What are the reasons for the burning of the G pole of the MOS tube?

S3S4S5S6

What are the reasons for the burning of the G pole of the MOS tube? [Copy link]

 

This is the schematic diagram. The P36 load is a solenoid valve with a rated value of 24V/0.2A (max). The G pole is burned out. The multimeter is used to measure

There is a short circuit between GS, but no short circuit between GD and SD. I don't know what happened. The board is mass-produced, and only this circuit appears.

maychang

This is quite strange. From the electrical schematic, the maximum voltage between the gate and source of the MOS tube can only be 12V (R206 and R214 divide the 24V power supply). Not to mention the 1N4740 protection.

maychang

"The boards are mass-produced, and this is the only way to go."

How many paths are there on the board? Does the same phenomenon occur on other paths?

mingplus

The gate pin burnout phenomenon seems to have a large current flowing through it. Check the PCB layout and whether there are any wrong parts, such as R206.

S3S4S5S6

maychang posted on 2021-11-10 12:14 "The board is mass-produced, and only this one road appears at present." How many roads are there on the board? There is no such phenomenon on other roads?

There are 20 channels on the board. Currently, 200 boards have been produced. Among these 200 boards, only one channel of the board has been burned out.

S3S4S5S6

maychang posted on 2021-11-10 12:11 This is quite strange. From the electrical schematic, the maximum voltage between the gate and source of the MOS tube can only be 12V (R206 and R214 divide the 24V power supply). What's more, there is a 1N4740 fuse...

是呀，G极应该是瞬间注入了一个大电流，但是大电流从什么地方来是一个问题，另外怀疑是手碰到了静电电压超过G极最大电压将这个击穿了，但是看图上这管子烧毁的状态来讲，静电也没有那麽大的能量啊

S3S4S5S6

mingplus posted on 2021-11-10 12:57 The gate pin burnout phenomenon seems to have a large current flowing through it. Check the PCB layout and whether there are any wrong parts, such as R206

This has been checked, the resistors are correct, and the device models are all soldered correctly

maychang

S3S4S5S6 Published on 2021-11-10 14:06 Yes, the G pole should be injected with a large current instantly, but where the large current comes from is a problem. In addition, it is suspected that the hand touched the static voltage...

The hand (or more precisely, the human body) touches the gate of the MOS tube and causes damage. This usually happens during welding. If the board has been welded, static electricity on the 10 kilo-ohm resistor will usually not generate a high voltage that will damage the MOS tube.

S3S4S5S6

maychang posted on 2021-11-10 14:56 The hand (or more precisely, the human body) touches the gate of the MOS tube and causes damage. This usually happens during welding. If the board has been welded, static electricity is on the 10 kilo-ohm resistor...

Now use an electrostatic generator to apply 2KV high voltage to the G pole, 5 times per second. The key is that after applying it, there is no continuity between GS measured by the multimeter. This...

se7ens

So what if 1N4740 is broken?

You can measure other related devices to see if they are also damaged.

chunyang

It is estimated that there is a long line in the G-pole loop on your PCB, which causes the transient coupled high voltage to break down the G-pole, and the 1N4740 does not work in time. You can try to connect a low ESL capacitor in parallel between the G-pole and the ground to absorb the spike.

S3S4S5S6

chunyang posted on 2021-11-10 18:00 It is estimated that there is a long line in the G-pole loop on your PCB, which causes the transient coupled high voltage to break down the G-pole, and the 1N4740 does not work in time. You can try to connect a parallel...

Yes, the G-pole wire is indeed very long. The busbar power supply voltage of the entire board is 24V, which is directly connected to the switching power supply. The board is a four-layer board, and the order of the layers is Signal——GND——Power——Signal

24V1 is far away from the G pole line, and there is a GND layer between the 24V2 power layer and the G pole line.

"Transient coupled high voltage"

How do you think the high voltage is generated? Static electricity? Grid surges cause instantaneous fluctuations in the output of the switching power supply.

I'll take a look at the coupling path.

，

chunyang

S3S4S5S6 Published on 2021-11-11 09:00 Yes, the G-pole wire is indeed very long. The busbar power supply voltage of the entire board is 24V. This is directly connected to the switching power supply. The board is a four-layer board. The order of the layers...

As long as it is a long line, when there is strong electromagnetic interference in the surrounding area, transient high voltage may be induced. It has nothing to do with whether it is coupled on the board. The existence of coupling on the board will only further deteriorate the problem. In addition, the problem needs to be analyzed, not guessed. This is much more important than the answer!

PowerAnts

S3S4S5S6 Published on 2021-11-11 09:00 Yes, the G-pole wire is indeed very long. The busbar power supply voltage of the entire board is 24V. This is directly connected to the switching power supply. The board is a four-layer board. The order of the layers...

If the gate-source resistor and the clamping diode ground terminal are connected to the inner layer GND through vias, there is no problem. Your problem is as maychang said, you can't tear it apart! ! You must learn to use a stethoscope like a doctor. Is it difficult to hang a waveform?

tagetage

There are 20 channels on a single board. Currently, 200 boards have been produced. Among these 200 boards, only one board has one channel burned out. This is an individual case and the cause cannot be determined. It may be an external reason.

niuyite

The problem can only be solved after the fault is restored!