DC Impedance bond and signal filter - metro application problem

Thread Starter


Joined Nov 22, 2008
The following sketch is of an old impedance bond in a track-circuit for a metro line operating on 800 VDC traction supply. In addition, there is a 75Hz 3 VAC voltage injected onto the rails for train detection. The 7 + 7 windings pose no resistance to the traction return current of 2000 A, however, the 75 Hz signal must see infinite impedance in the filter, i.e. not pass through the 14 windings. These windings are wound on a standard transformer laminate - secondary winding uf unknown number of turns is loaded with a 10micro Farad capasitor.

The question is: Disregarding the DC-current, what is the filter equation for this 75 Hz band-stop filter?

The second question: The 2x7 windings turn the same way, but with opposing currents, thus when carrying even currents, opposing fluxes will cancel each other under ideal conditions. If however the impedance bond is loaded unevenly, the iron core will saturate at some point from the DC current. What will happen to the filter characteristics?

Does anyone know a resource that discusses this particular filter. I am too old to have kept all my books from school :)



Joined Dec 7, 2008
Let me offer my humble ideas.
1. The receiver of AC track circuits should get the AC signal at 75 Hz, 83-1/3 Hz.
2. The receiver should not get the DC traction return current.
3. The high d.c traction current must be provided with a low impedance ground path.
The circuit does exactly this.
The transformer primary (with heavy gauge winding) does not impede the traction current. In fact the magnetic flux (dc or power harmonics) generated by the currents in two rails nearly cancels out. I X7 - I X7 ampere turns.
4. The secondary would typically have a turns ratio of 1 to 5 for older AC track circuit vane relays. The turns ratio ensures, good signal level at the receiving coil and some impedance matching.
5. The capacitor filters out any AC audio frequencies greater than 500 Hz, which may come from unbalance in rail current and residual ac in traction current. The capaicitor provides about 200 ohm shunt path for 75 Hz, which does waste some of the signal power. (for 500 hz noise, it is only 30 ohm)
So it is not a conventional band stop filter.
6. It is also possible to choose a turns ratio so that the inductance of the track (multiplied by n squared) resonates with the10 microF capacitor at 75 Hz. But this is rare.
Hope this helps.