# "Top coupling" in VHF bandpass filter

I am looking at a filter design in the ARRL handbook for radio communications. It is a bandpass filter designed to pass 144Mhz with a bandwidth of around 12Mhz.

The solid line represents a shield to prevent "stray coupling". $C_{c}$ is a capacitor that "top couples" the two sections of the circuit. The posts labeled "C" are trimmer capacitors designed to tune the resonance of each coil. They connect the top ends of the inductors to ground.

I'm planning on implementing this for a receiver I am building.

What is "top coupling"? I'm confused, because the book says that the two sections are inductively coupled, but yet they are shielded and the diagram suggests that they are capacitively coupled.

Practical question: To shield the two circuit segments, should I use aluminum foil, or something else?

• There seems to be a bit of confusion over the interpretation of the picture. Can you perhaps quote some more to make it clearer? Perhaps a schematic, or the description of the labeled components, particularly $C_C$? Oct 23 '16 at 4:26
• I added a few sentences, I hope those clear up the confusion. Nov 2 '16 at 23:07

That filters has 4 sections, which if we number them 0 to 3 ...

0 is coupled to 1 by mixed coupling which is mostly inductive
1 is coupled to 2 by capacitive coupling through Cc
2 is coupled to 3 as 0 is to 1

Using all one type of coupling results in a filter with the shape tilted noticably one way or the other. Mixing the coupling type results in a more symmetrical filter.

Using a septum across the middle of the filter not only enforces capacitive coupling from 1 to 2, it also inhibits coupling of any type between sections 0 and 3, capacitive, inductive, or wave propagation.

Filters are often made with a single coupling type when a canted shape is desirable, for instance when used before a mixer, steeper attenuation on the LO and image side is a useful feature.

• How would you implement the septum? Would you use aluminum foil and just cut a small hole for the capacitor? Oct 22 '16 at 22:39
• @HunterAkins that would work electrically, but it's a bit tricky mechanically. If using alli, I'd use some thin sheet that was stiff enough to drill, and mount it with screws. My usual preference would be for a bit of single sided PCB, and tack it electrically and mechanically in several places by soldering solder-tags round it which I'd screw to the sides of the box (looks plug-ugly, but simple). Screwing to the lid is not necessary for the passband, but needed if you're going for no returns up into the very high stopband. You could also use two small boxes screwed together side by side. Oct 23 '16 at 7:14

Note that the mounting terminals marked "C" are insulated from the case, while the terminals at the bottom that have two coils attached to them are grounded to the case.

Electrically speaking, the grounded end of the coil is the "bottom", and that area has the lowest impedance (relatively low voltage and high current), which is a good match for the input and output connections. This is why the external taps are connected here.

The ungrounded end of the coil is the "top", and has the highest impedance (high voltage and low current). A capacitor connected here can be very small and still pass relatively large amounts of power.

• The terminals marked "C" are actually supposed to be trimming capacitors to tune the resonant frequency of the coils. Oct 22 '16 at 22:19