Search the Community
Showing results for tags 'mosfet diode', ' switch' or ' switching power supply'.
Abstract: A high-quality switching power supply efficiency is as high as 95%, and the switching power supply loss is mostly from switching devices (MOSFETs and diodes), so the correct measurement of switching device losses is critical for efficiency analysis. So how do we accurately measure switching losses? First, switching loss Since the switch is a non-ideal device, its working process can be divided into four states, as shown in Figure 1. "On state" means that the switch tube is in the on state; "off state" means that the switch tube is in the off state; "on process" means that the switch tube is switched from off to on state; "off process" means that the switch tube is from conduction The conversion is turned off. In general, the main energy loss is reflected in the "on process" and "off process", a small part of the energy is reflected in the "on state", and the "off state" loss is very small, almost zero, negligible. Figure 1 Switching tube four state division The actual measurement waveform is generally as shown below: Figure 2 Switch tube actual power loss test Second, the conduction process loss The energy consumed by the transistor switching circuit during the conversion process is usually very large, because the parasitic signal of the circuit prevents the device from switching immediately, and the voltage and current in this state are in an alternating state, so it is difficult to directly calculate the power consumption. The voltage and current are considered to be linear, so the loss can be roughly calculated by finding the area of the triangle, but this is not accurate enough. For digital oscilloscopes, advanced math functions are provided, so the loss of the conduction process can be calculated using the following formula. Eon represents the loss energy of the conduction process Pon represents the average power loss (active power) during the conduction process Vds and Id represent instantaneous voltage and current, respectively Ts indicates the switching period T0, t1 indicate the start time and end time of the conduction process Shutdown process loss The closing process loss is the same as the conduction process loss calculation method, except that the start and end times of the integration are different. Eoff represents the loss energy of the shutdown process Poff represents the average power loss (active power) of the shutdown process Vds and Id represent instantaneous voltage and current, respectively Ts indicates the switching period T2, t3 indicate the start time and end time of the shutdown process Third, conduction loss In the on state, the switch tube usually flows a large current, but the on-resistance of the switch tube is very small, usually in milliohms, so the energy loss in the on state is relatively small, but it cannot ignore. Using an oscilloscope to measure conduction loss is not recommended for voltage-to-current integration because the oscilloscope cannot accurately measure small voltages during turn-on. For example, when the switch is normally turned off, the voltage is 500V, and when it is turned on, it is 100mV. Suppose the accuracy of the oscilloscope is ±1‰ (this is a very bullish indicator), and the minimum measurement accuracy is ±500mV. It is impossible to accurately measure 100mV. It is even possible that the measured voltage is negative (100mV-500mV). Since the small voltage at the time of conduction cannot be accurately measured, the energy loss error calculated by the method of integrating the voltage by the current is large. On the contrary, the current is large when turned on, so it can be measured accurately, so the current and on-resistance can be used to calculate the loss, as shown in the following formula: Econd represents the loss energy of the conduction state Pcond indicates the average power loss (active power) in the on state Id represents the instantaneous current Rds(on) indicates the on-resistance of the switch, which is given in the switch, as shown in Figure 3. Ts indicates the switching period T1, t2 indicate the start time and end time of the on state Figure 3 shows the relationship between on-resistance and current Fourth, switching loss Switching loss refers to the total energy loss, which consists of conduction process loss, shutdown process loss, and conduction loss, calculated using the following formula: Five, switching loss analysis plug-in High-end oscilloscopes usually also integrate switch loss analysis plug-ins. Because the on-state voltage measurement is not accurate, the calculation formula for the on-state can be modified. There are three main types: ● UI, U and I are measured values ● I2R, I is the measured value, R is the on-resistance, and the user inputs Rds(on) ● UceI, I is the measured value, and Uce is the voltage value input by the user to compensate for the problem of voltage and voltage uncertainty. It is generally recommended to use the I2R formula. The figure below shows the switching loss test diagram of the ZDS4000 Plus. Figure 4 Switching loss test results Summary Switching loss testing is critical for device evaluation. With a professional power analysis plug-in, the power loss of the device can be evaluated quickly and efficiently, which is simpler and more convenient than manual analysis. For MOSFETs, the I2R conduction loss calculation formula is the best choice.