Re: K3 in CQ WW RTTY


ve3ki
 

Hmm ... the 30 Hz Fc offset with the 1445 pitch setting is quite
clearly visible with the tools I am using (a zoomed
waterfall/spectrum display in MixW using RX noise with a dummy load).
At a 150 Hz DSP width (too narrow for decoding, but it makes the
offset easier to see) the mark frequency is within or on the edge of
the bandpass (maybe 5 dB down) but the space frequency is obviously
well outside (response down by around 30 dB).

The effect is not as large at 1275 and 2125 settings with only a 10
dB Fc offset, but it seems to me that I can still see it with my
filter - but as you say, perhaps that is mainly due to inaccuracy in
the IF filter offset.

73,
Rich VE3KI


--- In Elecraft_K3@..., "Joe Subich, W4TV" <lists@...>
wrote:


Rich,

Part of the difficulty with the 200 Hz filter in RTTY might
perhaps
be the inability to set the filter Fc accurately enough with the
current firmware to match some of the selectable RTTY tone pairs.
I asked this question some time ago and was assured that the filter
center frequency (e.g. NORM) is correct (e.g. 2210 in FSK D with
MARK = 2125) even though the shift steps are only displayed in 50
Hz increments. My filter measurements tend to confirm that is
indeed the case.

The larger problem is that the IF filter offsets can only be set
in 10 Hz steps but the real offsets does not fit in nice 10 Hz
buckets.

73,

... Joe, W4TV




-----Original Message-----
From: Elecraft_K3@...
[mailto:Elecraft_K3@...] On Behalf Of Richard Ferch
Sent: Wednesday, October 01, 2008 9:21 AM
To: Elecraft_K3@...
Subject: [Elecraft_K3] Re: K3 in CQ WW RTTY


Part of the difficulty with the 200 Hz filter in RTTY might
perhaps
be the inability to set the filter Fc accurately enough with the
current firmware to match some of the selectable RTTY tone pairs.

The worst case is the 1445 Hz pitch setting. The centre frequency
is
1530 Hz, but the current granularity of Fc is 50 Hz and Fc NORM
is
actually set to 1500Hz. At a filter bandwidth of 400 Hz, this 30
Hz
error is not important, but if you were trying to use a 200 Hz
filter
an error of 30 Hz would probably destroy the ability to copy.
Alternatively, you could retune to centre the received signal in
the
filter and establish copy, but at the price of losing the CWT
indicator and also of transmitting 30 Hz off in FSK and perhaps
being
outside the filters at the other end of the QSO.

Pitch settings of 1275 and 2125 Hz are better, because the error
in
Fc is only 10 Hz. The 915 pitch setting has zero Fc error, but
for
those using AFSK that might not be an optimal choice (second and
third harmonics both within the Tx filter bandpass).

I am not sure whether a granularity of 10 Hz for the front panel
shift control is reasonable or not, but even if it isn't, setting
the
NORM Fc to the exact value corresponding to the selected pitch
should
be possible (e.g. NORM Fc in PSK D is 1010 Hz), and I would hope
that
a 10 Hz granularity for the IS; software command would also be
feasible.

While on the topic of software control, I note that the new MG;
command gets and sets the mic gain setting; I'd like to see the
same
capability extended to the line input setting (I use Line 4 for
PSK31
and Line 8 for AFSK RTTY - I'd like to be able to have my
software
set these automatically when changing digital modes).

73,
Rich VE3KI


--- In Elecraft_K3@..., wayne burdick <n6kr@> wrote:

Ian White GM3SEK wrote:

The 200Hz 5-pole filter for the K3, on the other hand, does
deliver
pretty much the bandwidth it says on the label - and that is
definitely
too tight for 170Hz shift. Even if tuned very precisely, it
causes a
noticeable deterioration in copy.
Hi Ian,

We designed this filter specifically for narrow-shift RTTY,
PSK31,
and
CW use in high-QRM situations. The bandwidth was selected in
consultation with some of our field testers who specialize in
RTTY
contesting. I believe they've had very good results with it.

That said, you could easily widen a 200-Hz 5-pole filter as an
experiment. I just looked at the model. You only need to change
two
capacitors -- C4 and C5. I'd suggest using 820 pF or 750 pF,
which
result in a -6 dB bandwidth of about 270 Hz or 300 Hz,
respectively.
This should also decrease loss through the filter by about 1 dB
(set
your FLx GN menu entry accordingly). Return loss and passband
flatness
are essentially unchanged.

73,
Wayne
N6KR


---

http://www.elecraft.com


------------------------------------

Yahoo! Groups Links


Join Elecraft-K3@groups.io to automatically receive all group messages.