Hello Experts, Looking at the LTM9100 datasheet, on page 47(Figure 37), they have mentioned how to use a second N-MOSFET as a reverse conduction blocker, Could someone please explain how that second MOSFET(Q2) acts as a reverse current blocker?


Generally P-MOSFETS are ideally used as a reverse current blocker, but could someone explain how the above MOSFET works?

 

Hello Experts, Looking at the LTM9100 datasheet, on page 47(Figure 37), they have mentioned how to use a second N-MOSFET as a reverse conduction blocker, Could someone please explain how that second MOSFET(Q2) acts as a reverse current blocker?
View attachment 293502

Generally P-MOSFETS are ideally used as a reverse current blocker, but could someone explain how the above MOSFET works?
View attachment 293503

I think you have it backwards. Connect the source/body to V8 and the drain to R7. When the gate is at ground it conducts and when the gate is pulled up to the source it is off.

Like this:

The PMOS FET switches ON when the GATE is negative with respect to the SOURCE. The PMOS FET is OFF when V(G) = V(S)

 

Yes , you are correct, i just showed the intended way to use it on a general bread board as a reference to my main query, during normal operation current through R7 would be 1 Amps, if the polarity of V8 is reversed ,current drops to 0.

 

Yes , you are correct, i just showed the intended way to use it on a general bread board as a reference to my main query, during normal operation current through R7 would be 1 Amps, if the polarity of V8 is reversed ,current drops to 0.

You did ask for an explanation of how the device works -- right?
In your original drawing there would be a voltage difference between the gate and the source, even with with nothing to drive the source so the device would be ON in that configuration. Was that your intention?

 

Hi Papabravo,
Apologies for the confusion , but that question was directed to the circuit with the NMOSFET
"Could someone please explain how that second MOSFET(Q2) acts as a reverse current blocker?"

 

From that circuit fragment I'm not entirely certain what they are trying to do. Maybe you can suggest what the problem is they are trying to solve.

 

From that circuit fragment I'm not entirely certain what they are trying to do. Maybe you can suggest what the problem is they are trying to solve.

Okay, So CL is the DC bus capacitance here, which is paralleled to three half bridges, which connects to a load(motor):
Let me draw a rough sketch to illustrate this:

At any point in time , due to any reason if there is a reverse current (indicated by Irev) flowing , Q2 is supposed to block this current(and save my input), But i am not able to understand how exactly is this done?

 

Okay, So CL is the DC bus capacitance here, which is paralleled to three half bridges, which connects to a load(motor):
Let me draw a rough sketch to illustrate this:
View attachment 293506
At any point in time , due to any reason if there is a reverse current (indicated by Irev) flowing , Q2 is supposed to block this current(and save my input), But i am not able to understand how exactly is this done?

Thanks for the clarification. Because the source terminals of both Q1 and Q2 are connected to ground, the voltage on the gate will have the same effect on both transistors. Normally the gate voltage will be 10-20 volts above the source (held at ground), which will turn the MOSFETS ON. Resistor Rs should have a small value to measure both the magnitude and direction of the current. When reverse current is detected, the gate can be driven low to turn off both transistors. There is a time constant associated with this process, via the 10Ω resistor and 47 nF capacitor. This time constant is 470 nanoseconds which seems quite fast for switching MOSFET gates.

Does this answer your question?

 

Two MOSFETs of the same type (N or P channel) are often connected in anti-series like that to prevent current flowing via the body diodes.

 

Two MOSFETs of the same type (N or P channel) are often connected in anti-series like that to prevent current flowing via the body diodes.

When it comes to diodes, the ones in a MOSFET are not really stellar performers. If that is all it takes why not just use an actual diode?

 

It perhaps should be noted that a MOSFET can conduct equally well in either direction when on (looks like a resistor with the on-resistance value).
Thus, when used as a reverse blocking diode, it is configured to conduct normally in the reverse direction (the forward direction of the substrate diode), allowing it to block reverse voltage (the normal bias of the MOSFET).

 

Hi Papabravo, Alec crutschow!
I tried simulating the same and added a current source to simulate a reverse current at a predefined time, but the input still sees the reverse current.

It looks like it makes no difference if the second MOSFET is connected
One more point, even if I replace the second MOSFET with a diode, it still doesn't block the reverse current(in the simulation atleast..), Any idea why?
Here are the files if someone wants to try!
The LTM9100 can be found in the 'lib\sym\PowerProducts' folder.

 

I tried simulating the same and added a current source to simulate a reverse current

You can't use a current source.
In Spice the current source is ideal, so it will generate whatever voltage is needed to keep the curre3nt flowing.
You need to use a voltage source.