Research on the Oscilloscope as a Platform for Testing Switching Power Supply Measurement

　　Power supplies come in a wide variety of types and sizes, from traditional analog power supplies to highly efficient switching power supplies. They all face complex, dynamic operating environments. Equipment loads and demands can change dramatically in an instant. Even "daily" switching power supplies must be able to withstand instantaneous peaks that far exceed their average operating levels. Engineers who design power supplies or use power supplies in systems need to understand how the power supply operates under static conditions as well as worst-case conditions.

　　In the past, characterizing the behavior of a power supply meant measuring quiescent current and voltage with a digital multimeter and performing painstaking calculations with a calculator or PC. Today, most engineers have turned to oscilloscopes as their preferred power measurement platform. Modern oscilloscopes can be equipped with integrated power measurement and analysis software, simplifying setup and making dynamic measurements easier. Users can customize key parameters, automate calculations, and see results in seconds, not just raw data.

Power Supply Design Issues and Measurement Needs

　　Ideally, every power supply should work just like the mathematical model it was designed for. But in the real world, components are imperfect, loads vary, power supplies can be distorted, and environmental changes can alter performance. And changing performance and cost requirements make power supply design even more complex. Consider these questions:

　　How many watts can the power supply sustain beyond its rated power? For how long? How much heat does the power supply dissipate? What happens when it overheats? How much cooling airflow does it require? What happens when the load current increases significantly? Can the device maintain the rated output voltage? How does the power supply handle a dead short at the output? What happens when the input voltage to the power supply varies?

　　Designers need to develop power supplies that take up less space, reduce heat, cut manufacturing costs, and meet stricter EMI/EMC standards. Only a rigorous measurement system can enable engineers to achieve these goals.

Oscilloscope and Power Supply Measurements

　　To those accustomed to making high-bandwidth measurements with an oscilloscope, power measurements may seem simple because of their relatively low frequencies. In reality, power measurements present many challenges that high-speed circuit designers never have to face.

　　The voltage across the switching device may be high and "floating", that is, not grounded. The pulse width, period, frequency, and duty cycle of the signal may vary. The waveform must be captured and analyzed faithfully to detect anomalies. This is demanding for the oscilloscope. Multiple probes - single-ended probes, differential probes, and current probes are required at the same time. The instrument must have a large memory to provide recording space for long-term low-frequency acquisition results. And it may be required to capture different signals with greatly different amplitudes in a single acquisition.

Switching Power Supply Basics

　　The dominant DC power architecture in most modern systems is the Switch Mode Power Supply (SMPS), which is well known for its ability to efficiently handle varying loads. The power signal path of a typical SMPS includes passives, actives, and magnetics. SMPS minimizes the use of lossy components (such as resistors and linear transistors) and instead uses mostly (ideally) lossless components: switching transistors, capacitors, and magnetics.

　　The SMPS device also has a control section that includes components such as the pulse width modulation regulator, the pulse frequency modulation regulator, and the feedback loop 1. The control section may have its own power supply. Figure 1 is a simplified SMPS schematic showing the power conversion section, including active devices, passive devices, and magnetic components.

　　SMPS technology uses power semiconductor switching devices such as metal oxide field effect transistors (MOSFET) and insulated gate bipolar transistors (IGBT). These devices have fast switching times and can withstand unstable voltage spikes. Equally important, they consume very little energy in both the on and off states, with high efficiency and low heat generation. The switching devices largely determine the overall performance of the SMPS. The main measurements of the switching devices include: switching loss, average power loss, safe operating area, and others.

Figure 1. Simplified schematic of a switching power supply.

Figure 2. MOSFET switching device showing measurement points.

Preparing for Power Measurements

　　When preparing to make measurements on switching power supplies, it is important to select the right tools and set them up so that they work accurately and repeatably. Of course, the oscilloscope must have the basic bandwidth and sampling rate to accommodate the switching frequency of the SMPS. Power supply measurements require at least two channels, one for voltage and one for current. Equally important are some facilities that can make power supply measurements easier and more reliable. Here are some things to consider:

　　Can the instrument handle both the turn-on and turn-off voltages of a switching device in the same acquisition? The ratio of these signals can be as high as 100,000:1.

　　Are there reliable, accurate voltage and current probes? Is there an effective way to correct for their different delays?

　　Is there any effective way to minimize the static noise of the probe?

　　Does the instrument have sufficient record length to capture a long, complete line frequency waveform at a high sampling rate?

　　These characteristics are fundamental to making meaningful and effective power supply design measurements.

Measure 100 V and 100 mV in one acquisition

　　To measure the switching losses and average power loss of a switching device, the oscilloscope must first determine the voltage across the switching device when it is turned off and on, respectively.

　　In AC/DC converters, the voltage dynamic range across the switching device is very large. The voltage across the switch device in the on state depends on the type of switch device. In the MOSFET tube shown in Figure 2, the turn-on voltage is the product of the on-resistance and the current. In bipolar junction transistor (BJT) and IGBT devices, this voltage is mainly determined by the saturation conduction voltage (VCEsat). The voltage in the off state depends on the operating input voltage and the topology of the switching converter. Typical DC power supplies designed for computing equipment use a universal mains voltage between 80 Vrms and 264 Vrms.

　　The off-state voltage across the switching device (between TP1 and TP2) can be as high as 750 V at full input voltage. In the on-state, the voltage between the same terminals can be between a few millivolts to approximately 1 V. Figure 3 shows the typical signal characteristics of a switching device.

Figure 3. Typical signals of a switching device.

　　To accurately measure the power supply of a switching device, the off and on voltages must first be measured. However, the dynamic range of a typical 8-bit digital oscilloscope is not sufficient to accurately acquire both the millivolt-level signal during the on period and the high voltage that occurs during the off period in the same acquisition cycle. To capture this signal, the vertical range of the oscilloscope should be set to 100 volts per division. At this setting, the oscilloscope can accept voltages up to 1000 V, so a 700 V signal can be acquired without overloading the oscilloscope. The problem with using this setting is that the maximum sensitivity (the smallest signal amplitude that can be resolved) becomes 1000/256, which is about 4 V.

　　Tektronix DPOPWR software solves this problem by allowing the user to enter the RDSON or VCEsat value from the device data sheet into the measurement menu shown in Figure 4. If the voltage being measured is within the sensitivity range of the oscilloscope, DPOPWR can also use the acquired data for calculations rather than the manually entered value.

Figure 4. The DPOPWR input page allows the user to enter specification values ​​for RDSON and VCEsat.

Figure 4. Propagation delay affects power supply measurements.

Eliminate timing deskew between voltage and current probes

　　To make power measurements with a digital oscilloscope, you must measure the voltage and current between the drain and source of a MOSFET switching device (see Figure 2), or the voltage between the collector and emitter of an IGBT. This task requires two different probes: a high-voltage differential probe and a current probe. The latter is usually a non-intrusive Hall-effect type probe. Each of these two probes has its own unique propagation delay. The difference between these two delays, called time skew, can cause inaccuracies in amplitude measurements and time-related measurements. It is important to understand the impact of probe propagation delay on maximum peak power and area measurements. After all, power is the product of voltage and current. If two variables that are multiplied are not well calibrated, the result will be wrong. When the probe is not properly "time-skewed", the accuracy of measurements such as switching loss will be affected.

　　The test setup shown in Figure 5 compares the signal at the probe tip (lower trace display) and the signal at the oscilloscope front panel after a propagation delay (upper display).

　　Figures 6 - 9 are actual oscilloscope screen shots showing the effects of probe skew. It uses a Tektronix P5205 1.3 kV differential probe and a TCP0030 AC/DC current probe connected to the DUT. The voltage and current signals are provided through a calibration fixture. Figure 6 illustrates the skew between the voltage and current probes, and Figure 7 shows the measurement result (6.059 mW) obtained without correcting the skew between the two probes. Figure 8 shows the effect of correcting the probe skew. The two reference traces are overlaid, indicating that the delay has been compensated. The measurement results in Figure 9 show the importance of correcting for skew. This example shows that the skew introduces a 6% measurement error. Correcting for skew accurately reduces the peak-to-peak power loss measurement error.