How to completely shut down Pmos

程秋香

How to completely shut down Pmos [Copy link]

 

I have been working on a Pmos switching circuit recently, and I want to use a single-chip microcomputer to generate a PWM wave to control the PMOS on and off. The circuit schematic is shown in Figure 1. I first use a DC power supply to output 3.3v to control its ON/OFF to see if the PMOS can be turned on and off normally, but I found that:

1: When the G pole of Q2 (NMOS) is not powered, the voltage at point U1 to ground (Uds) is 7.8v, the voltage across R2 is 1.23v, the voltage across load R4 is 4v, and Q1 (Pmos) is in the amplification area

2: When the G pole of Q2 (NMOS) is powered by 3.3v, the voltage at point U1 to ground is 0v, the voltage across R2 is 9v, the voltage across load R4 is 9v, and Q1 (Pmos) is turned on. I would like to ask what is the cause of this problem. Is it because when the NMOS is turned off, Rds is not infinite? I changed R2 from 10K to 1k and 100k, but it didn't work. I suspect it is the PMOS leakage current?

qwqwqw2088

The internal resistance of IRLL014N is 140 milliohms. Is it good or bad? The internal resistance of MOS tube is certain, and the shutdown is infinite. I don't know how the OP understands that MOS tubes are generally used, and most of them work in the switching state.

chunyang

The original poster should measure the gate voltage of Q2. It should be 0 when the gate of Q2 is not powered. Otherwise, check R3. Only when Q2 is in a slightly on state, the gate voltage of Q1 will be pulled down and turned on. Note that the input impedance of the MOS tube is very high, and a little energy induced in space may turn on the tube, especially when there is a long connection in the gate loop. The value of the static bias resistor R3 must be able to make the gate voltage low enough when there is no drive to ensure that Q2 is cut off.

程秋香

chunyang posted on 2019-2-19 16:39 The OP should measure the gate voltage of Q2, which should be 0 when the gate of Q2 is not powered, otherwise check R3. Only when Q2 is in a slightly on state, the gate voltage of Q1 will be pulled down from...

I have confirmed that the gate voltage of Q2 is 0 when there is no power supply. I would like to ask if there will be leakage current in the GS of PMOS?

程秋香

qwqwqw2088 Published on 2019-2-19 16:26 The internal resistance of IRLL014N is 140 milliohms. Is it good? The internal resistance of MOS tube is certain, and the shutdown is infinite. I don’t know how the OP understands MOS tube...

Then will there be leakage current in the GS of PMOS?

qwqwqw2088

Cheng Qiuxiang posted on 2019-2-20 08:18 Then will there be leakage current in the GS of PMOS?

Regardless of PMOS or NPOS, tubes from different manufacturers will have different degrees of leakage current, but leakage current generally refers to the leakage current between D and S. When the GS voltage remains unchanged, it changes with temperature. Generally, as the temperature rises, the leakage between source and drain will increase.

程秋香

qwqwqw2088 Published on 2019-2-20 08:48 Regardless of PMOS or NPOS, tubes from different manufacturers will have different degrees of leakage current, but leakage current generally refers to the leakage current between D and S. When GS is not pressed...

Then when the power is not on, NMOS is in an open circuit, why is the gs voltage of PMOS not equal to 0?

qwqwqw2088

Cheng Qiuxiang published on 2019-2-20 10:12 So when the power is off, the NMOS is in an open circuit, why is the gs voltage of the PMOS not equal to 0?

Please look at the manual of the MOS tube, the volt-ampere characteristic line diagram ID---Uds, in different UGS line clusters

alan000345

Cheng Qiuxiang published on 2019-2-20 10:12 So when the power is not on, the NMOS is in an open circuit, why is the gs voltage of the PMOS not equal to 0?

Is it because the built-in diode causes the voltage value of the PMOS to be non-zero? To be honest, I rarely design the MOS circuit part myself. The company used other people's mature circuits. When our tester tested other people's protection boards, during the overcurrent test, we encountered some boards from different manufacturers with diodes in the MOS tubes. When the test was the second time, they were burned, and those without diodes burned at the first time.

chunyang

Cheng Qiuxiang posted on 2019-2-20 08:17 I have confirmed that the gate voltage of Q2 is 0 when no power is supplied. I would like to ask if there is leakage current in the GS of PMOS?

Of course there is leakage current, but it is extremely low, especially the leakage current through the gate. As long as the tube is intact, its leakage current cannot be measured by a conventional multimeter. First, find out your load impedance.

chunyang

alan000345 Published on 2019-2-20 10:57 Is it because the built-in diode causes the voltage value of PMOS to be non-zero? To be honest, I seldom design the MOS circuit part by myself. The company used...

Unless your load impedance is extremely high and the measuring instrument also has extremely high internal resistance, it is impossible.

alan000345

chunyang posted on 2019-2-21 17:57 Unless your load impedance is extremely high and the instrument used for measurement also has an extremely high internal resistance, it is impossible.

The internal resistance of the measuring instrument must be high enough. The problem is that the protection board is being tested. Because the protection board is connected to the lithium battery, its internal resistance is very low. When making a lithium battery protection board, many MOS tubes need to be used in parallel to reduce the resistance.

chunyang

alan000345 posted on 2019-2-22 11:46 The internal resistance of the measuring instrument must be high enough. The problem is that the protection board is tested. Because the protection board is connected to the lithium battery, its internal resistance is very low. The lithium...

Load impedance is different from the internal resistance of the protection board you mentioned. The load here refers to the load between the MOS tube and the ground during your actual test, which corresponds to R4 in the figure.

youganw

It should be that the Q2 gate does not reach zero when it is not powered.

alan000345

chunyang posted on 2019-2-22 16:01 Load impedance is different from the internal resistance of the protection board you mentioned. The load here refers to the load between the MOS tube and the ground during your actual test, which corresponds to the...

For the battery, the protection board needs to have a small internal resistance. For our tester, the internal resistance of the protection board is the load impedance, and it also needs to simulate the internal resistance of the battery. To clarify, I hope you really understand the original poster's question and answer it.

chunyang

alan000345 posted on 2019-2-25 13:33 For the battery, the protection board needs to have a small internal resistance. For our tester, the internal resistance of the protection board is the load impedance, and it also...

The concept must be made clear, otherwise it is inevitable to go in the wrong direction. The protection board is only part of the load, and the load also includes the battery, and not just the battery, there should be other circuits, such as charging circuits. For MOS tubes, the synthetic impedance should be examined. When testing, it is either a complete target circuit or a simulated load, and it cannot be taken for granted, especially under the wrong understanding of the concept.

alan000345

chunyang posted on 2019-2-26 16:17 The concept must be made clear, otherwise it is inevitable to go in the wrong direction. The protection board is only a part of the load, and the load also includes the battery, and not just the battery, it should also...

You didn't understand what I said, I mean that Q1 in the original poster's circuit has a diode, which can play the role of the diode. Our test instrument has defects, so Q1, which has a built-in diode, is not easily damaged, because this diode may cause leakage voltage. You actually pulled it to the load. You should ask me if I know resistors or capacitors, and tell me the principles of resistors and capacitors. So far, I don't have time to waste with you.

chunyang

alan000345 posted on 2019-2-27 08:46 You didn't understand what I said. I meant that Q1 in the OP's circuit has a diode, which can play the role of a diode. Our test instrument has defects...

It is necessary to investigate the size of the load and even the impedance of the measuring instrument. If it is really caused by leakage current, for weak current, there is no large enough load impedance and high enough internal resistance of the measuring instrument, it is impossible to observe obvious voltage. If the load is not large enough and the internal resistance of the measuring instrument is not large enough, then the observed voltage is not caused by leakage current. Ohm's law, we all learned it in junior high school, but whether we can use it is another matter. Since you spoke rudely, you are the one who suffers. I don't care about people like you at all. I can even say that the more rookies with bad attitudes and no skills, the better it is for those who work hard. Without comparison, there is no gap. Without a gap, how can we reward the diligent and punish the lazy?

alan000345

chunyang posted on 2019-2-27 16:51 It is necessary to investigate the size of the load and even the impedance of the measuring instrument. If it is really caused by leakage current, for weak current, there is no large enough load impedance to...

Let's not talk about technical matters first. As a moderator, you cursed on the forum just because others did not agree with your point of view. There are many people reading the forum every day, and the insiders know it very well. Whatever you say, I don't want to say another word to you, not for anything else, just because you are too poor.

chunyang

alan000345 posted on 2019-2-28 12:35 Let’s not talk about the technical issues first. As a moderator, you cursed on the forum just because others did not agree with your point of view. There are so many...

Come on, whoever claims, please provide evidence! Since you said "because others did not agree with your point of view, you cursed on the forum", provide evidence! At the same time, I also declare again: I have always been a respectful person. If someone speaks rudely first, I have the right to fight back, and I have always done so. If you think this is "poor quality", it doesn't matter. I never need your approval and don't care at all. But if you speak rudely, I will be rude to you just the same!