June 5, 1997

Modulator draws just 5 mA at 2.7V

Kevin Bilke, Maxim Integrated Products, Reading, UK

Single-sideband (SSB) modulation uses the frequency spectrum and generates output power more efficiently than does full-amplitude modulation. Although not used for data transmission, SSB modulation is still popular for the transmission of voice at HF and low-end-VHF. The circuit in Figure 1 generates 35- to 80-MHz SSB signals by combining wideband, low-voltage op amps with an IC that integrates all the necessary functions. All the ICs operate at 3V±10%.

The traditional method for producing SSB signals is to modulate a carrier and then filter the output to remove the unwanted sideband and carrier frequencies. This approach is wasteful because the system can dump as much as two-thirds of the generated power into a filter. (Note that filtering does not always occur at the output stage, so the system doesn't necessarily waste two-thirds of its transmitted power.)

An alternative method for SSB generation is the phasing, or algebraic, method, in which two modulators (mixers) produce the desired sideband while suppressing the unwanted carrier and other sideband. Two modulators for this purpose, normally used for I and Q modulation in a quadrature-amplitude- modulation signal, are available in IC₁. The resulting circuit offers several advantages, including low-power and -cost operation: The output signal includes the 4 and 6m amateur-radio bands; no filter is necessary; you can shift from upper to lower sideband operation by reversing two pairs of connections rather than changing a filter; and one IC provides the required tank oscillator, two modulators, and summing amplifier.

No filtering is necessary to suppress the carrier and sideband frequencies because frequency cancellation is inherent in modulation. Suppose, for example, ignoring signal magnitudes, that the carrier is sinlowercase omega_Ct and the modulation signal is sinlowercase omega_Mt. Adding 90° of phase shift to either quantity produces the cosine:

and

Modulation (mixing) means multiplying the carrier and modulation signals as follows:

If you shift each of these inputs (sinlowercase omega_Mt and sinlowercase omega_Ct) by 90° and then multiply them in a separate modulator,

Lower sideband, which appears as cos(lowercase omega_M­lowercase omega_C)t at the output of IC₁. Subtracting the outputs gives the upper sideband (cos(lowercase omega_M+lowercase omega_C)t).

Selecting the RC-phasing network for this design was based on simplicity rather than low component count. Using 5% components, the network produces a response of 300 to 3500 Hz with less than 1° of phase-shift error and less than 0.2 dB of magnitude error. IC₁'s suppression of unwanted carrier and sideband frequencies (­35 dB) is approximately 5 dB less than that expected from commercial equipment but is reasonable for levels of output power less than 5W. This performance depends somewhat on the presence of capacitive terminations (C₇ and C₈) for the unused I and Q modulator inputs. Figure 1 does not show the output stage, but you can use a single-transistor buffer, a class-C power amplifier, or whatever the application requires.

For simplicity, Figure 1 shows the circuit operating with IC₁'s internal free-running oscillator. If this arrangement has insufficient stability, you can either use this oscillator as part of a PLL or use an external source. Using an external source greatly extends the transmit-frequency range.

Measurements of the circuit operating with an oscillator frequency of 142 MHz and a carrier of 71 MHz indicate carrier suppression of ­27 dB, which is 8 dB short of the typical data-sheet performance specification. However, you can improve this feature. (This design uses only the single-ended mixer inputs and outputs.) Unwanted sideband suppression was at least ­36 dB, which was the noise floor of the test site. (DI #2029)

| EDN Access | Feedback | Table of Contents |