet repeat back from ISS,
when he did not see his packet repeat from ISS.

The answer is a little complex but can be summed in the words "Race

This packet condition I am going to talk about, only applies to a packet
which was correctly decoded and then retransmitted back to earth.

First a little history on ISS packet.  The existing TNC has suffered a
possible battery failure.  This caused the TNC settings to go back to
the default mode.  The ISS crew has been too busy to connect up a
computer to reprogram the correct TNC settings.  Dave Larsen (MIREX) has
provided me with a partial set of parameters which are his best guess on
what the current TNC settings on ISS have defaulted to (MIREX is
managing the TNC ISS project, thanks Dave)

Key ISS TNC settings
Slot		1 [10 milliseconds]
Persist		OFF
TXD		15 [150 milliseconds]
Axdelay	0

Settings used on Mir's 
Slot		10 [100 milliseconds]
Persist		63 [counter range 63-255]
TXD		150 [150 milliseconds]
Axdelay	150 [150 milliseconds]

I am going to invent a few word here to help explain the Race Condition:
RecieverTurnAroundTime =  The total time it takes your radio, to switch
from a Transmit mode to Full receive mode and start passing audio
signals to your TNC.

Comparison: MIR and ISS Settings.  
On the Mir system we had the ability to remotely adjust the TNC settings
(we liked to fiddle to find the most optimum for a space based TNC). 
The Minimum time to issue the command to transmit was 100 milliseconds
(slot = 10).  The minimum time from setting the transmitter on, until
actually sending packet data was minimum of 300 milliseconds.  This
combination gave Earth based transmitters over 400 milliseconds to get
ready to start receiving data from Mir.
(Slot + TXD + AXDELAY = ReciverTurnAroundTime)

The ISS TNC settings have not been optimized due to the temporary
battery problem.  We hope to either have the crew run the setup program
in the future or replace the TNC with a new TNC with the default
settings BURNED into the ROM.
The ISS Minimum time to issue the command to transmit is 10 milliseconds
(slot = 1).  The minimum time from setting the transmitter on, until
actually sending packet data is a minimum of 150 milliseconds.  This
combination gave Earth based transmitters over 160 milliseconds to get
ready to start receiving data from ISS.
(Slot + TXD + AXDELAY = ReciverTurnAroundTime)
Note:  The estimated 160 millisecond ReciverTurnAroundTime for ISS,
appears to be padded by other timers in the TNC.  Most of the actual
ReciverTurnAroundTime from ISS are much longer, which indicates there
are other factors which reduce the number of times the "Short -
ReciverTurnAroundTime" are observed.

There are 3 primary reasons you may not see your Return message coming
back from ISS, (Assuming  your message was re-transmitted from ISS).

1. The Transmit time delay on ISS may be too short for some Earth based
transmitters to recover fast enough to be in Receive mode before ISS
starts sending a packet. When the delay is High, your RX has a better
chance of seeing the packet. When the delay is Low, your RX may not be
ready fast enough to hear the full packet which will cause a CRC error
and prevent the packet from being decoded. Try setting you TNC PASSALL
to Enabled.  This will allow damaged packets to be viewed.

2.  Your receiver may be too slow.  The time it takes your receiver to
switch from Transmit to Receive may be too long to allow your TNC to
hear the first part of the packet message.  I do not have any
specification on VHF equipment, however I have reviewed a few documents
for HF equipment and it seems that most good HF equipment switches from
Transmit to RX in 3-10 milliseconds (which is good).
You then must allow a little more time for the AGC to fully wake up the
receiver.  And you need a few extra milliseconds for your TNC to wake up
(this is where the AXDelay values helps)
(Maybe there is an expert in this field that can shed more light on this
topic, how long does it take a VHF receive to switch from Transmit to
full RX?)

My old DJ-580 2-meter  HT, has a slow recovery on the Receiver.  There
is a Repeater in my town that has a courtesy beep right after you stop
transmitting. When I transmit with the DJ-580, I do not hear the beep
because my RX is too slow.  When other stations Transmit, then I can
hear the beep, my TM-733 always head the beep.

3.  Slow Earth TNC.  Some TNC's wait for the squelch to break, others
are constantly
monitoring the audio stream looking for packet data.  The TNC's that
monitor the
audio stream for packet data are usually a little faster than the TNC's
to wait for the Squelch to open.  My Kantronics TNC's support both
modes, Squelch and Internal.
CD = Software (listen to audio for a packet)
CD = Internal (wait for squelch)
(note:  I had an old kpc-3, which would go deaf progressively deaf after
for several days and a power cycle would cure the problem.  I have not
seen this bug on any newer Kantronics equipment.)
Read your TNC manual and try different modes to faster turn around

Temporary Solution:
Set up a backup radio and TNC configured to Receive only.  Then after a
pass review both logs or compare logs of a local station.  Set your TNC
to PASS all data.  If you are using a Squelch, set it to is lowest
setting.  Make sure you are sending the correct audio level into your

Problem Frequency:
I think this packet race condition is real, but is not a serious issue. 
I only hear what appears to be a race condition problem once per orbit. 
If the ISS crew has time in the future to reload the TNC, the new
settings should eliminate the packet race condition for all but the
slowest Earth Receivers.

ISS Voice Tips:
Over the past few weeks, Susan and Jim have been talking to West-Coast
USA Amateur radio stations.  The orbits of ISS change and repeat over an
8 week cycle.  For the month of June, the orbits of ISS favor Australia
and the Pacific Rim part of the world.

The ISS crew can only use the Amateur Radio station, during their
OFF-Times, when all other important work has been completed.  Voice
contacts are purely Random.  However your best times to find the crews
are during the evening hours on ISS (ISS crews are in UTC time.  Best
times between 1500 - 2200 UTC).

Pacific and European Voice contacts:
Note, under the current radio plans, the ISS crews change frequencies
which favor different parts of the world, the frequencies TX 145.200  RX
145.800 are for Amateur Radio stations in Europe, for the USA TX on
145.490 and RX on 145.800.

Good luck all, suggest you get your tape recorders ready and start
listening to the ISS channels.  Please observer the proper calling

1. Wait for ISS Crew to call CQ or QRZ.
2. Send only your call sign and wait for crew to acknowledge.
3. Listen closely for the call of the station she is talking to.
4. If you do not hear your call sign, do not transmit again until you
hear the ISS say CQ or QRZ

Please be courteous.

73 Miles WF1F MAREX-NA

New MAREX Web pages:
Check out our future ISS Projects and a large list if Mir related links
and tips on how to use the Chat room on ISS and voice frequencies.


Copyright 2001 Miles Mann, All Rights Reserved.  This document may be
distributed via the following means - Email (including listservers),
and World-Wide-Web.  It may not be reproduced for profit including, but
limited to, CD ROMs, books, and/or other commercial outlets without
written consent from the author.

Until we meet again

Via the sarex mailing list at AMSAT.ORG courtesy of AMSAT-NA.
To unsubscribe, send "unsubscribe sarex" to Majordomo@amsat.org