# Coupling-matrix filter — tuning sheet # rf.akzami.com Requirements centre 2400.000 MHz band 100.000 MHz (4.17 %) resonators 4 return loss 22.0 dB zeros ordered 2250.00, 2560.00 MHz zeros obtained 2250.00, 2560.00 MHz unloaded Q not given in-band loss 0.028 dB Coupling matrix, order S,1..N,L 0.00000 1.07648 0.00000 0.00000 0.00000 0.00000 1.07648 0.00000 0.94382 0.00000 -0.07435 0.00000 0.00000 0.94382 0.00000 0.75461 0.00000 0.00000 0.00000 0.00000 0.75461 0.00000 0.94382 0.00000 0.00000 -0.07435 0.00000 0.94382 0.00000 1.07648 0.00000 0.00000 0.00000 0.00000 1.07648 0.00000 Couplings S-1 M +1.07648 k +0.04485 1-2 M +0.94382 k +0.03933 0.952 mm 2-3 M +0.75461 k +0.03144 1.140 mm 3-4 M +0.94382 k +0.03933 0.952 mm 4-L M +1.07648 k +0.04485 1-4 M -0.07435 k -0.00310 out of reach cross External Q input 20.71 output 20.71 Resonator tuning 1 2400.000 MHz 2 2400.000 MHz 3 2400.000 MHz 4 2400.000 MHz Geometry: Rogers RO4003C, 0.508 мм resonator impedance 50 Ω hairpin arm 19.269 mm track width 1.097 mm fold gap 1.016 mm tap from the fold 3.536 mm Tuning procedure 1. One resonator alone. Excite it with a loosely coupled probe and trim the arm until the dip sits on the band centre. The probe must be loose enough that the dip stops moving when it is loosened further. 2. Each pair on its own. Place two resonators at the computed gap, measure the two resonances f1 and f2 and take k = (f2^2 - f1^2) / (f2^2 + f1^2). Move the gap until k agrees with the table above. There are as many pairs to do as there are distinct k in the table. 3. External Q. Connect the tap to one resonator and take the phase of S11: Qe = f0 / df, where df is the spacing between the points at which the phase has moved +-90 degrees from resonance. Move the tap until Qe agrees. 4. Assemble and check. What will disagree first are the couplings through a resonator, which the pairwise measurement did not see. The sign of a cross coupling is not set by the gap. It is set by which parts of the resonators face each other: at the open ends the coupling is electric, at the folds it is magnetic, and their signs differ. The table gives the sign — get it from the placement, and the magnitude from the gap. Caveats - computed without loss: no resonator Q was given. In a narrow-band filter that is exactly what eats the decibels, so the plot is kinder than the board will be - the coupling 0.00310 between resonators 1 and 4 is beyond this model: once the hairpins are far apart the resonance splitting sinks out of sight and the gap can no longer be read from it. A coupling that weak is not usually made with a straight gap between the arms anyway — it is obtained by bringing the ends of the resonators closer in a folded layout, which is a different geometry