OpenBCM V1.07b12 (Linux)

Packet Radio Mailbox

IW8PGT

[Mendicino(CS)-Italy]

 Login: GUEST





  
UT1HZM > PACKET   10.02.23 20:16l 40 Lines 1583 Bytes #999 (0) @ WW
BID : 17631_UT1HZM
Read: GUEST
Subj: Re: AX.25 Packet Length?
Path: IW8PGT<I3XTY<I0OJJ<LU4ECL<ZL2BAU<EI2GYB<UT1HZM
Sent: 230209/1357Z 17631@UT1HZM.KREM.POL.UKR.EU BPQK6.0.23


Hi Chris.

Its not mandatory to keep same Paclen for both end of link
but same values are good to keep proper L2 timings (acks, link check etc).
Its just a good tradition between operators.

On good VHF channel there are no problems to use full Paclen at 256 bytes,
or 236 bytes if netrom system used.
Also full 7 Maxframe could be used on single point-to-point link.
But if there are several connections active and most operators run old gear,
like obsolete KISS-tncs w/o proper control of modem-channel, better to limit
Maxframe to 2 or 3. 


On HF 300Bd its worth to use now improved FX.25 protocol with FEC code, 
available in two most used software-TNCs: Direwolf and (QT)Soundmodem.

And for FX.25 its mandatory to use fixed Paclen lengths,
for 300 Baud good values exactly 40 or 100 bytes.
If used any values in between of that, FX protocol will fill it up to
reasonable values (40, 100 bytes, etc) with "training" bytes - just transmission
time waste. So on HF FX.25 most usable values are 40 or 100 Bytes.
Also FX.25 are backward complimentary with classic AX.25, can be used in mixed.

For pure classic AX.25 protocol and 300 Bd HF Paclen values are something
between 64 to 80 bytes, exactly not critical value.
Just no sense send too small packets and too long too.
By practices Paclen >80-100 bytes rarely used on HF - QSB will kill long
packets.
But now FX.25 protocol helps a lot! +25-30% improve of performance!
Use of software TNCs even improves a lot more, over old h/w tncs.
Its suggested to all HF packet (aprs too) operators move to FX.25 now.




73, Sergej.



Read previous mail | Read next mail


 11.05.2024 20:56:41lGo back Go up