High-precision analysis of power loss using a digital phosphor oscilloscope

rain

High-precision analysis of power loss using a digital phosphor oscilloscope [Copy link]

As the demand for switching power supplies continues to grow in many industries, measuring and analyzing the power loss of the next generation of switching power supplies is critical. This article describes how to use the TDS5000O series digital phosphor oscilloscope and TDSPWR2 power measurement software to perform power loss analysis.

The new switching power supply (SMPS, Switch Mode Power Supply) needs to provide lower voltage and high current for high data transmission speed and GHz-class processors, which adds invisible new pressure to power supply designers in terms of power efficiency, power density, reliability and cost. In order to take these requirements into account in the design, designers have adopted new structures such as synchronous rectification technology, active power factor correction and increased switching frequency. These technologies also bring some higher challenges, such as: higher power loss on switching devices and excessive EMI/EMC.

Since the power dissipated by a switching power supply determines the efficiency of the power supply and its thermal effects, determining the power loss of the switching devices and inductors/transformers is an extremely important measurement task. The challenges faced by designers in accurately measuring and analyzing the instantaneous power loss of various devices include the following aspects: the test setup required to accurately measure power loss; correcting for errors caused by conduction delays of voltage and current probes; calculating power loss graphs for non-periodic switching changes; analyzing power losses during dynamic changes in load; and calculating the core losses of inductors or transformers.

Test Setup Required for Accurate Power Loss Measurements

Figure 1 shows a simplified circuit for switching. The MOSFET in Figure 1 is not connected to the AC power supply ground or the circuit output ground, that is, it is isolated from the ground, so a simple ground reference voltage measurement cannot be performed using an oscilloscope, because if the ground lead of the probe is connected to any terminal of the MOSFET, that point will be shorted to the ground through the oscilloscope.

Figure 1 Schematic diagram of the circuit inside the switch

In this case, differential measurement is the best way to measure the voltage waveform of the MOSFET. Differential measurement determines the drain-source voltage (VDS), which is the voltage across the drain and source terminals of the MOSFET. VDS can float over a voltage that can range from tens of volts to hundreds of volts, depending on the voltage range of the power supply. There are several ways to measure VDS:

*Leaving the chassis ground of the oscilloscope floating. This is not recommended because it is extremely unsafe and dangerous to the user, the device under test, and the oscilloscope.
*Use two regular single-ended passive probes and connect their ground leads together, then use the oscilloscope's channel calculation function to make the measurement. This measurement method is called quasi-differential measurement. Although passive probes can be used in conjunction with the oscilloscope's amplifier, they lack the common-mode rejection ratio (CMRR) function that can properly block any common-mode voltage. This setup cannot accurately measure voltage, but existing probes can be used.

* Use a commercially available probe isolator to isolate the oscilloscope chassis ground. The probe's ground lead will no longer be at ground potential and can be connected directly to a test point. Probe isolators are an effective solution but are expensive, costing 2 to 5 times as much as differential probes.
* Use a true differential probe on a wideband oscilloscope. VDS can be accurately measured with a differential probe, which is the best approach.

When measuring current through a MOSFET, clamp the current probe in place and then fine-tune the measurement system. Many differential probes have built-in DC offset trimming capacitors. After turning off the device under test and allowing the oscilloscope and probe to warm up completely, the oscilloscope can be set to measure the average value of the voltage and current waveforms. The sensitivity should be set to the value that will be used for the actual measurement. In the absence of a signal, adjust the trimming capacitors to zero the average value of each waveform to 0V. This step can minimize measurement errors caused by static voltages and currents in the measurement system.

Correcting for errors due to conduction delays in voltage and current probes

Before making any power loss measurements in a switched-mode power supply, it is important to synchronize the voltage and current signals to eliminate conduction delays. This process is called “deskewing.” The traditional method is to calculate the skew between the voltage and current signals and then manually adjust the skew using the oscilloscope’s deskew range, but this is a very tedious process.

A simpler method uses an offset correction fixture and a TDS5000 Series oscilloscope. To perform the offset correction, connect the differential voltage probe and the current probe to the test points of the offset correction fixture. The offset correction fixture is stimulated by the Auxiliary output or Cal-out signal of the oscilloscope. If necessary, the offset correction fixture can also be stimulated by an external signal source.

The deskew capability of the TDSPWR2 software automatically sets up the oscilloscope and calculates the conduction delay. The deskew function then uses the oscilloscope's deskew range and automatically compensates for the skew. The test setup is now ready to make accurate measurements. Figures 2 and 3 show the current and voltage signals before and after deskew.

Figure 2 Propagation delay of voltage and current signals

Figure 3 The signal shown in Figure 2 after “automatic offset correction” using the TDSPWR2 power measurement analysis software

Calculating Power Losses on Non-Periodic Switching Signals

Measuring dynamic switching parameters is simpler if the emitter or drain is grounded. However, the differential voltage must be measured on a floating voltage. If you need to accurately characterize and measure differential switching signals, it is best to use a differential probe. We can use the Hall effect current probe to see the current through the switching device without disturbing the circuit itself. This is where the automatic offset correction function of TDSPWR2 can be used to remove the conduction delay explained earlier.
The "Switching Loss" function of the TDSPWR2 software automatically calculates the power waveform and measures the minimum, maximum and average power loss of the switching device based on the captured data. This data is very useful when analyzing the power dissipation of the switching device. This data will be displayed as Turn on Loss, Turn off Loss, and Power Loss. This data is very useful when analyzing the power dissipation of the switching device. If you know the power loss when turning on and off, you can start to address the voltage and current transitions to reduce power loss.

During load changes, the SMPS control loop will vary the switching frequency to drive the output load. Note that when the load changes, the power loss of the switching device also changes. The resulting power waveform will be non-periodic. Analyzing non-periodic power waveforms is a very tedious task, but the advanced measurement capabilities of the TDSPWR2 can automatically calculate the minimum power loss, maximum power loss, and average power loss to provide relevant information about the switching device.

Power loss analysis during dynamic load changes

In actual operating environments, the power supply has continuous dynamic load changes. Therefore, it is important to capture the entire load change event and characterize the switching losses to ensure that the power supply is not overloaded by these.

Today, most designers use oscilloscopes with deep memory (2MB) and high sampling rates to capture events at the required resolution. However, the challenge is how to analyze the large amount of data generated by each switching loss point, as it puts a lot of stress on the switching device.

TDSPWR2's HiPowerFinder feature eliminates the challenges of analyzing deep memory data. Simply select a point of interest within the range and HiPowerFinder will find that point within the deep memory data. Once the point is found, TDSPWR2 can be used to zoom in around the cursor position to observe its activity in detail. This feature, combined with the switching loss measurement function mentioned earlier, allows users to quickly and effectively analyze the power dissipation of switching devices.

Calculating Power Losses in Electromagnetic Components

Another way to reduce power losses is related to the magnetic core. Looking at typical AC/DC and DC/DC circuit diagrams, inductors and transformers are other components that dissipate power, thus affecting not only power efficiency but also heat dissipation.

Inductors are usually tested using LCR, which uses a sine wave as a test signal. In a switching power supply, the inductor will be loaded with a high-voltage, high-current switching signal, but it is not a sine signal. Therefore, power supply designers need to monitor the behavior characteristics of the inductor or transformer in the actual power supply. Therefore, the test using LCR cannot reflect the actual situation.

The most effective way to observe the core characteristics is through the BH curve, because the BH curve can quickly reveal the behavioral characteristics of the inductor in the power supply. In the past, if you need to view and analyze the BH characteristics, designers must first capture the signal and then perform further analysis on a personal PC. Now, users can perform BH analysis directly on the oscilloscope through TDSPWR2 to observe the inductor behavior characteristics instantly. When doing in-depth analysis, TDSPWR2 can also provide a cursor link between the BH diagram and the captured data on the oscilloscope.

The BH analysis capability of the TDSPWR2 also automates the measurement of power losses in a real SMPS environment. If the core losses of an inductor or transformer need to be inferred, the power loss measurements can be made on the primary or secondary core. The difference between these results is the power loss in the core (core loss). These measurements reveal information about the area where the power is dissipated.