Table of contents
Two questions:
1. Can the conduction current of the MOS tube flow in reverse? D to S, S to D direction at random?
2. How much current can the body diode of the MOS transistor handle?
Why are there these two problems?
When we first learned about MOS tubes, we should all start with NMOS, and the direction of current is from D to S.
In the actual application circuit, NMOS will have a current from S to D, such as the following NMOS tube anti-power reverse connection circuit (just a schematic diagram, the actual circuit needs to consider some factors).
Let me explain the principle first.
1. When the power supply is connected normally
The positive pole VCC of the power supply is connected to the body diode through the post-stage load circuit, then the body diode will be turned on, so at this time the voltage of the S pole is about 0.7V (body diode conduction voltage).
At the same time, the gate G is connected to VCC, so Vgs=Vcc-0.7V>Vgsth, the NMOS tube will be turned on. After the NMOS transistor is turned on, the conduction voltage drop is basically 0, then Vgs=Vcc, and the MOS transistor remains in the on state.
In this way, the overall power path is connected, the power supply supplies power to the subsequent stage load, and the latter stage circuit works normally.
One thing that needs special attention here is that the current of the MOS tube is from S to D at this time, which is opposite to the usual D to S that we often see.
2. When the power supply is reversed (the power supply and ground are reversed)
The gate G is connected to the negative pole of the power supply, which is 0V, and the S pole is connected to the negative pole of the power supply through the load, which is 0V, so Vgs=0V, and the MOS tube is not turned on.
At the same time, D is extremely Vcc, S is extremely 0V, and the body diode is reverse-biased and does not conduct, so current cannot flow through the NMOS tube.
For the load, it means that the power supply is disconnected.
The reversed power supply will not hit the rear load, so the subsequent stage circuit will not be burned. We only need to connect the positive and negative poles of the front power supply correctly, then the subsequent stage circuit can work normally again. In this way, it is realized Anti-reverse connection function.
It needs to be said that the anti-reverse connection here does not mean that the power supply is reversed, and the post-stage circuit can still work. But the power supply is reversed, and the subsequent stage circuit will not smoke and burn out.
When I first saw this circuit before, it was actually a drum in my heart.
When the MOS tube is turned on, can the current flow in reverse? D to S, S to D does not matter?
In addition to the problem of the direction of the current, there is also the problem of the body diode of the MOS tube. How much current can this diode hold?
If you don't understand it, you will think that the current that this diode can flow is very small, because it also has a name called "parasitic diode", which is easy to be deceived by it.
The word "parasitic" can easily make people think of parasitic inductance and parasitic capacitance, and these two things are generally very small, so it is easy to mistakenly think that this parasitic diode is also weak and cannot pass a relatively large current.
FAQ
These two questions can be answered with one circuit, which is the BUCK circuit below.
You should all know that the above is a buck circuit. The lower tube is an NMOS tube. When the upper tube is disconnected and the lower tube is turned on, the current of the inductor comes from the lower tube.
That is to say, the current direction of the lower tube NMOS is from S to D, that is, it flows in the opposite direction, and this current can be very large, because the current of the inductor can be relatively large, which is related to the load.
In addition, we also know from the previous article "BUCK's Ringing Experiment and Analysis" that when the BUCK switch is switched, there will be a dead time (when the upper tube and the lower tube are not turned on). The current of the inductor cannot be interrupted, and the current of the inductor in the dead time is the body diode of the lower tube.
And because the current of the inductor depends on the load current, it can reach several amperes, so the current of the body diode of the lower tube can also be very large.
What is the maximum body diode current of the MOS tube? Need to consider when choosing a model?
Many MOS tubes do not mark this parameter, but some manufacturers have marked it, such as this NMOS tube SI9804
As seen from the above manual, the continuous current that can pass is 2.1A.
How did this come about?
I think this may be due to power consumption limitations.
If the current passing time is very short, then a larger current can be passed, and if the time is relatively long, the current flowing cannot be too large.
As can be seen from the figure above, the maximum power consumption at an ambient temperature of 25°C is 2.5W. Looking at it this way, the continuous current mentioned above is 2.1A, which should also be based on the power consumption limit.
According to a conventional silicon diode, when a current of 2.1A is passed, the conduction voltage drop is about 1V, so the power consumption is P=2.1A*1V=2.1W, which is not much different from 2.5W.
Of course, the above is just my guess, and I haven't found any more official statement.
a more detailed manual
After writing this, I found a more detailed MOS tube manual, Infineon's NMOS tube BSC059N04LS6, which describes in detail the overcurrent capability of the body diode, including continuous and instantaneous current.
This manual has convinced me of the above guess.
The following are the parameters of the body diode in the BSC059N04LS6 manual
It can be seen directly from the above table that the continuous current of the body diode can reach 38A, and the pulse current can reach 236A. At the same time, it can also be seen that the maximum conduction voltage of the diode is 1V.
You may be a little surprised, can the continuous current of this diode be as large as 38A?
Naturally, it cannot be used in practical applications. We need to pay attention to the above condition, that is, Tc=25°C, and c is case, that is, when the outer shell is kept at 25°C.
In our actual application, if no special heat dissipation measures are taken, it is definitely impossible to guarantee the temperature of the MOS case, and naturally the current of 38A cannot be continuously passed.
But it doesn't matter, we just look at the meaning of this parameter and want to know how it came from.
Let's look at the power consumption limit in the manual
It can be seen that when Tc=25°C, the power consumption limit is 38W. We know that the conduction voltage is 1V and the current limit is 38A. It happens that the power consumption limit is equal to the voltage multiplied by the current, which is too coincidental.
Therefore, the current that the body diode can pass is based on the power consumption limit.
At the same time, we see that at Ta=25°C, the power consumption limit is 3W, and this Ta is the ambient temperature, which should be closer to the actual usage (no special heat dissipation measures are used).
If this value is used to calculate, then the current that the body diode can continue to pass is about 3W/1V=3A. Of course, this is my guess, and it is not written in the manual.
At this point, at least we should know that the body diode can still pass a relatively large current.
Of course, there is another problem. The above mentioned continuous current, and there must be a problem of instantaneous current. How big can the instantaneous current be?
This problem is more important, because in normal use, we will not pass a long-lasting current to the body diode of the MOS tube. If there is such a need, wouldn't it be good if we just turn on the MOS tube, and the power consumption can be lower.
In the previous example of BUCK, the body diode will only pass current during the dead time, and this time is quite short.
Therefore, it is more worth seeing how much current can be generated at this moment.
We still look at the manual of BSC059N04LS6, because it is directly marked.
The conduction current of this tube can reach 59A, and the current that can pass within 10us is 236A, and the body diode is also 236A. The two are the same, and both are very large, which means that the instantaneous current of the body diode will not become Use the bottleneck.
Perhaps this is why we seldom pay attention to the current of the body diode of the MOS transistor, only whether the conduction current of the MOS transistor is large enough.
Content summary:
1. After the MOS is turned on, the current direction can actually flow in both directions, from d to s, or from s to d.
2. The continuous current of the body diode of the MOS transistor can be calculated according to the power consumption limit of the MOS transistor.
3. The current that can pass instantly through the body diode of the MOS transistor is equal to the current that can pass instantly after the NMOS transistor is turned on. Generally, it will not be a bottleneck
Originally written here, the article can already be over, but I still wonder if I can see the above content from the principle of the MOS tube.
The following are some of my understandings for reference.
The structure of NMOS tube
Let's take a look at the structure of the NMOS tube.
Taking NMOS as an example, as shown in the figure above, both S and D are N-type semiconductors with relatively high doping concentration, the substrate is P-type semiconductor, and the substrate and S poles are connected together.
When the Vgs voltage is greater than the threshold voltage Vth, that is, the gate is positively charged relative to the substrate, it will attract the minority carriers (electrons) in the P-type substrate to the P-type substrate, forming an inversion layer, that is, a conductive channel road.
At this time, we will see that S and D themselves are N-type semiconductors with many free electrons, and there are also many electrons between S and D, which can also conduct electricity.
That is to say, between S and D, it is connected, and there are free electrons everywhere, which can move.
Therefore, if we apply a voltage between S and D, a current will be formed, and no matter what the direction of the voltage is, as long as there is a voltage, a current will be formed, and there is no difference between the two.
In other words, current can flow in both directions, from D to S, or from S to D.
Let's look at the overcurrent capability of the body diode
Putting P and N-type semiconductors together will always form a PN junction, which is a diode. The body diode between S and D is actually formed by the drain D and the substrate, because S and the substrate are connected together, so there is an individual diode between D and S.
The reason why the MOS transistor is turned on is because the gate attracts the minority carriers (electrons) in the P-type substrate to form a conductive channel. This channel should be relatively narrow, but it can already support the current of Id ( The current when the MOS tube is turned on, each NMOS has this parameter).
Then, as a substrate with a large volume and a large area, the PN junction formed between it and the drain electrode has no problem with the naturally flowing current reaching Id (regardless of the temperature).
However, because the formed channel resistance is very low, it does not generate much heat, and the PN junction always has a conduction voltage drop, and it will generate heat when the current flows. This is a big disadvantage, so the body diode is subject to this heating problem.
So the final result is that we will see that the continuous current flowing through the body diode is limited by the power consumption of the MOS tube.
——————————————————
Original link: Can the current direction of the MOS tube be reversed? How much current can the body diode carry? _Mos tube current flow_Hardware Engineer's Blog-CSDN Blog