|
| (Technology News, 03 Feb 2012 ) |
| Thomas Neu, Texas Instruments |
|
Power consumption is among the most important system-design parameters for designers choosing high-speed data converters. Power dissipation is critical whether in portable designs requiring longer battery life or for small products that dissipate less thermal energy. System designers traditionally power the data converter from a low-noise linear regulator, such as a low-dropout regulator, rather than a switching regulator because they worry that switching noise will feed into the output spectrum of the converter and significantly degrade ac performance.
However, newer-generation, noise-optimized switching regulators, for use in mobile phones to minimize interference with nearby low-noise and power amplifiers, allow for a change in practice. They enable high-speed data converters to be powered directly from a dc/dc converter without significantly reducing ac performance. This design instantly improves power efficiency by 20 to 50 percent.
Modern high-speed converters reduce their power consumption by approximately 50 percent over previous generations, partly by lowering the power-supply voltage from 3.3V to 1.8V. As the supply rail goes lower in a low-dropout-regulator-based design, the regulator’s dropout voltage and the available power rails become more critical for power efficiency. On the digital section of the board, many voltage rails typically service the various core and I/O voltages of FPGAs and processors. On the analog section, however, only a few “clean”options, such as 3.3V and 5V, may be available.
For a high-speed data converter, 3.3V supply can be generated using a linear regulator from a common 5V rail. This 1.7V drop in the low-dropout regulator equates to a power loss of approximately 35 percent. When using a low-dropout regulator to derive the 1.8V supply of an ADC, such as the ADS4149 (Reference 1), from a 3.3V bus, the power loss in the linear regulator increases to approximately 45 percent, meaning that almost half the power dissipates in the low-dropout regulator. This example illustrates how easily inefficient power design can lose the 50 percent power reduction. The efficiency of a switching regulator is fairly independent of the input-supply rail and, therefore, offers significant power savings. With careful design, the effect on ac performance can be minimized.
Power-supply filtering A key component in isolating the switching noise from the ADC is the power-supply filter, which comprises a ferrite bead and the bypass capacitors. Several critical characteristics should be considered when choosing a ferrite bead. First, the ferrite bead must have sufficient current rating for the data converter, and must have a low DCR (direct-current resistance) to minimize the voltage drop across the bead itself. For example, a supply current of 200mA through a bead with a DCR of 1Ω leads to a 200mV drop in supply voltage. This drop may push the voltage at the ADC close to the edge or even below recommended operating conditions when standard supply voltage variations are factored in.
Second, the ferrite bead must have high impedance at the switching frequency and harmonics of the dc/dc converter to block the switching noise and spurs. Most available ferrite beads have an impedance of 100MHz, whereas the switching frequencies of modern dc/dc converters typically are 500kHz to 6MHz. For example, the ADS4149 evaluation module uses a TPS625290 switching regulator with a frequency of 2.25MHz (Reference 2). Because dc/dc regulators have a square-wave output, higher-order harmonics should also be considered. The NFM31PC276B0J3 EMI filter from Murata gives high impedance and low DCR in that frequency range.
Figure 1 compares the insertion loss of a traditional ferrite bead with a resistance of 68Ω at 100MHz with the Murata EMI filter. Power-supply circuits have low impedance, and the insertion loss is measured in a 50Ω environment. Hence, the insertion-loss magnitude of the power-supply filter may differ slightly, although the resonant frequencies do not change.

 The other components of the power-supply filter are the bypass capacitors. Choose the values of these capacitors so that their resonant frequencies, which create a low-impedance path to ground, are close to the switching frequency. Thus, switching noise passing through the bead is shorted to ground. The insertion-loss comparison of the power-supply filter in Figure 2 shows that proper bypass capacitor values create a resonance close to the switching frequency, even when this is combined with a traditional ferrite bead, such as the EXCML32A680. However, at low frequencies, it does not differ much if it is replaced with a 0Ω resistor. On the other hand, the Murata EMI filter provides approximately 20dB extra attenuation around the switching frequency. The power-supply filter in Figure 3 uses a 33μF tantalum capacitor for broad frequency decoupling, and the 10-, 2.2-, and 0.1μF ceramic capacitors have a narrower resonance frequency.

AC performance Depending on the PSRR of the data converter, a certain amount of noise on the power rail still makes it into the ADC and degrades its ac performance. The SNR and SFDR (spurious-free-dynamic-range) sweeps in Figure 4 compare a benchmark supply, such as a 1.8V, clean lab supply, with a low-dropout regulator and a dc/dc converter with different power-supply-filter options using the ADS4149 evaluation module.
Test results show SNR-performance degradation of approximately 0.3dB when powered by a switching regulator compared with a low-noise low-dropout regulator at a 300MHz intermediate frequency. The SFDR performance is also nearly identical between the setups. A closer look at the normalized FFT plot, which starts at the input signal and plots noise versus offset frequency, shows a slightly elevated noise floor across the Nyquist zone when using the suboptimal EXC ferrite bead but no evidence of any feedthrough of the switching frequency (Figure 5).

Power efficiency The main advantage of using a dc/dc converter instead of a linear regulator is power savings. In all of the experiments on the ADS4149 evaluation module, an external 3.3V supply, a common analog supply rail, powers both the low-dropout and the switching regulators. Table 1 illustrates the measured power efficiencies and their respective quiescent currents. This comparison shows that the low-dropout regulator consumes almost as much power as does the ADC. The switching regulator dissipates only 32mW more than an ideal approach, achieving an efficient power design. The low-dropout regulator’s efficiency can be further improved by stepping down the input voltage—first from 3.3V to, for example, 2.5V or 2.2V—at the expense of increased system cost and size.
 Despite having more external components than the low-dropout design, the footprint of a dc/dc converter design overall may be smaller because newer dc/dc converters have higher switching frequencies that drastically reduce the inductor’s size, making it, for example, approximately 2.2μH for 2.25MHz instead of 33μH for 500kHz.
Conversely, linear regulators may require less power-supply filtering, but they also have size constraints because they typically dissipate more power. From a cost perspective, a switching regulator may be slightly more expensive due to higher component count. Still, the increased efficiency can save cost in thermal-dissipation techniques and the system power budget (References 3 and 4).
As system designers push for more power-efficient components, changing the power architecture on a high-speed-data-converter design to switching regulators can bring a large power saving. A low-power, high-speed data converter should be powered directly from a switching regulator without significantly degrading its ac performance.
References 1. “ADS4126, ADS4129, ADS4146, ADS4149 12-/14-bit, 160/250MSPS Ultralow Power ADC,” Texas Instruments, January 2011.
2. “TPS62590 1-A Step Down Converter in 2 x 2 QFN Package,”Texas Instruments, April 2011.
3. E2E Community, Texas Instruments.
4. Neu, Thomas, “Improving the Power Efficiency of High-Speed ADCs,” Texas Instruments, March 18, 2011.
Acknowledgment This article originally appeared on EDN’s sister site, Power Management Designline.
About the author Thomas Neu is a systems engineer for the high-speed-data-converter group at Texas Instruments, where he provides application support. Neu received his master’s degree in electrical engineering from Johns Hopkins University in Baltimore.
|
| |
|
|
|
|
| |
|
|
Average Rate:
No rating yet |
| |
| |
|
|
|
|
| |
|
|
| |
|
|
| 25/4/2012 |
|
| 25/4/2012 |
|
| 25/4/2012 |
|
| |
|
|
|
|
|
|
|
| |
|
| |
|
| 30/3/2012 |
|
| 22/3/2012 |
|
| 1/3/2012 |
|
| |
|
|
|
|
|