Failure analysis of fast recovery diode

Failure analysis of fast recovery diode

Damage to fast recovery diodes includes overheating failures, overcurrent failures, overvoltage failures, and failures caused by dynamic avalanches.

1. overheating failure

Overheating failure means that the power consumption generated during the operation of the fast recovery diode causes the junction temperature to increase, which exceeds the maximum junction temperature Tjm allowed by the device, resulting in thermal breakdown of the device. Overheat failure is related to the operating temperature of the device. An intrinsic temperature Tint is usually used to predict the failure mechanism within the device as the temperature increases. Tint refers to the temperature at which the carrier concentration ni( T) when heat generation causes a temperature increase is equal to the substrate doping concentration ND. When the temperature is higher than Tint, the carrier concentration increases exponentially with the temperature, and heat generation becomes the dominant factor. Tint is related to the background doping concentration,

Generally, the Tint of high-voltage devices (ND is approximately 1013cm-3) is much lower than that of low-voltage devices (ND is approximately 1014cm-3). Due to factors such as materials and processes, device Tjm is usually much smaller than Tint.

Since the actual device does not operate in a thermal equilibrium state, the relationship between the device operating mode and temperature needs to be considered. For example, the power consumption caused by surge current in the on-state, the power consumption caused by leakage current in the off-state, and the power consumption caused by high reverse voltage during the reverse recovery period will cause the operating temperature of the device to increase and cause positive feedback between temperature and current, and the device will eventually experience thermal breakdown. Therefore, thermal breakdown occurs if the heat generated power density is greater than the dissipated power density determined by the device packaging system. In order to avoid thermal failure of the device, its operating temperature is usually limited below Tjm.

Overheat failure is usually manifested as localized melting of the device. If the local temperature is too high, it occurs in point-like areas, which will also lead to cracks in the die. If the operating frequency of the fast recovery diode is very high, high frequency switching between the off-state and the on-state will generate a large power consumption, and the overheating failure profile of the device may be different at this time. As the temperature increases, first of all, the blocking ability is lost, and almost all planar terminal devices will break down at the edge. Therefore, the damage point is usually located at the edge of the device, or at least a small portion of the edge.

2. overcurrent failure

Over-current failure refers to the failure caused by high on-state power consumption generated when surge current passes during the conduction period of the fast recovery diode. During inrush current, high losses are caused by high currents and high voltage drops, resulting in an increase in temperature. The highest temperature occurs at the crimp joint, melting the metallized electrode on the surrounding surface. As shown in Figure 1, the failure location is usually within the active area and is manifested by melting of the metallization layer near the bond wire pins.

3. overvoltage failure

Overvoltage failure is mainly caused by the voltage endured by the fast recovery diode when working exceeding the rated voltage. Damage caused by overvoltage usually occurs in the junction edge termination area. For a 1.7 kV diode, the damage point caused by the overvoltage is located between the active region and the first field limiting ring in the terminal region, as shown in Figure 2a. This is due to the high electric field strength there. Judging from the failure morphology, the damage point is small, indicating that the failure point does not pass through a large current. It may be caused by the voltage applied to both ends of the device during use exceeding the rated voltage, or it may be caused by defects introduced during the device manufacturing process. As shown in Figure 2b, for a 3.3kV diode, the overvoltage caused most of the active area and part of the junction terminal area to be burned, indicating that a large current flowed after failure. Therefore, it can be considered that the damage point first appeared in the terminal area and then extended to the suture line of the active area.

4. Failure caused by dynamic avalanche

Current concentration caused by dynamic avalanche during reverse recovery of fast recovery diodes will lead to local failure of fast recovery diodes. Figure 3 shows the failure waveform and legend of the fast recovery diode caused by dynamic avalanche. It can be seen from Figure 3a that when the fast recovery diode flows through 360A of IRM(corresponding to a current density of about 400A/cm2), the reverse voltage has risen to 2 kV within 200ns, and it failed shortly thereafter. As shown in Figure 3b, the failure caused by dynamic avalanche occurs in the active area, with a small melting channel accompanied by cracks distributed at an angle of 60°, which is consistent with the damage caused by point-like stress acting <111>on the silicon wafer in the crystal orientation. This indicates that there is current concentration in a small area, and the current density and temperature are extremely high.

In addition, due to issues such as material or process uniformity (such as defects, diffusion uniformity, life control and electrode contact), failure can also be caused when a large current is passed in a small area. For example, gold-doped diodes will fail directly under high reverse peak voltages, which is caused by the non-uniform lifetime distribution caused by gold diffusion, resulting in avalanche breakdown in a small area within the device.