Active antenna for frequency range 10 KHz to 40 MHz
Active E-field antenna for reception of vertically polarized radio signals with high sensitivity
and high dynamic range
The explanation of the “how and why” for this amplifier design follows after the schematic diagram. Do not blame me when expectations do not come true when you did not read that part
of the document and think that an active antenna can replace a passive antenna in every
The receiving element is a
vertical rod with length between 1 meter and 2 meter, depending on the height of the antenna base point. The counterpoise is a
Placed on a massive ground
plane the conversion from E-field in Volts per meter to unloaded output voltage is equal to half the length of the vertical element. For antenna element length 2
meters the output of the antenna rod is one volt unloaded referred to ground for one volt per meter E-field strength.
The impedance of the element is about 10 pF per meter. The
active antenna is in fact just a voltage buffer with high input impedance and low output
impedance. The input
capacitance is in the order of 5 pF. The protection diodes
contribute 2 pF together and the J-FET input capacitance of this design is 3 pF.
The output impedance is low and the feedback circuit must be protected for overload by insertion of a resistor with value between 30 and 50 Ω. This gives also a proper matching to the
coaxial cable for signal transport to the receiver.
My antenna is placed in the garden, far away from local interference. The ground reference is a 4 meter iron metal pipe (blown into the ground just with water pressure; no special equipment, job done in three minutes). My antenna has base-point height of 140 cm and the appropriate antenna length for good sensitivity is only one meter.
To prevent interference coming out of my house (TV, computer, switch-mode power supply etc.) entering the antenna via the ground I used two balanced transformers in the total path. Simple solution is a balanced-unbalanced transformer to increase the impedance of this path. Just use a ferrite core and make a coil with the coaxial cable; twenty times through the core.
IN – GND -- OUT
1 meter element
Interference Ground reduction Filters Ground
The coaxial cable is single-sided impedance-matched so the frequency-dependent transfer is without surprises.
Now some more background information.
Radio waves propagation is described by the Maxwell relations. Far from the radiating source (more than 10 lambda distance) there is a fixed ratio between E and H field amplitude. The ratio of E-field in volt per meter to H-field in Ampere per meter is equal to the impedance of the ether and is 377 Ω. This can be found in text books or obtained from Wikipedia:
The active antenna is an E-field sensor and converts E-field in voltage with a factor:
has dimension “meters”. H (effective) = V (out_unloaded) / E(volt/meter) and H_eff
External noise floor
The antenna sensitivity is described by an equivalent noise floor. The antenna noise floor is preferably lower than the received noise from the atmosphere and the galactic. This noise floor is expressed in the apparent temperature of the radiation resistance. A curve of the temperature of the radiation resistance as a function of the frequency is obtained from a CCIR report and is given here. See for more information and a clear explanation
Interpretation of the radiation
resistance temperature and the output
signal of the antenna via
Johnson/Nyquist thermal noise
The assumptions are: 1; The antenna is loss-free
; The feeding point impedance is
close to 50 Ω
1 Loss-free is not a must for a receiving antenna. The loss resistance at ambient temperature 300 Kelvin, in series with the radiation resistance, gives a minor increase of the high noise from the high-temperature radiation resistance. (For transmission the loss resistance directly affects the efficiency!)
What does this mean for the S-meter reading?
At 3.8 MHz nighttime the temperature is 100 million degrees. The noise voltage of your 50 Ω
antenna, not the described active antenna, is SQRT (4 * K * T * R * BW) = SQRT (4 * 1.38E-23 * 100E6 *50 * 2700) = 27.3 µV (unloaded) or 13.65 µV terminated. When 50 µV corresponds to S9 and one S-point is 6 dB then the meter reading is slightly over S7. Daytime noise is about S5.
For twenty meters the temperature corresponds to 300 thousand Kelvin and outcome is S3. When the band is open (conditions) then also the noise level is higher. This “band is open”
situation is not included in the CCIR graph.
Noise as field strength
The temperature can be estimated with a
straight line in the graph. For 1 MHz the
temperature is seldom below 100 mega-
Kelvin, for 10 MHz it is almost 1 mega-
Kelvin and for 100 MHz it is about one
thousand Kelvin. A coarse estimation for 2this line is “MHz * MegaKelvin = 100”.
That is the red dotted line.
Short vertical monopole
The conversion factor from field strength
to unloaded output voltage (EMF) of a
short (related to the wavelength) vertical
rod is 0.5 times the length. The radiation
resistance of such an antenna is
For a rod with length 1 meter the radiation resistance as a function of the antenna is calculated
according the formula and the temperature according the red-line estimation.
MHzRr (Ohm)Mega KelvinEMF/rtHzEMF in 3 kHz 1,00,00221003,47947E-091,91E-07 3,20,0219103,47947E-091,91E-07
Frans Sessink 1992 December
The outcome is a frequency-independent constant noise field strength of 380 nV/m in SSB bandwidth of 3 kHz.
To have the antenna sensitivity to below the noise, also when the band is “dead” (no
conditions), the sensitivity should be improved by 10 dB below the red line indication. This makes no real sense.
The antenna output noise can be expressed as voltage in the load and is rather frequency-independent 200 nV rms in 3 kHz bandwidth, corresponding to 14.53 dB noise over 50 Ω at
300 Kelvin. That fits to most modern transistor receivers, but not to older transceivers.
The antenna is connected to a home-made receiver with frequency-independent noise figure of 12 dB. (The receiver is an up-conversion (IF1 = 45 MHz, filter realized with cheap 27 MHz third overtone crystals) with first mixer SRA1-H). Reception of weak signals never gives me the idea that it could be better with a special antenna. Except for real DX where a vertical tuned antenna with a good ground gives a few dB better signal to noise-and-interference ratio. Never
discovered an intermodulation product caused by the antenna.
Reception of “local signal” like Radio Waddenzee with low power at distance 200 km amazes
me every time I hear the signal (daytime, ground wave). Reception of medium wave pirates at
distance 125 km (1620-1670 kHz, recorded 2010 May 1, around 11 pm) is astonishing (evening or night-time; sky wave). Greek pirates (evening or night-time) are loud and clear, as far as their modulation allows “clear”. Medium wavefrom the USA DX (1700 kHz, 2009
January 11, 7:45 am, probably Texas) is possible in the winter time: November till March, around sunrise.
The DX-window on 80 meters, when not jammed by idiots, gives nice signals from Japan and USA, most of the time much better than my (sky-wave) tuned dipole antenna. Forty meter is also perfect with this antenna.
Above 10 MHz there is some doubt about the reception superiority of the antenna: a tuned dipole gives better results for real DX reception.
It is just a vertical reception antenna with its own noise contribution.
Document 2010 May 15
(Interference reduction filters) This is the isolator where the
cable enters the house. This is a practical solution for the
Antenna mounted on wooden support Picture made after 5 years use.
The antenna element is covered by an insulating PVC pipe to reduce the static-rain noise. The antenna element is an aluminum pipe that fits into the standard-size PVC pipe.
Wooden support: 140 cm (antenna base)
Antenna element length: 100 cm
Conversion factor of antenna (H_effective unloaded) is roughly 1.5 meter. There is some signal attenuation due to the mounting position under a tree.
C_source (antenna element) = 10 pF,
C_load (active antenna input capacitance) = 5 to 6 pF
H_eff loaded (at FET gate input) = 1.5 meter * 10 / (10 + 6) = 0.93 meter.
Active antenna output impedance is 45 Ω, load is 50 Ω. Voltage transfer from antenna
output (before 39 Ω) to load is 50 / (45 + 50) = 0.52.
Total conversion factor in 50 Ω load is 0.93 meter * 0.52 = 0.49.
Field strength of 747 kHz is 14 mV/meter (day time) at my location (Nuenen). Voltage in load is 7 mV (dry weather) and 4 mV (rain or wet leaves of the tree above the antenna).
The wet-leaves condition makes reception not worse, since the external noise is also attenuated by this capacitive screening.
Pages 6 and 7 added on 2010 June 8