Hi James and All
One feature that has not really been discussed is the average feature
offered in JT65. It allows the user to average sync'd sequences that
have not been decoded. jt65 can positively identify the snyc
information down to -33. This would allow JT65 to copy down to -33
given the determination to do so. I have had much success using this on
the 2m EME path and would only expect it to work better on LF given the
greater stability of the path.
The biggest disadvantage to anything longer than QRSS3 on DX LF paths is
the fading. It has driven me nuts watching a signal come up, copy a
couple of letters or parts there of and then back to nothing. A viable
DX mode on LF must be quick enough to deal with fairly short lifts and
smart enough to remember the bits it did get in a lift to be used in the
next lift...
My gut tells me that the LF DXer of the future will be using modes like
WSPR, JT65 or JT4A or similar in relatively short periods using an
integration feature in software to grasp every scrap of signal received
and put it to use. This is where the operator becomes the critical
link. They must be able to make judgments on what to add to the buffer
and what to reject and navigate the plethora of other issues that comes
up while scraping the bottom of the band.
One final note, a JT65 user should play with the SYNC setting. After
alittle experience with the software under fire you'll know what to do
with it.
73 Scott
VE7TIL
PS - I did take a considerable amount of data during Jay's runs and can
say that on a QRSS3 ARGO grab not much is visible in the -25 and beyond
range...
James Moritz wrote:
Dear David, LF Group,
The WSJT modes are certainly effective, and quite easy to use .
However it should be noted that there is something of a disparity in
terms of ERP - The WE2XGR licences permit up to 200W ERP. I don't know
what the ERP of XGR/2 is, but looking at the station details on Jay's
web pages, I guess it must be a good deal more than 1W! So getting
signals from UK -> USA is likely to be a bit more difficult.
The dB level refers to the SNR in a bandwidth of 2.5kHz, and this
reference point is used for all the WSJT programmes. The WSPR beacon
signals have decoded successfully down to about -30dB, so should give
a bit more margin still - M0BMU has been decoded a few times across
the pond using WSPR at the 1W ERP level; I have not tried JT65A yet.
It is a bit hard to compare the signal levels required to produce
readable QRSS and successful decodes of JT65 / WSPR by looking at
off-air signals. I got the impression looking at the XGR/2 signals
using Spectrum Lab that the levels producing successful decodes were a
good QRSS3 level, and perhaps a marginal QRSS1 level, but more
systematic tests would be required. The bandwidth required by QRSS3 is
about 0.3Hz, which would improve the SNR by 39dB compared to the
2500Hz WSJT reference bandwidth. I reckon you need 6dB or so SNR to
copy QRSS, so that would put QRSS3 at about -33dB on the WSJT scale,
but this takes no account of varying signal levels.
I think the most important benefit when these modes are compared with
QRSS, etc. is the increased data rate. If you were to send "WE2XGR/2
FN31" as QRSS, it would take around 160 bit periods, about 8 minutes
at QRSS 3, as against 1 minute for JT65A and 2 minutes for WSPR. DFCW
would be quicker, at about 3.3 minutes. The reduced time period should
allow you to take advantages of short periods of higher signal
strength in the QSB. It seems to be unusual to see periods of several
minutes when a deep fade does not occur
Cheers, Jim Moritz
73 de M0BMU
----- Original Message ----- From: <[email protected]>
To: <[email protected]>
Sent: Wednesday, December 31, 2008 9:01 PM
Subject: LF: Re: WE2XGR/2 in JT65A mode last night
Interesting result Jim / Jay.
I assume that the figure that varies between -17 and -23 is the receive
value in dB.
Given the observation about local noise then it may be the case that
UK -
USA could be achieved with 6 to 9dB less power.
|
