[600MRG] 630m WSPR SNR v QSO modes: Will 630m WAS be possible(?)

[Ralph Wallio, W0RPK]

During our cold and snowy January ... well, for some of us, there has 
been a lot of 630m WSPR-2 activity with numerous transmit and receive 
stations across the US ... well, minus Hawaii.  Collectively we are 
building reasonable expectations of WSPR SNR performance over several 
paths.  Given a new 630m Ham band, these expectations can be applied to 
SNR requirements for several modes that might be used for ARRL 630m WAS 
awards.

Will 630m WAS be possible even for our CONUS 48-states, let alone for 
50-states including Hawaii and Alaska?  Perhaps we can partially answer 
this question using collected WSPR SNR performance over several paths 
and required SNR for a few digital modes.  Perhaps we can arrive at 
reasonable expectations before we crowd-fund volunteer 630m DXpeditions 
to distant states ...

We can make an educated guess regarding probable regulations for a new 
630m Ham band.  The ARRL Petition for Rule Making asking for a new 630m 
Ham band includes two important issues 
(http://apps.fcc.gov/ecfs/document/view?id=7022073018):

1] Our -CONUS- 630m Ham band power limit may be 5Weirp (not ERP). 
Perhaps only stations in Alaska could be limited to 1Weirp.

2] Our 630m Ham band bandwidth limitation may be 1000Hz with the same 
97.307(f) standard as with HF bands.

We currently have a few 630m Part-5 experimental stations transmitting 
with more than 5Weirp so we have to be careful to derate any WSPR SNR 
performance collected for these stations. Part-5 WD2XSH stations have 
been limited to modes with bandwidth much narrower than 1000Hz.  The 
WE2XGR group and other individual stations are licensed for and have 
experimented with modes with much wider bandwidth.  (See 
http://www.w1tag.com/600M_Modes.htm for a lengthy and informative 
discussion examining digital mode testing on 630m by John Andrews, W1TAG 
WE2XGR/3.)

Minimal QSOs just for WAS and other awards can happen down to -26dBsnr 
with WSJT-X using JT9 for 4-6min QSOs.  Experimental use of JT9-30 may 
allow QSOs down to -40dBsnr but they would take hours to complete.

http://www.physics.princeton.edu/pulsar/K1JT/wsjtx.html
http://www.physics.princeton.edu/pulsar/K1JT/WSJT-X_Users_Guide_v1.1.pdf

With better SNR, keyboard-to-keyboard QSOs could use CMSK8 down to 
-21dBsnr.  MultiPSK and PSK10 could be used down to -17dBsnr.

http://www.qsl.net/zl1bpu/CMSK/cmsk.htm
http://www.qsl.net/zl1bpu/CMSK/cmskhelp.htm

http://f6cte.free.fr/index_anglais.htm

With even better SNR, we can use CMSK31 down to ~-16dBsnr (needs testing 
to confirm).  PSK31 is useful down to -11dBsnr.  Finally for experienced 
brass pounders, CW is useful down to WSPR ~-8dBsnr.

Nominations are open for crowd-funding 630m WAS DXpeditions ...

TNX es 73 de Ralph Wallio, W0RPK
http://showcase.netins.net/web/wallio/