In practice, unexpected problems uccurred. The main problem was a limit in suppression of sidebandnoise, even when inserting extra notches on the receiver frequency.
The couse was intermodulation between the second trnsmitter harmonic and the noise energy in the non-suppressed sideband. This can occur in te circulator or in the receiver input stages.
Calculation for a 70cm repeater on RB0 (Rx freq. 434.6 MHz, Tx freq. 433.0 MHz).
|Ft||2x Ft||Fs (Ft-Fr)||Tx lower sideband|
This calculation shows, that a non-lineair component can generate a noisesignal on the receiver input frequency Fr . The signal components are:
a. The noise energy on the non-suppressed lower transmittersideband frequency: Ft-Fs
b. The second harmonic of the transmitter carrier frequency: 2 x Ft
By reducing either one a better result can be achieved.
|1.||Noiselevel in lower sideband (see fig.1):||-70 to - 80 dB|
|2.||2nd harmonic of transmitter carrier:||fi. -80 tot -90 dB|
|3.||Mixing losses in the intermodulating part:||fi. 30 dB|
|4.||Resulting intermod signal level:||-120 to -130 dB.|
Intermod level is 20 to 30 dB above the receiver noise level. In this case, the receiver sensitivity can degrade 30dB (5 S-points!).
Most effective showed an extra notch on the non-suppressed under sideband (ft-fs). A sharp bandpass filter helps, but can shown to be not effective enough.
Adding an extra notch on the mirror frequency (70 cm).