KXV3: (1) RX ANT signal level during TX; (2) spectra at buffered IF output

[wayne burdick]

Hi all,

I'll address both of these issues.

1. RX ANT input signal handling during TX

The KXV3's ANT IN jack includes a carrier-operated relay circuit (COR).
If the COR kicks in too soon, they relay will switch off/on during
keying. So its threshold is set high. It will still serve its primary
function, which is to protect the transceiver when unsafe signal levels
are present, such as when the RX antenna is in the near field of a KW
transmit signal, or when a transmitter is accidentally connected to
this port.

This leaves a gray zone where the COR doesn't kick in, but the transmit
signal may still be very high. High-power contest stations sometimes
use additional external T/R switching if they can't avoid the use of
closely-coupled receive and transmit antennas.

We recently to a close look at this gray zone, and found that we could
make an "RX ANT jack on steroids" by adding two more diode isolation
sections. The resulting parts count is considerably higher, but we feel
that the added protection during transmit would be worthwhile. So this
change will be included in production units (once we use up our present
supply of assembled KXV3 modules).

Meanwhile, if your antenna configuration and power level warrant this
higher degree of isolation, and you already have a KXV3, you can modify
it. We're designing a small add-on board to make this easier. It will
be supplied free of charge, on request.

This unit has not yet been assigned an Elecraft part number, and isn't
available yet, so PLEASE don't request it until we announce the
details.

2. Buffered I.F. output spectra

The pick-off point for the K3's buffered I.F. output -- right at the
output of the 1st mixer's post-amp stage -- was selected to give the
widest possible bandwidth. Thus there is no filtering of any kind from
here to the BNC jack on the KXV3. This creates the opportunity to do
*very* wide spectral sweeps, if required -- limited only by the
bandwidth of the K3's band-pass filters.

As some (armed with spectrum analyzers) have noted, this output signal
includes the usual spectra one would expect from a high-level mixer.
(And it *is* a very high-level mixer.)

A mixer's output includes a broad range of products relating to the sum
and difference of its inputs (RF and LO). It also includes products
derived from harmonics of the sum and difference, etc. Collectively,
these will appear as a "comb" pattern on an analyzer. This is why the
mixer must be followed by band-pass filtering. (In transmit mode this
function is performed by the K3's narrow ham-band band-pass filters; in
receive mode, by the I.F. crystal filter and DSP filtering.)

But the buffered I.F. output is at a different point in the signal
chain. So any filtering needed for a panadapter must be included either
in the panadapter itself, or in series with it. Some panadapters will
require no filtering at all, while some will need various degress of
L-C filtering (the L-C filter used on our I.F. noise blanker is a good
starting point).

If you were only interested in a narrow band sweep, you could use a
ceramic or crystal filter. Our FM crystal filter, at about 13 kHz wide,
would be a good choice here.

Excellent shielding should be used, and possibly additional buffering,
to prevent re-radiation of 8.215 MHz back into the K3.

A related topic is the exact frequency of the buffered I.F. output. It
is nominally 8.215 MHz with our 8-pole filters, and typically 0.8 to
0.9 kHz lower with our 5-pole filters. In addition, the portion of this
passband actually present at the audio output is determined by the DSP
controls (SHIFT, WIDTH, etc.). We can work with panadapter suppliers to
provide this data via the RS232 port, and of course any future Elecraft
panadapter would also have this information available.

I'll be adding this subject to the Owner's Manual. If you're
interfacing a panadapter to the K3 and have any further questions, feel
free to e-mail me directly.

73,
Wayne
N6KR


---

http://www.elecraft.com

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
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

[Stewart Baker]

Wayne,
Thank you for your very comprehensive mail, it certainly has put
my mind at rest re potential issues involving separate receive
antennas into the K3.

73
Stewart G3RXQ
On Sun, 3 Feb 2008 14:42:19 -0800, wayne burdick wrote:
> Hi all,
>
> I'll address both of these issues.
>
> 1. RX ANT input signal handling during TX
>
> The KXV3's ANT IN jack includes a carrier-operated relay circuit
(COR).
> If the COR kicks in too soon, they relay will switch off/on
during
> keying. So its threshold is set high. It will still serve its
primary
> function, which is to protect the transceiver when unsafe signal
levels
> are present, such as when the RX antenna is in the near field of
a KW
> transmit signal, or when a transmitter is accidentally connected
to
> this port.
>
> This leaves a gray zone where the COR doesn't kick in, but the
transmit
> signal may still be very high. High-power contest stations
sometimes
> use additional external T/R switching if they can't avoid the
use of
> closely-coupled receive and transmit antennas.
>
> We recently to a close look at this gray zone, and found that we
could
> make an "RX ANT jack on steroids" by adding two more diode
isolation
> sections. The resulting parts count is considerably higher, but
we feel
> that the added protection during transmit would be worthwhile.
So this
> change will be included in production units (once we use up our
present
> supply of assembled KXV3 modules).
>
> Meanwhile, if your antenna configuration and power level warrant
this
> higher degree of isolation, and you already have a KXV3, you can
modify
> it. We're designing a small add-on board to make this easier. It
will
> be supplied free of charge, on request.
>
> This unit has not yet been assigned an Elecraft part number, and
isn't
> available yet, so PLEASE don't request it until we announce the
> details.
>
> 2. Buffered I.F. output spectra
>
> The pick-off point for the K3's buffered I.F. output -- right at
the
> output of the 1st mixer's post-amp stage -- was selected to give
the
> widest possible bandwidth. Thus there is no filtering of any
kind from
> here to the BNC jack on the KXV3. This creates the opportunity
to do
> *very* wide spectral sweeps, if required -- limited only by the
> bandwidth of the K3's band-pass filters.
>
> As some (armed with spectrum analyzers) have noted, this output
signal
> includes the usual spectra one would expect from a high-level
mixer.
> (And it *is* a very high-level mixer.)
>
> A mixer's output includes a broad range of products relating to
the sum
> and difference of its inputs (RF and LO). It also includes
products
> derived from harmonics of the sum and difference, etc.
Collectively,
> these will appear as a "comb" pattern on an analyzer. This is
why the
> mixer must be followed by band-pass filtering. (In transmit mode
this
> function is performed by the K3's narrow ham-band band-pass
filters; in
> receive mode, by the I.F. crystal filter and DSP filtering.)
>
> But the buffered I.F. output is at a different point in the
signal
> chain. So any filtering needed for a panadapter must be included
either
> in the panadapter itself, or in series with it. Some panadapters
will
> require no filtering at all, while some will need various
degress of
> L-C filtering (the L-C filter used on our I.F. noise blanker is
a good
> starting point).
>
> If you were only interested in a narrow band sweep, you could
use a
> ceramic or crystal filter. Our FM crystal filter, at about 13
kHz wide,
> would be a good choice here.
>
> Excellent shielding should be used, and possibly additional
buffering,
> to prevent re-radiation of 8.215 MHz back into the K3.
>
> A related topic is the exact frequency of the buffered I.F.
output. It
> is nominally 8.215 MHz with our 8-pole filters, and typically
0.8 to
> 0.9 kHz lower with our 5-pole filters. In addition, the portion
of this
> passband actually present at the audio output is determined by
the DSP
> controls (SHIFT, WIDTH, etc.). We can work with panadapter
suppliers to
> provide this data via the RS232 port, and of course any future
Elecraft
> panadapter would also have this information available.
>
> I'll be adding this subject to the Owner's Manual. If you're
> interfacing a panadapter to the K3 and have any further
questions, feel

> free to e-mail me directly.
>
> 73,
> Wayne
> N6KR
>
>
> ---
>
> http://www.elecraft.com
>
> _______________________________________________
> Elecraft mailing list
> Post to: [hidden email]
> 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

_______________________________________________
Elecraft mailing list
Post to: [hidden email]
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