Analysis of Improved Direct Conversion Receiver Design Based on Near Zero Intermediate Frequency

Whether used to transmit voice or data, RF communication links are an essential part of modern life. The transmitter modulates the information to the radio frequency, and the radio receiver processes the receive and demodulation processes. Modern receivers typically downconvert RF signals to baseband where they are digitized and further processed. In the history of radio development, the receiver uses a superheterodyne approach, using two or more stages of downconversion, each of which usually brings the signal closer to the baseband frequency, but at the cost of increased complexity. Large, including filters that may require very expensive and cumbersome, also require multiple local oscillators, which may introduce spurious responses that are difficult to filter out, which are inherent problems with superheterodyne techniques.

Direct conversion is achieved by changing the signal directly from the RF frequency to the baseband, eliminating the need for intermediate stages. Compressing multiple stages into a single stage makes the design simpler. Direct conversion receivers are sometimes referred to as zero intermediate frequency (IF) receivers because IF is 0 Hz. The local oscillator used to downconvert the RF signal has the same frequency as the RF signal, which significantly reduces the cost of the overall design. However, this technique is not without challenges, and in particular may be susceptible to DC offset effects from various sources. The biggest problem is the DC variation due to the amplitude-changing out-of-channel blocking signal, which are difficult to remove because they are not easily distinguished from the useful signal. This is very common in systems employing TDMA and can result in a pulsed signal envelope. In adjacent receivers, this pulse envelope produces DC artifacts through a process called second-order intermodulation. The pulse rate of the signal is very noticeable at DC and is almost indistinguishable from the signal. When the useful signal is weak, unwanted signals from nearby transmitters may cover the useful signal. This problem becomes more common as many of the digital modulation schemes used today follow this short burst mode rather than longer bursts or continuous transmissions (e.g., in FDMA systems).

Near zero intermediate frequency

From private/terrestrial mobile radios to cellular systems, more and more are being dominated by digital transmissions, and receivers are exposed to pulsed signals operating in the same frequency band. As this trend increases, the DC offset problem associated with direct conversion becomes more apparent, but there are still many trade-offs that need to be weighed against digitization. With the solution provided by CML Microcircuits (CML), a very popular design can be constructed that applies direct conversion very close to the baseband frequency and uses a very low intermediate frequency (or "near zero IF") within the signal bandwidth.

For example, for a typical 12.5 kHz channel system, the IF can be in the range of 3 to 6 kHz. The higher the IF, the stronger the resistance to the above problems. For example, in a TDMA system such as a DMR with a repetition rate of 33 Hz, the signal at DC will have a spectrum with a peak at a harmonic of 33 Hz, although the amplitude of these harmonics will attenuate at higher orders, but the higher harmonics still have There is a problem with high enough power, so the IF needs to be high enough to be immune to these harmonics.

The analog-to-digital converter bandwidth required for near-zero IF conversion is slightly increased. If the I/Q output does not have a flat response, a digital filter is required for roll-off compensa-TIon. There are other challenges in the near zero IF approach, including an image response that falls into adjacent channels on one side, which can result in channel rejecTIon, for example 65 dB on one side, another The side is only 30 dB, which in turn may cause the receiver to lose some control functions. There are various solutions to this problem, the basic principle being either using some form of calibration scheme to enhance image rejection or using dynamic local oscillator control to avoid problematic channels (ie optimizing interference immunity).

Processing requires power

The problems associated with direct transform and near-zero IF are inherent in software-defined radio (SDR) architectures. A plausible solution is to use higher digital processing power for problems, but digital processing may require a lot of power, in some The application may be prohibitive. The solutions developed by CML include a range of direct conversion ICs with very advanced features that provide the best results at the lowest power consumption. One of these products is the CMX994/A/E, which uses a proprietary technology called PowerTrade to dynamically balance power consumption with required performance. Figure 1 shows a typical system level design based on the CMX994 series.

Figure 1: Example of a system-level design based on the CMX994.

All devices in the CXM994 family have a low noise amplifier at the input that inputs the signal to the downconverter section and the baseband filter (see Figure 2). The filter stage is capable of removing the out-of-channel blocking signal prior to signal amplification and then filtering further. The differential I/Q output of the CMX994/A/E is fed into the ADC and fed into the filtering and demodulation stages. This solution is a high-performance direct conversion/near-zero IF conversion chipset optimized for low-power operation while reducing design complexity.

Figure 2: Functional block diagram of the CMX994 family of devices.

The perfect solution

Since the "back end" of these systems typically runs in the digital domain, one temptation is to continue to apply processing cycles to achieve maximum performance. In fact, the processing cycle requires the cost of the corresponding power consumption, and should also consider the time and cost of development. CML has extensive experience in the development of direct transform and near-zero IF transform systems. By looking at specific aspects of the air interface and evaluating how to apply these techniques to solve problems, identifying the problem is the first step.

This involves using those interfering signals to measure the results to truly identify the problem. Of course, the ultimate goal is to be completely immune to interference. Near-zero IF is becoming a very effective method when it comes to receiver design for the first time.

By carefully designing and selecting other RF modules (such as local oscillators), the CMX994E is able to achieve comparable immunity to radio systems based on traditional superheterodyne technology while still benefiting from direct conversion or near-zero intermediate frequency conversion. Many advantages, for example, can avoid the spurious signal response and complexity issues inherent in superheterodyne technology.

The ultimate goal of a radio system is to achieve a solution that delivers all the high quality factors, including performance, cost, size, flexibility and power. Using a dedicated chipset from a professional vendor such as CML can help achieve this goal because these solutions address the problems of directly transforming the design flow.