Ok Paul
Well , Opera and Mark's systems are not the same , Opera is
recovering data , with afc/fec , and the DS , devised by
Marcus , is correlating the Opera encoded data pattern and requires
high stability etc
But there are swings and roundabouts , Opera will decode with up to
50% of the data missing , block or randomly distributed ..
From observation's there could be 6 or more db difference in the
reported s/n levels , Op is the average over the tx cycle , wspr
gives a peak level , Opera is complimentary to the ROS data mode
and indicates the path suitability for data tx /rx
Minimum Op, levels depend on various factors , one being , keeping the
'pulse' as close to original as possible , I gained -3 dB using a
'cheap sdr' on 477K over the racal Rx ..
OP32 min as used on 136 KHz , gives ~ -41 dB min , so OP2H
should run - 47 , I note Dex reached -48 , which is -54/55 on
Marcus's scale
Tnx , G,
--------------------------------------------------
From: "Paul Nicholson" <[email protected]>
Sent: Saturday, March 08, 2014 5:58 PM
To: <[email protected]>
Subject: Re: LF: 27.925 is correct dial
Thanks for that Graham.
I'm just trying to see how far short we are of an opera decode.
I just got another 0.84 dB S/N improvement by shifting the loop
a little further from the prevailing noise, and by compensating
for the slight elliptical polarisation of the signal.
Markus reported:
> 2014-03-07 06:51:53 ... -48.9dBOp 100% 19.4dB
so I guess that would now be -48.1dBOp.
But still 2 or 3dB short I think?
--
Paul Nicholson
--
|
