A notch filter
Alan Melia (G3NYK) and Hugh (M0WYE)
After discussions with Hugh M0WYE, I sent him the circuit and parameter value for a notch filter that should reduce the strength of the German utility station DCF39 on 138.83 kHz by 30 to 40dB without giving more than a couple of dB loss inside the amateur 136kHz band. He made a quick practical example of the circuit with a small inductor and achieved results very similar to my simulations. This is offered as a possible experimental solution with no guarantees that it will work in every situation.
The simulation has been made with a coil resistance of 0.1 ohm, so a quite easily achieveable Q. Hugh managed to get a small ferrite bobbin coil (rewinding it with a centre tap) to meet this requirement. The higher the Q of the coil the higher is the resistance required to set the best notch depth. According to Terman (Radio Engineer's Handbook 1943) This should be a quarter of the parallel resonant impedance of the LC circuit. As far as I can see at the moment the only problem is that the output of the filter must not be loaded too heavily, so a FET source follower is probably called for ( J310 running reasonable amount of source current to keep the IMDs down ??). The simulation used a 10k0 load and lower values lead to a reduction in notch depth and sharpness. For practical purposes the capacitor C needs to have some propotion of the capacity variable to tune the notch in "on the nose". Then the resistor is adusted for maximum attenuation. It is very sharp, and it may not be possible to achieve best attenuation with a single sweep trimmer. You will need to experiment with the components you select to use. The simulation uses the AC Analysis program available as PD freeware in my Programs area
Return to homepage