The biggest problem with mixers feeding directly into a crystal filter is,
the mixer always has TWO outputs (the sum and the difference frequencies).
One of them is the desired frequency (i.e. 9 MHz), the same as the crystal
filter, and the second one is on another unwanted frequency.
The crystal filter has a design impedance which should be designed to match
that of the output of the mixer, but it is only that impedance on its design
frequency (i.e. 9 MHz). It terminates the desired frequency just fine,
but...it's this second signal that causes the problem, because it is not
properly terminated by the crystal filter. This lack of proper termination
for this one specific signal causes standing waves at that point.
So the culprit is specifically the standing waves caused by improper
termination of the unwanted signal, which may causes ~30 dB of degradation
the IP.
The best way to deal with this is with a "diplexer", which is a splitter
which separates the two signals, routes the wanted signal to the crystal
filter (where it is properly terminated) and routes the unwanted signal to a
tuned circuit or trap, resonant on its frequency.
The diplexer is inserted between the mixer output and the crystal filter and
traps the unwanted signal, preventing it from reaching the crystal filter,
thus preventing the standing waves and maintaining a high IP.
So as I see it, it is not so important to have "broadband termination" as it
is to have "TWO specific terminations", one for the desired signal and one
for the unwanted signal.
Broadband termination is also a way of addressing this, and it also works,
but it is not nearly as effective as using a diplexer.
Les Hayward, W7ZOI popularized this concept in the mid 1970s.
If I dig deep enough in my files, I can probably find one of his articles
explaining this, which as I recall, were published in Ham Radio Magazine.
BTW, as I recall, Ten-Tec introduced a diplexer with its first OMNI but that
was long ago and my old grey cells may be failing me on that point. Easy
enough to check if someone has an Omni schematic handy.
73
Rick, DJ0IP
-----Original Message-----
From: TenTec [mailto:tentec-bounces@contesting.com] On Behalf Of Barry N1EU
Sent: Tuesday, March 12, 2013 11:54 PM
To: Discussion of Ten-Tec Equipment
Subject: Re: [TenTec] Corsair vs Corsair II
On Tue, Mar 12, 2013 at 10:32 PM, GARY HUBER <glhuber@msn.com> wrote:
> Mike,
>
> You are welcome. I regret that I cannot locate the document where I
> originally saw the reference to Corsair / Corsair-II having ANZAC
> receiver design and later commentary on later TEN-TEC receiver design
> being changed to avoid Copyright or similar legal issues.
>
Here you go, posted 2007 to this list by Jerry K0CQ.
73,
Barry N1EU
The bipolar RF (and first IF) stage uses transformer feedback with a three
winding transformer. Its a circuit developed for the output stages of cable
TV distribution amplifiers where acceptable intermod is 70 dB down and in
Ulrich Rohde's book on receiver design he sees nothing close to having its
performance. A double gate MOSFET doesn't come close. I've run Kenwood with
MOSFETs and those MOSFETS don't come close to the strong signal handling of
my Corsair II.
That circuit is covered by an Anzac patent or two and that may be why Tentec
went to the grounded gate JFET circuit in the Omni V and VI.
Those use four JFETS in parallel with individual source resistors to make
them balance better. I've not seem their performance compared directly to
the Anzac circuit, but the Omni V and VI seem to do quite well in the
intermod department.
Schottky ring mixers are hard to beat and the higher the LO power (so long
as the mixer is designed for it) the better the strong signal performance.
Its also important that at least two of the three ports of the ring mixer be
terminated over a wide range of frequencies, else NF, mixer loss, and
intermod can be a lot worse. The mixer output often sees a crystal filter
and that is the worst thing that can be done, one book on mixers says that
can cost 30 dB intermod range. The Corsair II, Omni V and VI use a broad
band IF stage with lots of dynamic range to provide that IF port broad band
load and it shows. It works.
Don't know about the Omni V and VI, but the "attenuator" switch on the
Corsair II actually removes the RF stage from the circuit. There is no
attenuator.
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|