Continuous Tracking Challenges DC-DC Converters
Feb 1, 2005 12:00 PM
By John Tucker, Applications Engineer, Texas Instruments, Dallas
Click here for the enhanced PDF version of this article
(Click for a PDF of this article, containing full graphics and charts.)
News & Features From Auto Electronics
Committed to improving hybrid electric cars
New Motors for Hybrid Vehicles
Battery Firms Battle for Hybrid Hegemony
Innovative Bipolar Plates for Fuel Cells
See More Headlines
Top Articles
Exploring Current Transformer Applications
Ultracapacitor Technology Powers Electronic Circuits
Buck-Converter Design Demystified
Sensorless Motor Control Simplifies Washer Drives
PET Resources
Buyer's Guide
Conferences
Engineering Jobs
Power Electronics Events
Rent Our Lists
Spotlight on Digital Power
Typically, dc-dc converters are designed for use at a fixed output voltage. Some applications, such as dual data rate (DDR) memory bus and other high-speed logic bus terminations, require the output voltage to be a fixed percentage of an external variable voltage source. These applications typically are high-performance designs requiring fast transient response and the ability to track an external voltage reference within close tolerances, as well as current source and sink capability.
Standard dc-dc converter solutions are not well equipped to meet these requirements. Most integrated solutions have an internal voltage reference that regulates the output voltage to a fixed dc value. To operate as continuous tracking regulator, the controlling device must allow for an external reference source to be used.
TI's TPS54×72 family of integrated dc-dc converters can be used to provide simple, low-cost solutions for designs when continuous tracking is required. The device features a REFINv pin that will override the internal voltage reference if it is below 1.75 V.
A common application requiring continuous voltage tracking is the termination supply, V
In this application, the circuit generating V
The TPS54672 uses integrated, low on-resistance high- and low-side FETs operating in synchronous mode to achieve high efficiencies. This high level of integration also is important in reducing parts count and circuit size as the V
For this design example, consider the schematic in Fig. 1. This circuit will derive an output voltage, V
The output filter in this circuit is composed of L1, C7 and C8. The minimum inductor value that can be used depends on the maximum allowable inductor ripple current. For a given set of operating conditions, smaller inductor values will have a larger ac ripple current. And because this ac current is all shunted to ground through the output filter capacitors, smaller inductor values will result in a relatively high output ripple voltage. However, smaller inductor values also will allow the output current to change more rapidly in response to load changes so the transient recovery will be better. The minimum inductor value is given by:
K
The important design factors for the output capacitor are dc-voltage rating, ripple-current rating and ESR. The dc-voltage and ripple-current ratings cannot be exceeded. The ESR is important because, along with the inductor current, it determines the amount of output ripple voltage. The actual value of the output capacitor is not critical, but some practical limits do exist.
Consider the relationship between the desired closed-loop crossover frequency of the design and LC corner frequency of the output filter. In general, it is desirable to keep the closed-loop crossover frequency at less than one-fifth of the switching frequency. With high switching frequencies, such as the 700-kHz frequency of this design, internal circuit limitations of the TPS54672 limit the practical maximum crossover frequency to about 70 kHz.
For this design, a 50-kHz closed-loop crossover is desired. To allow for adequate phase gain in the compensation network, the closed-loop crossover frequency should be at least three times the LC corner frequency for this design. Assuming typical placement of the compensation zeros, this limits the minimum capacitor value for the output filter to:
For L
Given that the output must not vary by more than ±40 mV, then for a 2-A step at a 1-A/µs slew rate, the use of 100-µF, low-ESR (2-mΩ) capacitors at C7 and C8 will result in a calculated voltage change of 22 mV.
In addition to the output filter, the other essential component to achieve good transient response characteristics is the compensation network in the feedback path. The high-unity-gain bandwidth (3-MHz minimum) and fast slew rate (1.4 V/µs) of the internal error amplifier permit the use of a high closed-loop crossover frequency of 50 kHz. Placement of the poles and zeros of the type-3 compensation network in this design are optimized for early phase boost below the output filter resonant frequency and a phase margin of 50 degrees at the closed-loop crossover frequency.
When responding to a 1.5-A to 4.5-A load-current step change with a slew rate of 0.1 A/µs, the peak voltage overshoot is measured at 30 mV, lower than the required ±40 mV. Larger load-current swings or extremely fast transient may require the use of more output-filter capacitance.
Load-urrent changes are not the only source of voltage tracking errors. The dc-dc converter circuit must be able to accurately track the supplied reference voltage, in this case one half of V
The output voltage deviation is less than 4.5 mV from the REFIN voltage input over the entire range. Because the source of the REFIN reference voltage is itself a power supply voltage, load-current changes on the V
Fig. 3 shows the response of the output voltage to an input step change. The V
For more information on this article, CIRLE 339 on Reader Service Card

