EQUIPMENT: MFJ Phasing Unit initial comments

[wa1ion@xxxxxxxxxxxxx (Mark Connelly)]

---------------------------------------------------------------------
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
Categories: ADMIN, BULLETIN, CONDITION, CONTEST, EQUIPMENT, FM, INTERNET,
LIST, LOG, MEETING, NEWS, PIRATE, PUBLICATION, SCHEDULE, QSL, QUESTION.
For more information, please check http://www.iki.fi/rko/hard-core-dx/
or email Risto Kotalampi, rko@xxxxxxx
---------------------------------------------------------------------
Initial Review of MFJ-1026 Phasing Unit
Mark Connelly, WA1ION - 20 JUL 1997

I've just received a new MFJ "Deluxe Noise Cancelling Signal Enhancer", 
model number MFJ-1026.  It offers the ability to null interference, 
whether from electrical noise sources or actual transmitters.  This is 
accomplished by creating a 180 degree phase shift between two antennas 
that are presenting equal-amplitude "pest" signals that cover desired 
DX.  This is, of course, the same principle used in the many "homebrew" 
/ "d-i-y" units out there (and also other commercial boxes such as JPS 
ANC-4 and SEM QRM Eliminator).

In many cases, the unit could produce good nulls on steady groundwave 
broadcast signals and electrical noises.  The first thing I noticed, 
though, was very high insertion loss on the lower half of the medium 
wave band.  Loss measured 5 dB or less above 1600 kHz, about 10 dB at 
1300 kHz, 20 dB at 1000 kHz, 25 dB at 800 kHz, and a whopping (and 
totally unacceptable !) 30 dB at 600 kHz.  The MFJ advertisement states 
"Works on all modes, VLF to VHF".  Strong local beacons "LQ" on 382 kHz 
(Lynn, MA) and "LW" - 402 (Andover, MA) are over S9 straight off the 
antenna, but were barely audible with the MFJ-1026 in line.  We're 
talking about a loss in the 50 dB range on LF/VLF.  

A quick examination of the circuit revealed the problem: Components 
L4/C8/L3/R26 on the main antenna input and L5/C16/L6/R27 on the 
auxiliary antenna input create a substantial amount of attenuation on 
frequencies much below 2 MHz.  Throwing warranty coverage to the wind, 
I attacked the circuit board with the diagonal cutters and removed L5, 
L6, R27, L4, L3, and R26.  What a difference !  Now there's plenty of 
signal across the entire spectrum from 150 kHz to 30 MHz.

Two other obstacles were encountered: The phase adjustment range has 
SLIGHTLY LESS than 180 degree coverage.  Swapping the two antennas 
makes this a non-issue because enough "lag" can be produced in cases 
where enough "lead" can't be (and vice versa).  I put a DPDT "swap 
switch" in to allow the auxiliary antenna to go to the main path and 
the main antenna to go through the auxiliary chain.  Either the "as 
built" or reversed-inputs configuration can now be selected and nulls 
can be obtained in almost every circumstance.  Besides the phase 
adjustment range problem, there's the not surprising problem of 
active-broadband-circuit overload caused by 3-mile / 5 km distant 
WRKO-680 (50 kW).  The simply remedies here are to use tuned inputs 
such as loops or L-C tuned whips/wires - or, in cases of untuned wire 
inputs, to just use less gain (as selected by the two 250-ohm input 
level pots).  Passive L-C and passive delay-line phasing units 
typically do not have overload / intermodulation product problems, even 
with fire-breathing WRKO down the street.

Over all, I would say that the MFJ-1026 ... WITH THE ABOVE 
MODIFICATIONS ... is a very capable phasing unit.  I feel that it is 
clearly superior to the ANC-4.

A formal review / article will be issued in a month or so, after I give 
the unit a serious beach-DXpedition "shake-down cruise".