[HCDX]: MFJ-1026, other phasing unit designs

[wa1ion@xxxxxxxxxxxxx (Mark Connelly)]

Thoughts About the MFJ-1026 and Possible Future Designs
Mark Connelly - WA1ION - 14 OCT 1997

Greetings to all.  Al Merriman reported some troubles with
intermodulation spurious responses (mixing products) on his
MFJ-1026 phasing unit.  I have noted a few of these myself 
recently.  Generally these are out-of-band (e.g. just below
530 kHz and just above 1610 kHz), but a few on 5 kHz multiples
within the medium wave band have been observed around local
sunset.  I've determined that these result from SWBC stations
in the 49 m band mixing with those in the 41 m band.  Field
tests done on Saturday (11 OCT) at Red River Beach in Harwich
(Cape Cod, MA) and at the Chatham Lighthouse indicated a few 
"problem frequencies" before MW signals are up to full night-
time skip strengths.  There's an RTTY (or similar) navigational 
station on 325 kHz in Chatham and a powerful LORAN station in 
Nantucket on 100 kHz.  Second and third harmonics of these 
stations were audible when the MFJ-1026 was used.  At home in
Billerica, MA, 3 miles / 5 km from WRKO-680, harmonics (1360, 
2040) are similarly audible.  These observations indicate that 
the various FET amplifier stages in the MFJ-1026 are not as 
"muscular" as they could be.  In urban areas especially, 
operation of a broadband-amplifier-based phaser could offer 
challenges.  The same sort of problems had previously been 
noted (and to a worse extent) in the JPS ANC-4.  It's clear
that the designers of these commercial units were not sitting
in the same Third Order Intercept classes that Dallas Lankford
and Ulrich Rohde must have attended.

The passive R-L-C and passive delay-line phasing units that 
have served DXers well for many years are, of course, far less 
susceptible to second and third order spurious products (mix
spurs, harmonics).  Still, if a bit of forethought is exercised, 
some good results can be extracted from commercial units such
as the MFJ-1026 and JPS ANC-4.

In-line attenuators may be useful in some situations.  Or just
use shorter antennas (although this might rob you of MW strength
while the SWBC signals causing intermod's may still be strong
enough to be a problem).  The right way to manage things in strong-
signal areas would be to use passive L-C based tuned circuits
ahead of each phasing unit input.  These could be bypass-able
for generally bandscanning and then switched in only when doing
critical listening (where the possibility of spurious responses
must be kept to a minimum).

Very few broadband high-impedance-in / low-impedance out amplifiers 
can handle very strong signals without spur-causing distortion and 
still maintain a superb noise figure for weak-signal work.  Dallas 
Lankford's whip amplifier designs (see National Radio Club reprints) 
come close to the ideal, though some consider the gain of the 
circuits as inadequate.

Some of the things I'll be researching over the next year:

1. A phasing design with a variation of MFJ's bridge phase-
   shift circuits, but with less amplification (or amplifiers
   with higher input intercepts)

2. A Broadband Active Loop that can be phased against a
   co-located active whip to provide cardioid patterns.
   Presently the loop-whip cardioid application (useful for 
   antennas mounted on a car roof) requires the loop, and 
   sometimes the whip, to use remote varactor tuning.  Having 
   both antennas broadband would increase DX productivity on
   beach (and other) DXpeditions where site space and privacy
   are not adequate for better types of antennas.  The trick 
   to making this work is a high-gain amplifier design having
   a good noise figure for weak signals and a high intercept
   point to prevent spurs in challenging urban areas.

3. A series of Input Conditioning units to use ahead of the 
   various commercial and homebrew Broadband Active Phasers.

   These will include -

*  Active Antenna adapter.  Three active whips (or loops) get
   coupled through switches to go to the phasing unit's two
   inputs.  Each active whip input port can be the actual 
   active whip amplifier's output or a passive input consisting
   of the pick-up from the cable going to the associated antenna.
   This scheme yields three primary pairing combinations of active
   inputs (1 & 2, 1 & 3, 2 & 3), three similar primary passive
   combinations, and 30 additional unique combinations (e.g.
   Channel 1 = Active Ant. 1; Channel 2 = Passive Ant. 2 added
   to Active Ant. 3).  If three 30 m / 100 ft. cables are run 
   out to the respective active antennas at 120 degree horizontal 
   angular spreads, a very versatile antenna array is created
   without using a great deal of land.  The equilateral triangle 
   formed by the three active antennas can give cardioid patterns 
   every 60 degrees around the compass.  If the broadband active 
   whips (or loops) are at bearings of 15, 135, and 255 degrees 
   (clockwise from north), cardioid peak-null axes at 45, 105, 
   165, 225, 285, and 345 degrees can be produced with the primary 
   pairing (active) combinations: 1 versus 2, 1 versus 3, 
   2 versus 3.  With 33 other pairing possibilities (including 
   blends of some antennas operated as active and others as 
   just end-fed unpowered "feedline longwires"), the variety 
   of nulling / peaking patterns is quite extensive.

*  Controller adapter.  This is similar to the head-end of the
   DCP-2 Dual Controller / Phaser.  It allows control of remotely-
   tuned loops and whips and also of Byan-type Vactrol remotely-
   controlled Beverage / rhombic terminations.  The outputs of
   the selected two remotely-controlled antennas are fed out to 
   the two inputs of the phasing unit.

*  Noise-reducing adapter.  Two balanced-feed input pairs are
   switched through 4:1, 1:1, or 1:4 balun transformers.  This
   is similar to a dual version of the "Bevmatcher" circuit
   described in NRC's reprints.

*  Tuner head-end adapter.  Two L-C tank circuits are used.
   Each is placed between the given antenna and the associated
   phasing unit.

>From time to time, I'll keep the DX community aware of the progress
of the research and development efforts.

73 / good DX ... Mark
------------------------------------------------------------------------
This is a message from wa1ion@xxxxxxxxxxxxx (Mark Connelly)
to hard-core-dx@xxxxxxxxxx list. To unsubscribe the list, send
"unsubscribe hard-core-dx" in mail body to majordomo@xxxxxxxxxxx
For more information, please check http://www.iki.fi/rko/hard-core-dx/
or email Risto Kotalampi, rko@xxxxxxx
------------------------------------------------------------------------