[Home] [Digital Signal Processing route] [Published Articles]


[Local QRM/noise reduction] [Very small vertical magnetic loop]

[Medium size vertical magnetic loop] [Vertical magnetic Alford loop]

[Vertical magnetic loops in real life] [Circular polarization]

[Broadband amplification] [Broadband amplifier] [Single chip amplifier]

[Dual loop antenna system] [Hints]

[Phaser 80 10 meters] [Balancing and closed loop antennas]


A Two-Transistor Phaser for 80 10 meters


After using the MFJ-1025 Noise Canceling box for a few years, I noticed that I really did not use it in the array application. Mainly because it is not well-equipped for the job. It is difficult or impossible to find the right settings for the phase difference and gain ratio.


Now is time to think about a useful Phaser.

Phaser is a much better name for it when using it in the array application.



Desirable Phaser specifications

phase control

         must cover the full 360

         linear scale for maximum phase control spreading **

         knowledge of the actual phase setting

         logical scale for application in an array (linear from -90 via 0 to +90)

         reproducible and stable control

gain control

         implemented as a balance (ratio) control

         accurate by using a gain control spreading **

dynamic range

         good large signal behaviour

         negligible noise contribution


         phase and gain both frequency independent

(gain and phase control must be also independent)


         easy to build and no expensive components


** phase and gain control spreading is essential for obtaining deep nulls. Just like bandspreading is for accurate frequency control.



Phasing network

The MFJ-1025, but also the ANC4, uses a well known phasing network to control the phase from one of the antennas. The phase shift is controlled by a variable resistor.

The next schematic shows the circuit used in the MFJ-1025 and the ANC4.



The next graph shows the phase shift at 10MHz as a function of the variable resistor Rp.



The main problems of this circuit are clear. It is not possible to cover the full 180 and the control of the phase shift is very non-linear with the variable resistor. And at other frequencies this non linear behaviour is different, so it is not practical to calibrate the scale.



Linear phase shift

It is possible to get an almost perfect linear phase shift when using this circuit for both antenna signals. How: by controlling the phase shift of both antenna signals in the opposite direction at the same time with one knob.

The next schematic shows the basic new transformed circuit using a differential capacitor (Ca and Cb) for tuning the phase shift. A variable capacitor makes a much better control than a variable resistor (contact wear out).



The next graph shows the phase difference at 10MHz as a function of the differential capacitance (Ca or Cb) setting.



At 10MHz and below it is very linear. At higher frequencies, up to 28MHz, it becomes some what less linear. The full 180 is easy feasible!

The next graph shows the phase shift over frequency for 11 capacitor values (7pF, 17pF, ., 107pF).





A Two-Transistor Phaser for 80 10 mtr


(click picture to enlarge)


The gain of antennas signals A and B are first set by Pa and Pb and then the exact ratio is set by the fine balance control Pfb. If you use the broadband antenna amplifier no overload protection is necessary.

Rpa3, Rpa4, Rpb3, Rpb4, Cpa1, Cpa2, Cpb1, Cpb2 are added to compensate for the not-ideal behaviour of the transformers T3 and T4 at the higher frequency bands.

Switch S1 exchange the antennas A and B. Switch S2 inverts one of the antenna signals and makes the other 180 possible. Relays can be used instead of switches.

Important to note is that the differential capacitor can also be constructed using two normal tuning capacitors. The controlling knob must be isolated (some distance) from the capacitors.

Two pictures of the Phaser prototype may help:

Front view

Back view


Because of the much better control of the phase difference and the gain ratio, it is much easier to get deep nulls. Knowledge of the actual phase setting, linear phase control and phase/gain spreading are a must.

Noise level is about the same as the MFJ-1025 noise level, because the phasing networks are essentially the same. But large signal behaviour is much better.



Last updated: September 24, 2006