This article interprets the four fundamental characteristics of RF circuits from four perspectives: the RF interface, small desired signals, large interfering signals, and adjacent‑channel interference. It also highlights important factors that require special attention during PCB design.
The RF Interface in RF Circuits
Wireless transmitters and receivers can be conceptually divided into two parts: baseband and RF. The baseband includes the frequency range of the input signal to the transmitter, as well as the frequency range of the output signal from the receiver.
The baseband bandwidth determines the basic rate at which data can flow through the system. Baseband processing is used to improve the reliability of the data stream and to reduce the burden imposed by the transmitter on the transmission medium at a given data rate.
Therefore, designing baseband circuits on a PCB requires a great deal of knowledge in signal processing engineering. The RF circuitry of a transmitter converts and upconverts the processed baseband signal to the designated channel and injects it into the transmission medium.
Conversely, the RF circuitry of a receiver extracts the signal from the transmission medium and downconverts it to baseband.
Transmitters have two primary PCB design goals:
They must deliver a specified output power while consuming as little power as possible.
They must not interfere with the normal operation of transceivers in adjacent channels.
For receivers, there are three main PCB design goals:
First, they must accurately recover small signals.
Second, they must reject interfering signals outside the desired channel.
Finally, like transmitters, they must consume very little power.
Large Interfering Signals in RF Circuits
The receiver must remain sensitive to small signals even in the presence of large interfering signals (blockers). This situation occurs when trying to receive a weak or distant transmission while a nearby powerful transmitter is broadcasting in an adjacent channel.
The interferer may be 60‑70 dB stronger than the desired signal and can block reception either by overwhelming the receiver's input stage or by causing excessive noise generation in that stage.
Both problems arise if the input stage is driven into nonlinearity by the interferer. To avoid these issues, the front‑end of the receiver must be highly linear.
Therefore, linearity is also an important consideration in receiver PCB design. Since the receiver is a narrowband circuit, nonlinearity is characterised by measuring intermodulation distortion. This involves driving the input with two sinusoidal signals of similar frequency, both within the band, and measuring the intermodulation products.
In general, SPICE is a time‑consuming and expensive simulation tool for this purpose, because it must perform many cycles to achieve the frequency resolution needed to understand the distortion.
Small Input Signals in RF Circuits
A receiver must be sensitive enough to detect small input signals. In general, the input power to a receiver can be as low as 1 μV. The sensitivity of the receiver is limited by the noise generated by its input circuitry.
Therefore, noise is an important consideration in receiver PCB design, and the ability to predict noise using simulation tools is indispensable.
In a typical superheterodyne receiver, the received signal is first filtered and then amplified by a low‑noise amplifier (LNA). It is then mixed with the first local oscillator (LO) signal to convert it to an intermediate frequency (IF).
The noise performance of the front‑end circuitry is mainly determined by the LNA, the mixer, and the LO. Although traditional SPICE noise analysis can find the noise of the LNA, it is useless for the mixer and LO because the noise in these blocks is severely affected by the large LO signal.
The small input signal requires the receiver to have a very high gain, often as high as 120 dB. With such high gain, any signal coupling from the output back to the input can cause problems.
An important reason for using the superheterodyne architecture is that it distributes the gain across several frequency stages, reducing the likelihood of coupling. It also keeps the first LO frequency different from the input signal frequency, preventing large interfering signals from "contaminating" the small input signal.
For various reasons, some wireless communication systems use direct‑conversion or homodyne architectures instead of superheterodyne. In this architecture, the RF input signal is directly converted to baseband in a single step, so most of the gain is in the baseband, and the LO is at the same frequency as the input signal.
In such cases, the impact of even small coupling must be understood, and detailed models of stray signal paths – such as coupling through the substrate, coupling between package pins and bondwires, and coupling through power lines – must be established.
Adjacent‑Channel Interference in RF Circuits
Distortion also plays an important role in transmitters. Nonlinearity in the output circuitry of a transmitter can spread the bandwidth of the transmitted signal into adjacent channels – a phenomenon known as spectral regrowth.
Before the signal reaches the power amplifier (PA), its bandwidth is constrained; but intermodulation distortion within the PA causes the bandwidth to increase again.
If the bandwidth expands too much, the transmitter will fail to meet adjacent‑channel power requirements. When transmitting digitally modulated signals, it is virtually impossible to predict spectral regrowth using SPICE.
This is because approximately 1000 digital symbols must be simulated to obtain a representative spectrum, and the high‑frequency carrier must also be included – making SPICE transient analysis impractical.
About Maxipcb