In my previous blog, What Defines Analog Integration & Why We Should Applaud Its Benefits, I mention that the RF-DAC partition is really a form of analog integration. Why? Here's a little background.
The RF-DAC solves a difficult signal conversion problem by bridging the gap between signal processing in the digital domain, and high-frequency signal conversion in the real-world analog domain. Digital signal processing (DSP) with direct digital synthesis (DDS) and digital quadrature modulation (DQM) are powerful technologies used in wireless communication systems. However DSP, DDS, and DQM operate in the digital domain and thus are pure digital functions; ultimately wireless communication systems need to transmit real world analog radio frequency signals. The RF-DAC performs this real-world conversion.
The RF-DAC harmonizes digital signal processing with high-performance analog signal conversion. The DSP, DDS, and DQM functions can be realized with an FPGA, ASIC, or ASSP like the MAX5880 QAM modulator. The DQM and DDS move conventional analog functions into the digital domain. Basically this leverages Moore's Law and capitalizes on the fact that digital scales better than analog in terms of lower power consumption, faster speed, smaller die area, and lower cost. However, these benefits can only be realized if direct signal conversion from the digital domain to analog domain is achievable. The RF-DAC is the enabling technology that makes this possible.
For example, a multi-Nyquist RF-DAC can synthesize directly to RF a wideband block of multiple carriers in the fourth Nyquist zone. Devices like the MAX5881 (a 12-bit, 4.3 Gsps DAC) can synthesis 12 34.5 MHz, 8 PSK channels (approx. 460 MHz occupied bandwidth) at 1.82 GHz carrier. Typical applications are "direct-to-home" TV satellite uplink systems.
Because of this capability, the RF-DAC is a form of analog integration when considering that the RF-DAC translates a digital RF signal to a discrete time signal (D/A conversion) and acts as a frequency up converter (RF mixer) -- in other words, analog integration. So what's the benefit of this partition?
The advantages of RF-DAC direct conversion include better carrier/sideband suppression, very accurate phase/amplitude matching, flexible frequency planning, and software defined radio capability, as well as fewer components and lower operating power.
Consider how a conventional radio transmitter is architected with an analog quadrature modulator (AQM) versus a DQM using an RF-DAC. Figure 1 compares a typical 2x2 MIMO LTE cellular base station transmitter using complex-IF (CIF) versus direct conversion with the RF-DAC.
The CIF architecture requires five analog functions plus 32 discrete L-C filter components. The RF-DAC architecture replaces that with three devices. Yes, the FPGA digital processing functions need to be higher performance in terms of digital processing capability and I/O speed. However, we know digital scales better than analog, so the cost-power tradeoff for adding more digital area yields a better solution in terms of reducing cost, power, component count, and size while delivering better RF performance.
This is why a disruptive technology like the RF-DAC, which diverges from conventional ways to solve a difficult problem, is a form of analog integration. Can anyone think of similar examples?
For more information on the RF-DAC, check out this app note: Implementing a Direct RF Transmitter for Wireless Communications.