Transmitters
To deliver 10 W average with a signal of 10 dB peak-to-average ratio, a class B amplifier draws 40 W from the supply. Three quarters of it becomes heat.
01Transmitter architectures
A transmitter generates the modulated signal, brings it to the channel frequency and gives it power. The architectures differ in where each step is done:
| Architecture | How it works | Advantage and limit |
|---|---|---|
| Heterodyne | Modulates at a fixed intermediate frequency and then mixes up to the channel | Easy filtering at the IF; more components |
| Direct conversion | The I/Q modulator works directly at the channel frequency | Simple and integrable; sensitive to imbalances and to the amplifier "pulling" the oscillator |
| Polar | Separates amplitude and phase: modulates the oscillator phase and the amplitude through the amplifier's supply | Allows efficient nonlinear amplifiers |
| Direct digital | A fast converter generates the signal in the final band | Maximum flexibility; requires GHz converters |
The blocks have already been studied: the I/Q modulator, the synthesizer and the power amplifier. The transmitter is the system that combines them while meeting a standard.
02What the standard requires
The spectrum authority sets what a transmitter may emit. In Argentina, ENACOM (the national communications regulator) certifies equipment; worldwide, the ITU recommendations and the standards of each service define:
- The power, expressed as effective radiated power or equivalent isotropically radiated power: \( EIRP = P_t - L_{cable} + G_{antenna} \), in dBm or dBW.
- The spectral mask: how much it may emit outside the channel, at different distances from the carrier.
- The adjacent-channel leakage ratio (ACLR): the power that falls in the neighboring channel relative to that of the channel itself.
- The spurious and harmonic emissions, far from the channel.
- The modulation quality, measured as EVM.
Adjacent-channel leakage is produced mostly by the nonlinearity of the power amplifier: the third-order intermodulation between the components of the signal itself widens its spectrum.
03The peak-to-average ratio and efficiency
An FM signal has a constant envelope; a filtered QAM or an OFDM, with many subcarriers, have peaks several dB above their average power. So as not to clip them, the amplifier is sized for the peak and works most of the time far below it, where its efficiency is low. With an ideal class B amplifier whose peak output reaches saturation, in the approximation that the signal is almost always near its average level:
A transmitter that has to deliver 10 W average. According to the peak-to-average ratio of the signal, the bars show the power a class B and a class A amplifier draw from the supply, and how much of that power is heat.
04In the lab
Measure with the spectrum analyzer the output of a low-power transmitter and compare it with the mask of its standard.
Raise the power of an amplifier driven with a QPSK signal and measure how the adjacent-channel leakage grows as it approaches compression.
Calculate the EIRP of a Wi-Fi link from the equipment power, the cable losses and the antenna gain, and compare it with the regulatory limit of the band.
05Common mistakes
- Sizing the amplifier for the average power. The peaks are clipped and the spectrum widens.
- Forgetting the cable losses in the EIRP. At high frequencies they can be several dB.
- Confusing peak power with peak envelope power. The second is the average over one RF cycle at the peak.
- Measuring the mask with an inadequate RBW. The standards set the measurement bandwidth.
06Self-assessment
What EIRP does a 30 dBm transmitter have with 2 dB of cable loss and a 6 dBi antenna?
34 dBm, about 2.5 W.
What efficiency does the formula estimate for class B with 10 dB of PAPR?
\( 78.5 \% \cdot 10^{-0.5} \approx 25 \% \).
What produces adjacent-channel leakage?
The intermodulation of the power amplifier, which widens the spectrum of the signal.
What advantage does the polar architecture have?
It allows the use of nonlinear amplifiers, which are more efficient, because the amplitude is applied through the supply.
07Further reading
- Qizheng Gu. RF System Design of Transceivers for Wireless Communications. Springer, 2005. Transmitter design with its power, linearity and spur budgets.
- Steve C. Cripps. RF Power Amplifiers for Wireless Communications. 2nd ed., Artech House, 2006. Efficiency with variable-envelope signals, the Doherty and envelope tracking.
- Behzad Razavi. RF Microelectronics. 2nd ed., Prentice Hall, 2011. Integrated transmitter architectures and their problems.