Back to the index.

Wideband ferrite antenna

For measuring purposes I needed a wideband receiving antenna for magnetic fields.
It had to work in the 150 kHz - 1600 kHz range (long wave and medium wave band).

For this I used a ferrite rod 10 mm in diameter and 200 mm long.
The ferrite material is: 3B1  (datasheet_3B1.pdf).
I bought these rods from  Farnell (order number 2103384) .

Wideband ferrite antenna - version 1

The ferrite rod is made thicker at the ends with some tape to make it fit nicely in a 16 mm PVC tube.
When the rod is in the tube it is fixed with some glue, so it won't fall out the tube.

Then a coil of 10 turns (0,8 mm enamelled copper wire) is wound over the centre of the tube.
The wires are connected to a BNC connector.
When this antenna was connected to a spectrum analyser (50 Ω input) I could clearly receive a strong local medium wave station.
The two minima in sensitivity are in the direction of the axis of the antenna.
There are two maxima in sensitivity, in the directions perpendicular to the axis of the antenna.
I expected these two maxima to have the same amplitude, but this was not the case.
In one maximum the local station at 1008 kHz was measured with - 77 dBm.
And when the antenna was turned 180 the other maximum measures - 80 dBm.
So, there is a difference of 3 dB between the two maxima of this antenna.

In my opinion, this is caused by the fact that the antenna not only receives magnetic fields, but also electric fields.
The ferrite rod is electrical conductive (about 50 kΩ end to end), and picks up a voltage in the electric field of the radio station.
This voltage is capacitively coupled to the coil, and adds or subtracts to the signal caused by the magnetic field.

In one orientation of the antenna, the signals add, and you receive more signal.
When the antenna is turned by 180, the signals subtract because the magnetic field now flows the other way through the ferrite rod.
So you then have less received signal.

Wideband ferrite antenna - version 2.

Now an attempt to make the sensitivity of the antenna in both maxima the same.
A grounded screen will be placed between ferrite rod and coil.

The coil is removed, and a screen of copper tape is placed around the tube.
There must be a small gap between the ends of the tape, otherwise it will short circuit signals picked up from the magnetic field.
A thin copper wire is soldered to the screen.

A 2 mm thick layer of tape is wound over the copper screen.
The coil is then wound on top of the tape.
The wire from the copper screen is connected to ground.
The turns of the coil are fixed in there position with some glue.

The wideband ferrite antenna version 2.

With this version 2 antenna, the two maxima in sensitivity measure almost the same amplitude of my local station on 1008 kHz (-78,8 dBm and -78,5 dBm).
The grounded screen seems to do it's job very good.

Received signal at 1008 kHz (in dBm) for different orientations of the wideband ferrite antenna version 2.
An orientation of 0 will say: the BNC connector is pointing towards the transmitter.


Some measurements.

Here are some measurements done with the antenna version 2 connected to a spectrum analyser.

Spectrum of the modulation of strong local station "Groot Nieuws Radio" on 1008 kHz  (100 kW at 41 km distance).
We see the used bandwidth is about 12 kHz, good for about 6 kHz audio bandwidth.
This measurement was done a few days before there last transmission on medium wave.
It is the last nation wide medium wave transmitter being closed down here in the Netherlands.

Also visible are carriers of foreign stations on 999 kHz and 1017 kHz.

Here I zoomed in to the carrier of the station, with the marker we measure the carrier frequency to be 1007998 Hz, so 2 Hz too low.
I don't know if the station or my spectrum analyser is off frequency.
Carrier amplitude is here -78.67 dBm, but this is constantly going up and down by about 0.5 dB caused by the modulation of very low audio frequencies.
For precise level measurements of the carrier, I use the average function of the spectrum analyser, with which it displays the average spectrum over several (for instance 10 or more) measurements.

Measuring the frequency response of the ferrite rod antenna

In the next measurement, two ferrite rod antennas are used.
One ferrite rod antenna is connected to the tracking generator of the spectrum analyser, and is transmitting this signal.
The output level of the tracking generator is 0 dBm.
The other ferrite rod antenna is placed at some distance and is receiving this signal, which is measured with the spectrum analyser.

The test setup

The received signal in the frequency range 0 to 2 MHz at several distances between the ferrite rods.
The maximum sensitivity is at 500 kHz.
The response is reduced by 6 dB at 155 kHz and 2050 kHz, this is caused by two antennas (both transmitting and receiving antenna).
I think we can conclude that one antenna has a -3 dB bandwidth of 155 kHz to 2050 kHz.

This is the frequency response of the test setup between 0 and 20 MHz.
At 20 MHz the response for transmitting plus receiving antenna is reduced by 26 dB.

Received signal in the test setup at 500 kHz, and the tracking generator at 0 dBm.


Back to the index.