Topband: RX 4 square

[Richard (Rick) Karlquist]

Tom, W8JI wrote:

 > First, someone posted that a simple Minicircuits transformer
 > could replace the voltage follower. This appeared on
 > Towertalk a year or so ago, and again on Topband just now.
 > I'm very surprised that comment was made.

 > The statement a voltage follower can be replaced by a
 > conventional transformer with no change in performance is
 > very clearly not even remotely close to being factual,
 > although there are certain cases where a transformer can be
 > made to "work".

I was the person who made the postings.

If a voltage follower is replaced by a transformer, there
will be a change in the GAIN of the system, but not the
SENSITIVITY of the system.  As Tom has pointed out many
times, the GAIN of a receiving antenna is of no consequence
as long as the sensitivity is limited by external noise.
We can debate the semantics of whether "performance" refers
only to sensitivity or also to gain.  If you insist on equal
gain, you can always put a 50 ohm to 50 ohm preamplifier between
the transformer and the receiver.

But, then in that case, what are the advantages over the
voltage follower?

1.  In a four square, you only need one amplifier following the
combiner, rather than four amplifiers at the antennas.

2.  You don't have to build the amplifiers into weathertight boxes
and supply DC power to them remotely.

3. The 50 ohm amplifier is cheaper than  the DXE voltage follower,
which is only available as part of a $250 product.  Many 50 ohm
amplifiers are for sale for much less than that price.

4.  It can have a lower noise figure, not being constrained by
needing to be built with a FET, etc.

5.  It is compatible with a BCB reject filter with 50 ohm ports.
I don't know how to make a BCB reject filter with 3000 ohm ports.
The DXE tuned circuit provides only limited BCB rejection.

6.  A transformer has very little stray capacitance to tune out.
A  FET follower has more capacitance to tune out.  There is a lot of
variation in capacitance from FET to FET, which may result in
variations in phase.  Transformers are extremely repeatable.

Next, let's look at the question of noise figure.  My FT-1000D has a
noise figure of about 12 dB.  If I preceed it with a T36-1 transformer,
it has a noise figure of about 13 dB, if I assume 1 dB loss in the
transformer.  Only now, its input impedance is 1800 ohms instead of
50 ohms.  If I am in a very quiet area or use a very short whip, I
may not be limited by external noise.  In that case, I can make a preamp
with < 1 dB noise figure very easily with a 2N5109 transistor.
I recommended using a somewhat longer whip to make sure that the
sensitivity is indeed limited by external noise, that making the preamp
unnecessary.  I would submit that certainly anyone in an urban area
will not need a preamp with a 20 ft whip.

In the case of the voltage follower, the system is constrained by the
noise figure of the voltage follower.  Its probably better than my FT1000D,
but not as good as a 2N5109.  Let's estimate 3 to 6 dB.  Thus on
a noiseless desert island, the active whip is better unless a preamp is
used, in which case the transformer is better.

The issue of transformer ideality was raised.  I admit that I was skeptical
of transformers myself, until I tried some in the lab.  They are actually
quite good.  You really can present the whip with upwards of 1000 ohms.
In fact, you can cascade transformers and achieve 5000 ohms.  Now with the
T36-1, you are presenting 1800 ohms to a 3000 ohm source, if a CB whip is
used on 160m.  This results in a mismatch loss of only about 1 dB.  If 
the whip is
increased to 20 ft, the source impedance drops below 1800 ohms and even this
1 db mismatch loss goes away.

EXPERIMENTAL RESULTS:

I purchased two of the DXE active whips.  On one of them, I replaced the 
amplifier
with a transformer.  The other one was unmodified (as a "control").  In 
A/B tests,
both were equally sensitive, being limited by external noise.  The 
S-meter reading
was lower on the transformer equipped whip, of course, but the signal to 
noise
ratio was the same.  I added a  50 ohm preamp to the FT1000D, and then 
the two
systems were completely indistinguishable from each other, except that 
the BCB
birdies were not present on the transformer version.  I suggest to anyone
who does not believe this: try it yourself.

Rick N6RK