Power devices have been very popular in the market in recent years, especially the #MOS tube, which is mainly used in power adapters, battery management systems, inverters and motor control systems.

With the explosion of computer motherboards, AI graphics cards, servers and other industries, low-voltage power MOS tubes will once again usher in explosive market demand.

In the field of switching power supply application, since the Controller of the power supply has been very perfect, and most of the controllers are pure hardware control, manufacturers generally optimize the layout and the driver of MOS, so the MOS burn out situation in the application of switching power supply is relatively small, most of the performance is overheating.

In the battery management system, motor control system and inverter system, the burnout probability of the MOS tube becomes very large, the reason is that the protection of the battery management system instantaneous current mutation, the MOS load in the motor and inverter system is a very large inductive load, especially the motor control is also faced with the reverse electromotive force brought by braking. The operating voltage and current of MOS tube are more challenging.

Today we analyze the six most common failure modes of MOS tubes

1. Avalanche failure

Avalanche failure refers to overvoltage breakdown, that is, we often say that the voltage between the drain and the source exceeds the rated voltage of the MOSFET, and reaches the limit of MOSFET tolerance, resulting in the failure of the MOSFET.

2. SOA failure

SOA failure refers to overcurrent damage, that is, the failure caused by the current exceeding the safe working area of the MOSFET, generally because the Id exceeds the maximum value determined by the device specification, making the MOSFET heat loss is too large, and the failure caused by long-term heat accumulation.

3. Electrostatic failure

Electrostatic failure is better understood, and almost any electronic component faces electrostatic problems, especially in the dry winter in the north. You know, the general electrostatic resistance of the MOS tube is 500V, which is very fragile, so we still try to use anti-static bracelets and tweezers when operating the MOS tube in winter.

4, grid breakdown

Gate breakdown means that the gate is subject to abnormal voltage, resulting in the failure of the gate oxide layer. Generally, the Vgs of the MOS tube is set at 12V. Although the device manual indicates that the Vth is generally 2-5V, different Vgs will correspond to different Rdson. Therefore, we usually choose 12V or 15V to ensure that the MOSFET is fully turned on. And this voltage can not be as high as the MOS Vds voltage resistance, Vgs is generally limited to 20V, more than 20V will likely breakdown the gate.

After the gate breakdown, the general use of a multimeter can be measured that the short circuit between GS, and the normal high resistance state between DS.

5, resonance failure

Whether it is a battery management system, or the field of inverter and motor control, we usually use the multi-parallel design of MOS, due to the inconsistency of MOSFET itself parameters will lead to different gate and circuit parasitic parameters of each MOSFET, when switching together, due to the order of opening caused by switching oscillation, Further damage MOSFET, so when used in parallel must pay attention to the layout and wiring, as well as the Vth selection of MOS and supply chain management, which I will devote another article to discuss.

6, body diode failure

In motor control, bridge rectifier and LLC and other control systems, we need to use the MOSFET body diode for continuous flow, generally the reverse recovery time of the body diode will be relatively slow, so it is easy to overpower and lead to the failure of the body diode. Therefore, in a system with a relatively high control frequency, we need to parallel a fast recovery diode or Schottky outside the MOSFET.

Below, we will analyze its failure process and preventive measures in detail on overvoltage breakdown and overcurrent burn

Avalanche failure and its prevention

In simple terms, MOSFET on the power supply board due to the busbar voltage, transformer reflection voltage, reverse electromotive force of the motor, leakage peak voltage and so on after the high voltage overlap in the system, will be superimposed between the MOSFET drain and source.

MOSFET manuals generally include parameters such as single-side impulse avalanche Energy Eas, repeated-pulse avalanche energy Ear, and single-pulse avalanche current Ias, which reflect the avalanche capability of the power MOSFET.

‍ In fact, there is also a parasitic transistor in the actual MOSFET, just like the continuous current diode between the drain and source, you can see the following internal diagram and the corresponding equivalent circuit diagram:

We can see that this parasitic BJT is directly in parallel with the MOSFET, so when there is a large current Id and high voltage Vd in the MOSFET drain, the ionization effect inside the device is intensified, and a large number of hole currents appear. These currents flowing through the Rb resistance into the source cause the base potential of the parasitic triode to rise, that is, Vb will rise, then the parasitic triode will be on, resulting in avalanche breakdown, so its internal is due to overvoltage generated current into the parasitic triode, triode conduction, is equal to MOSFET also conduction.

Preventive measures:

Avalanche failure is ultimately a voltage failure, so we focus on voltage to prevent it. Specific can refer to the following ways to deal with.

1: Reasonable use of derating. At present, the derating in the industry generally selects 80%-95% derating, and the specific situation is selected according to the warranty terms and circuit concerns of the enterprise.

2: Reasonable transformer reflection voltage.

3: Reasonable RCD and TVS absorption circuit design.

4: The high current wiring should be as thick and short as possible to minimize the parasitic inductance of the wiring.

5: Select a reasonable grid resistance Rg.

6: In high-power power supply, RC shock absorption or Zener diode can be added as required.

SOA failure machine prevention

SOA failure refers to the abnormal large current and voltage superimposed on the MOSFET at the same time during the operation of the power supply, resulting in instantaneous local heating caused by the damage mode. Or the chip and the radiator and the package can not reach the heat balance in time leading to heat accumulation, and the continuous heating causes the temperature to exceed the oxide layer limit and causes the thermal breakdown mode.

The parameter limits for each line of SOA can be seen in the following picture, which is found in the data sheet of each @MOSFET.

Below we analyze the meaning of the five areas marked in the figure below

1. This place mainly limits the maximum rated current and pulse current, because at the moment the transverse axis display voltage is very low, so more is the SOA failure caused by high current.

2. The area of 2 is a safe area for current and voltage, but it also depends on the junction temperature of the device (depending on the Rdson size), if the junction temperature exceeds 150 degrees, it will also cause SOA failure.

3. In the No. 3 region, we can see that it has been extended three times according to different times, corresponding to 10ms, 1ms and 100us, here mainly to see the dissipated power of the device, which is essentially able to withstand the maximum current value of 10ms.

4. In zone 4, this is an area where the current value is capped, which refers to the maximum limit of the pulse current, beyond which the SOA will fail.

5. In Zone 5, this is a voltage capping zone, which mainly limits the voltage on the Vds.

The MOSFET in our circuit, as long as the device is within the range of the above limit area (2 and 3), it can effectively avoid the power failure problem caused by MOSFET.

Preventive measures:

1. Ensure that under the worst conditions, all power limits of the MOSFET are within the SOA limit line.

2. The OCP function must be precise and detailed. In the OCP point design, it is generally possible to take 1.1-1.5 times the current margin, and then start debugging the RSENSE resistance according to the IC protection voltage such as 0.7V. In addition, some MOSDriver also integrates overcurrent protection function, you can also try, is expensive.

3. Reasonable thermal design redundancy is also very necessary for reliability testing of rated current and maximum current operating time, bearing in mind the superimposed operating environment temperature.