Ah, the mystery of the 5NN unfolds (well, some of the 5NN that we saw
last weekend, anyway :-). Read the attached posting from the Elecraft
reflector.
By the way, I spend a good part of CQ WW contest SWLing and watching
the built-in RTTY decoder in the K3: the K3's built in decoder (even
with the dual passband nonsense) was throwing more errors than the
particular software modem which I was using.
For the K3 internal decoding, I was using a 400 Hz roofing filter
paired with the dual passband DSP filter, with an external crossed
ellipse indicator to help the K3 tune in an RTTY signal properly.
Without precise measurements, it is hard to tell where the built-in
decoder stands among other RTTY software and TU. Alex VE3NEA (Mr.
Rocky, Mr. CW Skimmer, etc) had a couple of years ago published curves
for a couple of Windows based modems on this reflector.
Although it appears to throw more errors, to measure the K3's error
rate curve precisely, I first need to build a QSE to modulate with an
HF Channel Simulator. It is something in my stack of things to do. I
have an unused SoftRock v6.2 that I can use to generate a low level RF
signal that is modulated by an HF Channel Simulator . That way, I can
feed the K3 with an RF signal with known SNR and known and repeatable
propagation model. In preparation for that, I have already modified
my channel simulator to output in-phase and quadrature signals for the
SoftRock.
Someone with a Mac and a SoftRock QSE can eventually replicate the
experiment for the Icoms' built-in RTTY decoders, too.
The Dual Passband filter in rigs is another pet peeve of mine. Good
software demodulators that use Matched filtering get their filters
unmatched by these "dual passband" filters. Among others, K6STI's
RITTY decoder for example uses a Matched filtering. Matched filters
that are designed for RTTY will result in a specific double peak
response (because the baud rate is so much lower than the RTTY shift),
but when you cascade an external filter that does not have a flat
passband with a true Matched filter in a modem, the concatenated
filter is no longer matched to an RTTY keying signal.
The only time a dual peak filter in a rig helps is if the modem itself
has the wrong filtering relative to the keying signal.
73
Chen, W7AY
> From: Arie Kleingeld PA3A <pa3a@hccnet.nl>
> Date: September 29, 2008 11:12:01 AM PDT
> To: elecraft@mailman.qth.net
> Subject: RE: [Elecraft] CQWW RTTY - K3 Experience
>
>
> I had a few spare moments this weekend and took the CW-paddle for
> about
> 100 RTTY QSO's during the CQWW.
>
> Boy, that was very easy and, most of all, FUN. Only had to change my
> routine from 5nn a4 to 5nn 14 :-)
>
> Using the CW memories made it even more easy. I think it must have
> been
> a bit strange on the other side of the QSO seeing the long pause after
> the 5nn because the "1" takes a reletive long time in CW.
> That's also why a kept going on with the 5nn instead of 599.
>
> Anyway, flawless operation and I love the rtty dual passband. Very
> good
> filtering.
>
> 73,
> Arie PA3A
>
>
> _______________________________________________
> Elecraft mailing list
> Post to: Elecraft@mailman.qth.net
> You must be a subscriber to post to the list.
> Subscriber Info (Addr. Change, sub, unsub etc.):
> http://mailman.qth.net/mailman/listinfo/elecraft
>
> Help: http://mailman.qth.net/subscribers.htm
> Elecraft web page: http://www.elecraft.com
_______________________________________________
RTTY mailing list
RTTY@contesting.com
http://lists.contesting.com/mailman/listinfo/rtty
|