The car array exists of 4 quarter wave whips on magnet mounts.
That would be 50cm whips on 145MHz.
Element spacing should be up to 0,25 wavelength, 50cm on 145MHz.
Be sure to have the antennas in a symmetric square, preferrably centered on the car roof.
The third switching diode in the classic V2,3 schematics -the one that switches the bottom leg in the homebase version antenna- should be deleted.
The coax shield should be hard wired to the magnet mount cup.
The size of the array should not greatly exceed 0,35 wavelength.
But it may be quite a bit smaller.
With the 145MHz array on the photo I had good results tracking down 35MHz RC transmitters for model airplanes.
In spite of the constant digital pulse (FM) on these signals, and the fact that the frequency is 1/4 of the array designing frequency,
it was easy to hunt down these transmitters.
However, being so far off the design frequency, you may have to recalibrate the RDF to be spot on.
In a city environment, multipath signals are a constant menice, resulting in erratic movements of the RDF display.
There is quite a bit of difference in performance between different arrays, some are more resistant to multipath than others.
A good symmetric radiation pattern for all 4 whips is very important.
On average, any directivity will enlarge the amplitude of multipath errors.
N0QBF did some interesting tests with standard and modified magnet mount solutions.
First he tested a standard mag mount system with aluminum crossarms, and recorded the bearing deviation in 200 points over a measuring route.
Losts of severe deviations are clearly visible on the below chart:
1, replacing the aluminum crossarms by non-metallic versions and
2, using alufoil under each magnet for improved capacitive coupling to the metal roof of the car,
the following improvement was recorded on the same test route:
The deviations are much smaller and much less common compared to the first test.
So, very good capacitive coupling is essential for good performance of the mobile array.
This is what we have to keep in mind.
A standard mag mount has poor capacitive coupling, and the coupling is easily disrupted by the curvature of the roof and by
dirt particles between the roof and the magnet.
Only the edge of the metal cup in which the actual ceramic magnet is held, is really close to the roof metal.
This is easy to improve: Just glue a thin layer of felt or thick cloth to the bottom of the magnet.
Put a clean flat sheet of alufoil beneath it, (no folds!) and wrap the foil upwards around the metal cup of the magnet.
Make sure there’s good, reliable electrical contact between foil and cup, and we have created a magnet mount that has a very high and constant capacitive coupling to the metal roof.
Thanks to the layer of felt or cloth in between, the foil will adapt to the curvature of the underlaying roof and at the same time provide a softer surface which is less likely to damage the car paint.
I measured the coupling capacity of a 80mm diam. cup magnet put directly on a painted steel sheet with an absolutely clean and flat surface, it was 473pF at best, not that bad after all.
Only realize this was a best case situation.
With the usual rubber sheet in position the capacity dropped to 155pF, way too low for a good RF ground.
But with the above construction using felt and aluminum foil it was almost 1700pF, which is really good enough for 145MHz and higher.
Ok, now let’s start building!
In the following pictures it is illustrated how I made my magnet mounts.
The magnet itself is a 80mm ferrite bus magnet with a theoretical pulling force of 70kg on a massive, polished block of iron.
You will need a strong magnet if you want to follow through with the above description of a good magnet mount.
If you have trouble finding the right magnets I can offer them to you for about 19 euro each.
Looking down into an antenna switcher.
You can clearly see the 1N4148 diode and the 1uH SMD coil
on the small PCB.
The housing is made of 32mm PVC end caps with a short
piece of pipe to connect the lower end upper cup.
The lower cup has a M8 bolt screwed in the bottom which
fits the M8 bus of the actual magnet.
The M8 bolt is also securing a 2,5mm solid copper wire of
which two ends are visible on the left side.
The small piece of PCB is soldered to this copper wire to
make a good reliable ground. Also the coax mantle is soldered
to this joint.
The top end cap holds a PL259 chassis part to accept the actual antenna whip.
The lower cup with connecting tube glued into it using PVC glue.
The top cup is glued in place using polymax glue.
This joint can be re-opened if necessary.
Before closing, only the core of the PL295 is connected to the switcher PCB using the shortest possible wire.
The shield of the PL295 remains unconnected, to minimize capacitive coupling
over the switching diode.
And now the magnet itself:
First glue a sheet of 1,5mm thick felt to the bottom
of the magnet.
Then put it on a sheet of thin aluminum foil.
Fold the four corners of the foil over the top,
and gently squeeze the foil to the magnet cup.
No real wrinkles, remember?!
Here comes the antenna switch
Just break the bolt through the foil in the top and bolt it in place.
The foil will have excellent electrical contact with the magnet cup.
This picture shows a somewhat different approach just to make it look better:
I cut the alu-foil half way the side of the magnet cup, and fixed it using black gaffer-tape.
The rest of the magnet and switcher is simply painted black.
This way it is a bit lower profile, especially on my black car.
The alu foil at the bottom will wear pretty quickly if you don’t take proper precautions handling it,
but a new sheet of aluminum is taped on in less than a minute.
One magnet mount with 1700pF coupling capacity!
The whips itself? Just cut equal lengts of about 1/4 wave of steel wire (or copper clad welding wire!) for every band you want to use the RDF for, and solder them to the core of the PL239 plug.
The length is not too critical, just make them identical.
I cover my whips with heat shrink tubing and seal them using silicone glue or polymax. Also the space between whip and the PL259 housing is filled with this mass, to make it really stable and waterproof.
Warning! Never rely on the magnetic force of the magnet mounts if you might drive at speeds higher than 50kmh.
Securing them with some strong dralon or nylon or whatever rope will keep things safe for you and the other traffic members.
Here’s the mobile antenna combiner, with 4 antenna switchers at the ends of the 4 equal length coax arms.
It is advised to tape all cables flat on the car’s roof.
This will reduce the effect of cables picking up RF.
A close-up of the combiner, just before closing the box.
The cable entries have a tight fit, but they will be sealed to prevent water to enter the box.
I didn’t bother to use a metal (shielded) box.
The isolation against direct radiation is very good anyway.
I barely hear any signal entering the receiver when all antenna switches are off.
And the isolation of switched off antennas will even improve when one antenna is selected and thus presenting a low impedance to the combiner.
This combiner uses 1N4148 diodes and 1uH SMD inductors.
Here’s is the schematics of the array.
For the RDF41 stick to this schematic.
Note that for the conventional V2,3 doppler the 1k resistors may be changed into 1uH inductors, like stated in the V2,3 schematic.
Of course, arrays may be scaled up or down for other frequency bands.
With the above technology and dimensions, good results will be achieved not only on the designing frequency, but also over a considerable frequency span up to 20% above and 20% below the designing frequency.