The "Flevo" medium wave transmitting stationPieter-Tjerk de Boer, PA3FWM firstname.lastname@example.org
(This is an adapted version of part of an article that I wrote for the Dutch amateur radio magazine Electron, December 2015.)On September 1, 2015, the big mediumwave transmitter on 747 kHz in the "Flevopolder" (a large piece of reclaimed land in the center of the Netherlands) was switched off, after over 35 years of service. At the time it was built, this station was a marvel of technology, judging by  and , both written by a former employee of Telefunken, the station's manufacturer.
The antenna system is unique and was developed specifically for this station; it features a special combination of properties [5,6]:
1) Its pattern is not purely omnidirectional, but has a maximum towards the south-east, because in that direction the soil is dryer, causing more attenuation for the groundwave. This is achieved by using two masts and feeding them with a phase difference such that to the north-west the signals are almost in opposite phase and thus almost cancel each other.
2) Simultaneous use of two masts on two frequencies (747 and 1008 kHz); this makes the masts more complicated, but was apparently cheaper than building a total of four masts (namely two for each frequency).
3) It is an "anti-fading" antenna: an antenna which radiates mostly at low elevation angles, with a minimum at elevations of about 45 degrees. This ensures that even at the edge of its intended service area (namely the entire country of the Netherlands), the sky wave is still much weaker than the ground wave, so they cannot cancel each other, regardless of their phase.
4) The entire mast is grounded, so static charges cannot build up.
Properties 2, 3 and 4 are achieved together by the special construction of the masts.
At first glance, each mast looks like a simple pylon mast, with a "cage" of six wires around it, and in the middle a climbing ladder for maintenance. However, in fact this whole construction forms two coaxial cables or transmission lines: an (inner) coax consisting of the ladder as its center conductor and the pylon mast as its shield, and a second (outer) coax with the pylon mast as its center conductor and the cage wires as its shield. Both coaxes are cut in half by isolators at a height of 90 m, the feed point. The coaxes themselves do not radiate; the only thing that can radiate is the current on the outside of the cage, and on the outside of the top part of the pylon mast where there is no cage. See the figure: at the left a photograph, in the middle a sketch of the construction, and at the right an equivalent schematic.
Let's follow the signal from the transmitter. It enters the mast at the bottom, and via the "inner" coax (i.e., climbing ladder + pylons) it goes up to the feedpoint. There it enters the actual radiator: a vertical dipole formed by the cage wires going up and down.
The current can go up to the top of the cage, 75 m above the feedpoint. This 75 m is a quarter wave on 1008 kHz. Between this point and the rest of the radiator, namely the top part of the pylon mast, is a 75 m long coax stub which is short-circuited at the bottom; this stub is formed by the cage as its shield and the pylon mast as its center conductor, stub A in the figure. For 1008 kHz this stub is a quarter wavelength long, so it will display a high impedance at its top and thus not pass the signal. For 747 kHz the stub is shorter than a quarter wave, and therefore behaves inductively; via this "inductor" the current can go up to the top part of the mast. This is the same principle as the "trap" used in many multi-band amateur antennas, except that amateurs usually build it using a coil and a capacitor, rather than a coax stub.
Downward from the feedpoint, the current can pass through the cage wires to just above the ground. From there, there are two possibilities which are in parallel: to ground via some coils and capacitors, or to ground via stub C, consisting of the lower half of the pylon mast and the cage wires around it. All of this is used to make the current distribution for both frequencies such that the sky wave is suppressed as much as possible.
Finally, there is stub B, consisting of the climbing ladder and the pylons in the top part of the mast. This stub is connected between the two connections of the feedpoint. The stub serves as a DC path from the top half of the mast to the lower half, and thus ground the former, but of course it also influences the feedpoint impedance for both frequencies.
All in all it's a very nice piece of engineering!
One of the engineers who was involved in the design back then is Bernd Waniewski; nowadays he publishes technical descriptions and photographs of many long and mediumwave stations on his website, including many that have been (co)designed by himself, like Flevo. See .
References: W. Burkhardtsmaier: 75 Jahre Sendertechnik bei AEG-Telefunken, 1978.
 W. Burkhardtsmaier: Antennen- und Anlagentechnik bei AEG, 1987.
 J.J. Bliek: Het nieuwe MG-zendstation Flevoland, Radio Bulletin 4/1980.
 J.J. Bliek: Zendantenne van het MG-zendstation Flevoland, Radio Bulletin 6/1982.
 http://radio-tv-nederland.nl/am/am.html (mast photo)