Fig 3—Block diagram of phasing-method audio quadrature generator.

Fig 3—Block diagram of phasing-method audio quadrature generator.

The other inputs to the mixers, lc and Qc, are carriers 90 degrees out of phase from each other at a constant frequency of 1477 Hz, the center of the input audio passband. The mixer outputs go to the inputs of the low-pass filters, and the filter outputs are the desired signals la and Qa-

The low-pass filters pass frequencies from dc to 1152 Hz, half the audio passband. The low-pass filters block frequencies of 1477 Hz and above. This is a very steep skirt:

Fig 4—Block diagram of Weaver-method audio quadrature generator. Ic and Gc are 1477-Hz carriers. Qc is 90° out of phase from lc.

stopband / passband = 1477 Hz / 1152 Hz = 1.28. An eighth-order elliptic filter will provide this performance and is available in IC form as the Linear Technology LTC1064-1.

The passband in la and Qa is thus dc to 1152 Hz. The original audio passband is folded into the la, Qa passband, so the audio passband is 1477 - 1152 = 325 Hz to 1477 + 1152 = 2629 Hz. Two frequencies in the audio passband, 1477 Hz-f and 1477 Hz + f, are converted by the mixers to the same frequency f in the la, Qa passband. For one of the two, la ieads Qa by 90 degrees, and for the other, la lags Qa by 90 degrees.

As in the phasing method, these phase relationships cause the 1477 Hz - f signal to appear on one side of the RF carrier, and the 1477 Hz + f signal to appear on the other side of the RF carrier. Thus, the exciter RF output signai is double-sideband, suppressed-carrier emission with different information in the two sidebands. This signal appears to be an SSB signal to a receiver. A conventional receiver dial will be set 1477 Hz away from the RF carrier frequency in the transmitter for proper reception.

Both the phasing method and the Weaver method will produce unwanted outputs if the phase or amplitude balance of the la and Qa channels are not correct. These unwanted outputs are on the other side of the RF carrier frequency from the desired outputs.

For the phasing method, these unwanted outputs are outside the SSB passband and interfere with other stations on adjacent frequencies. For the Weaver method, these unwanted outputs are in the SSB passband and appear as audio distortion rather than as interference with other stations. This is a major advantage of the Weaver method. Good audio quality requires only 20 dB of attenuation of unwanted outputs, while the FCC requires out-of-band spurious outputs to be reduced more than 40 dB.

One unwanted output in the Weaver method is especially troublesome. Dc offsets in la or Qa produce an RF carrier output in the center of the passband. Imbalances in the RF mixers have the same effect. This output can be reduced enough for a transmitter, but limits the dynamic range achievable in a receiver built with the Weaver method.

Wright had trouble with unbalances in his audio mixers, and with getting steep enough filter slopes. Modern parts solve these problems completely.

Was this article helpful?

0 0

Post a comment