15 Jun 2012
The Art of Selecting Power Supplies
Robert Green, James Niemann, and Qing D. Starks of Keithley Instruments describe some of the issues about selecting the correct power supply.
Power supplies are considered among the simplest instruments because they tend to be single-function DC devices that output a controlled voltage, so most engineers believe they understand them very well.
However, there is much more to them than that, such as the multiple design topologies developed to serve a wide range of power requirements and the variety of functions developed to address an expanding range of applications.
No matter the application, the supply’s specifier must ensure the model selected covers the load’s power requirements, which aren’t always totally straightforward.
Also, depending on the application, some important subtleties must be taken into account, such as noise considerations for low power applications and noise-sensitive systems such as low power components, modules, and devices specifically used in either low voltage or low current applications.
Many low temperature physics experiments demand an absolute minimum of noise. For modules and devices that require multiple voltages and voltages with independent (floating) references, isolation between channels of a multi-channel power supply can be critical. Finally, other important test requirements might call for a special type of supply, such as a sensitive measurement of load current, the need to characterize the pulses drawn by certain types of load, or bipolar operation for device characterization. This article explores these issues to identify some key parameters for selecting the optimum supply for a specific application.
The first step is to ensure the supply can produce sufficient power for the device under test (DUT). Different types of power supplies can have very different power envelopes. One type has a highly versatile rectangular power envelope (Figure 1a), in which any current can be supplied to the load at any voltage level. A second type (Figure 1b) has multiple rectangular envelopes for multiple ranges. This type of power envelope has the advantage of permitting higher values of one parameter at the expense of the other parameter. For example, a supply with this type of envelope can output a higher level of current but only at a lower maximum voltage. Some supplies output a hyperbolic envelope (Figure 1c), which provides a more continuous transition than a multi-range power supply. Here, one parameter is inversely proportional to the other. High power output supplies tend to have either a multi-range or a hyperbolic envelope. Take the time to evaluate the power levels your application requires to ensure the supply you select can deliver the levels of voltage and current at which you’ll be testing.
Figure 1a. Rectangular power supply envelope. Any current level can be delivered at any voltage
Figure 1b. Multi-range output. This characteristic allows higher voltages at lower currents and higher currents at lower voltages.
Figure 1c. Hyperbolic output characteristic. Maximum voltage and current follow the curve.
In circuits that operate at a very low voltage or those that use or measure very low currents (such as a transducer detector), external sources of noise can create measurement problems. The external noise produced by the power supply itself can be broken into two components: normal-mode noise and common-mode noise. For these applications, linear power supplies usually provide much lower normal-mode output noise than switching power supplies.
The tradeoff is that linear power supplies have low power-conversion efficiency and can be heavy and large. Switching supplies offer the advantages of greater power-conversion efficiency and higher output power in a smaller enclosure. For noise-sensitive circuits, linear supplies typically produce just one-fifth to one-tenth the level of normal-mode noise as switching supplies and generally also have lower common-mode noise.
Common-mode noise is generated whenever voltage transients on either the primary or the secondary windings of an isolation transformer couple current across the primary-secondary barrier. Any noise current generated on the primary (secondary) must return to the primary (secondary) in order to complete the circuit. Whenever this current flows through an impedance, a noise voltage is generated, which under certain circumstances, can cause either interference with DUT performance or measurement inaccuracies.
The magnitude of the noise term is directly related to the dv/dt of the AC line and the unshielded capacitance of the power supply’s isolation transformer. Other sources of common-mode noise include voltage transients from rectifier diodes (on the secondary) turning on and off and either the 50/60Hz line voltage variation or the abrupt voltage transient common with switching power supplies’ primary circuits. If noise is a significant concern, use a linear supply whenever possible.
The level of isolation of a power supply’s output from the power line is an important indication of the supply’s quality. A power supply with high isolation minimizes noise on the supply’s output. A very good isolation impedance is greater than 1 giga-ohm in parallel with less than 1nF and shielded well enough to support less than 5 micro-amps of common-made current.
Unfortunately, not all instruments exceed or even meet these figures. Linear designs may meet the common-mode current specification but still fall short of the DC resistance and capacitance figures; switching designs may have low capacitance and higher DC isolation but excessive common-made current. In some applications, such as testing a floating power supply, the DC isolation resistance and capacitance are the more important specifications
Alternatively, a supply used to power a low voltage resistive divider or a very low current pico-ammeter may require low common-mode current, regardless of the isolation impedance.
Figure 2. Keithley’s Series 2200 line of DC programmable power supplies offers a choice of five supplies from 86W to 150W with maximum output voltages from 20V to 72V.
Page 1 of 2 | Next >
About the author
Robert Green (photo) is a senior market development manager at Keithley Instruments, Cleveland, Ohio, which is part of the Tektronix test and measurement portfolio. During his career at Keithley, Green has been involved in the definition and introduction of a wide range of instrumentation. He holds a BS in electrical engineering from Cornell University and an MS in electrical engineering from Washington University in St. Louis, Missouri. As a staff engineer at Keithley, James Niemann is responsible for designing instrumentation used in low-level measurements. He earned a Bachelor of Science degree in Electrical Engineering from the University of Akron. He has been awarded three patents for his work and has 23 years of experience in instrumentation design. Qing D. Starks is a Keithley staff applications engineer. Prior to joining Keithley in 2006, she served in engineering roles at Infineon Technologies/Qimonda and Cypress Semiconductor. She earned a BSc in electrical engineering at the University of Calgary and an MSc in electrical engineering at Stanford University.
With more than 60 years of measurement expertise, Keithley Instruments has become a world leader in advanced electrical test instruments and systems. Our customers are scientists and engineers in the worldwide electronics industry involved with advanced materials research, semiconductor device development and fabrication, and the production of end products such as portable wireless devices. The value we provide them is a combination of products for their critical measurement needs and a rich understanding of their applications to improve the quality of their products and reduce their cost of test. In 2010, Keithley Instruments joined Tektronix as part of its test and measurement portfolio.
Most popular articles in Power management
Share this page
Want more like this? Register for our newsletter