Transmitter Board Layout

[NZ0I]

The schematic design for the Mini 2-Band ARDF Transmitter has been tweaked a bit more. The latest schematic is here: https://drive.google.com/open?id=1yk6tQTWNo6adcF94_UHlC6f1EtOyIl8k

Most changes are minor, but a significant change to the low-pass filter designs has been incorporated:

Modeling and local testing of the baseline ON7YD LPF design used for performance comparison (https://qsl.net/on7yd/atx80.htm) have had it fall short of the FCC requirement that spurious emissions be at least 43dB below the power of the fundamental frequency. The ON7YD design was replaced in the previous transmitter design by 7th order Chebychev filters, which theoretically would achieve the target 43dB attenuation of the 2nd harmonic and higher.

In the latest schematic the 7th order Chebychev filters were redesigned to lower the 3dB cut-off frequency closer to the upper ends of the 80m and 2m amateur bands, and an additional capacitor was added in parallel with one of the inductors (a strategy utilized in the ON7YD design), a technique that provides greater attenuation of the 2nd and 3rd harmonics at the expense of slightly less attenuation of the higher order harmonics. SPICE models of the new filters indicate that they should provide about 65dB of attenuation of the 2nd harmonic and over 100 dB of attenuation of 3rd order and higher harmonics. That would provide a margin of over 20dB of extra harmonic attenuation above what the FCC requires. I don't expect the actual hardware to perform as well as the SPICE model, but even if the hardware falls short of modeled performance by 20dB it should still meet FCC requirements.

Another advantage of the new design: no more inductor hand winding. All the inductors are standard values of 5% tolerance parts available from DigiKey. So the inductors are more accurately wound, smaller, and cheaper than winding coils by hand from Amidon cores and magnet wire. So I expect the filters to perform closer to the SPICE models than was achievable when hand-wound parts were used.

The part footprints have all been laid out on a circuit board designed to fit inside the target Hammond box. All parts fit on one board with room to spare. I hope to order the first prototype boards before the end of 2018.

The power supply will reside on a separate board that stacks below the transmitter PCB, connected to it by board-to-board pins and socket, and fits inside the same Hammond chassis box. The two boards will constitute a fully self-contained dual-band transmitter that utilizes no point-to-point wiring, no configuration switches or jumpers, and will hopefully prove to be very rugged and reliable.