This filter became a requirement for my station because I co-located a 2m yagi within an EME array of four 70cm yagi antennas. See the picture on the right. Antenna simulations and experience have shown that the 2m yagi will cause negligible deterioration of the 70cm EME array performance as long as it is positioned exactly in the center of the four 70cm yagis.

Still, as I had expected, the coupled 432MHz RF energy into the 2m yagi was higher than what my 2m receiver front end could tolerate. I actually measured 100mW at the 2m antenna when transmitting 300W into my 70cm EME array. This is a -35dB coupling between antennas. It is actually better than what I had expected, but it is still probably too much for the 2m rig to cope with. Maybe there is a low pass filter at the input of the 2m rig, but I did not want to try it!

Hence the need for a filter. A low pass filter combined with a notch filter appeared to me as a good way to do this. A commercial diplexer might have done the trick, but I did not want to spend, I wanted to build!

THE CIRCUIT

The main design objectives were: At least 30dB of attenuation of the 432MHz signal. Minimal insertion loss (S21) at 144MHz, < 0.15dB Low VSWR (S11 or return loss) at 144MHz, < 1.5:1 Be able to support at least 100W of 144MHz RF. The design activity was a combination of theory, gut feeling and trial-and-error. The resulting filter circuit consists of a low pass filter stage (L1, C1, L2) followed by a Series-LC notch filter stage (L3,C2). This is shown on the figure to the right. Simulations were performed using Ansoft Serenade SV v8.50. I used a Q (quality factor) of 50 for the inductors and 500 for the capacitors. The resulting simulation plots are shown on the right. With these results on hand, it was time to heat up the soldering iron!

Click on the figure to enlarge it. Click on the figure to enlarge it.

CONSTRUCTION

For the enclosure, I used a similar type filter box recovered from an AM-6155 FAA RF amplifier. It had compartments which may or may not improve the 432MHz notch performance. As well, since the enclosure had more compartments than needed, I connected one end of the filter circuit to the N type connector using a piece of semi-rigid UT-141 coaxial cable. This is visible on the picture below, in the upper-right of the enclosure. You may build a box using PCB material or use an off-the-shelf enclosure. This is not really critical. I would pick a large enough box so that the inductors do not come too close to the walls.

The variable capacitors I used are of 350V ceramic type . These will work fine with a 144MHz RF power of at least 100W. The suggested capacitance ranges are written on the circuit schematic. As for the inductors, their values and construction details are also provided on the schematic above. I used enameled copper wire, but bare copper wire will work the same. Of course, never use stranded wire for winding coils...

Click on the figure to enlarge it.

TUNING AND RESULTS

The tuning of this filter consists of first nulling the notch filter by using a 432MHz signal source (RF generator level below 1mW or 0dBm) and an RF power meter or 432MHz receiver. Adjusting C2 for minimum signal is the objective. L3 winding spacing can also be changed and C2 re-adjusted to make sure that the best null can be achieved. Tuning is quite sharp, so go gentle on the tuning screwdriver. Using an RF power meter and a signal generator could be tricky as the power meter will also pick up the generator's harmonic spurs at 864MHz and 1296MHz. A double check with a 432MHz receiver or a spectrum analyzer is probably a good idea. The second step in the tuning process consists of minimizing the VSWR at 144MHz. This is done by tuning C1 for best VSWR when a 144MHz transmitter is connected to a dummy load or antenna through the filter. Use a low enough power to minimize the risk of transmitter damage. If a good VSWR cannot be achieved, change L1 winding spread and re-tune C1 for best VSWR. Adjustment of L2 may also help dip the VSWR. Expect little interaction between the two filter stages. But as a final pass, the above two steps should be repeated so that any interaction, even small, between the two filter stages is taken into account. The resulting performance should resemble what I have measured on my filter. This is shown on the three figures on the right hand side. The first figure highlights the very small insertion loss at 144MHz. I measured less than 0.1dB. The second figure shows the resulting notch at 432MHz, a depth of greater than 60dB. The third figure shows the S11 reflection loss close to -50dB at 144MHz. This is a 1:1 VSWR. With such a filter, I don't feel nervous for my TS-2000 if I feed 300W at 70cm into my EME array. Next, I need to make sure that I don't blow up my 70cm preamp when I feed 100W into my 2m Yagi!...

Click on the figure to enlarge it Click on the figure to enlarge it Click on the figure to enlarge it