Hello,

I am trying to switch on\off a 5V load using an N-channel MOSFET, with a microcontroller supplying the logic signals (also 5V). I soon realized, after connecting my microcontroller directly to the gate of the MOSFET, that something was very wrong .

After a little research it seems that I need a driver circuit, so I attempted to build this one:

http://incredulist.blogspot.com/2012/10/high-side-switching.html

With the collector of the PNP transistor attached to my oscilloscope, everything seemed fine—the voltage alternated between a full 12V and then down to 0V, just as it should. However, when I wire it to the gate of the MOSFET, the drain stays at 5V... A 100-ohm pull-down resistor on the gate helps a *little* (though it gets HOT), but the MOSFET drain won't drop below 2.5V.

Is my problem that the PNP transistor in the diagram can't sink enough current? (I've attached a schematic of my MOSFET circuit in case it's needed).

 

Very little gate current is needed to switch a MOSFET, so that's not your problem. What MOSFET are you using? How much current does the load require?

ETA: I looked at the attachment, and see that you are using an IRF630. I assume you don't have a logic level MOSFET available?

ETA2: Are you actually using 12V on the gate, or is that just a representative picture that you posted?

 

No, I do not have a logic level MOSFET on hand. I tested the circuit with a 3A load but it will be a 10A load when I get it working (I have attached a large heatsink to the MOSFET, so I think it should be fine).

 

Are you actually using 12V on the gate, or is that just a representative picture that you posted?

 

I am actually using 12V.

 

With the collector of the PNP transistor attached to my oscilloscope, everything seemed fine—the voltage alternated between a full 12V and then down to 0V, just as it should. However, when I wire it to the gate of the MOSFET, the drain stays at 5V... A 100-ohm pull-down resistor on the gate helps a *little* (though it gets HOT), but the MOSFET drain won't drop below 2.5V.

No, I do not have a logic level MOSFET on hand. I tested the circuit with a 3A load but it will be a 10A load when I get it working (I have attached a large heatsink to the MOSFET, so I think it should be fine).

First of all the IRF630 is rated for a maximum of 9A, so it's underpowered for your application.

Second, you don't want a 100ohm pulldown resistor on the gate; you want something more like 100k.

Third, you must have something wrong with the IRF630 wiring. The gate is on pin 1, the drain is on pin 2, and the source is on pin 3; I suggest you recheck that carefully.

 

The PNP driver you show can source current but it can't sink any to pull the gate to zero volts. A totem-pole driver would work better to drive a MOSFET gate since it can both source and sink current.

 

The mosfet that you used is not match what you need.

I'm not sure how is your frequency, so I didn't change the other values of resistors.
Before you using some other mosfets that you can try the circuit as below:

The IRL540 is more better for you need.

 

Before you using some other mosfets that you can try the circuit as below:

Thanks! The resistors on the PNP output bring the voltage down to 5V, which made the MOSFET get pretty hot (probably wasn't getting switched on fully). I tried using 20V (which drops to 9V), which made it a little better, but not fully.

On a datasheet, what do they call the voltage needed to switch the MOSFET on fully? Is it "Gate Threshold Voltage"?

Also I found a 52A "NDP6030L" MOSFET on an old motherboard that should work better.

 

Thanks! The resistors on the PNP output bring the voltage down to 5V, which made the MOSFET get pretty hot (probably wasn't getting switched on fully). I tried using 20V (which drops to 9V), which made it a little better, but not fully.

On a datasheet, what do they call the voltage needed to switch the MOSFET on fully? Is it "Gate Threshold Voltage"?

Also I found a 52A "NDP6030L" MOSFET on an old motherboard that should work better.

The mosfet get hot because the Id of mosfet is not enough and the internal impedance 350 mOhm is too high when the current flow through Rds reach to 5A, it will get a lot of heat.

You can see the data on page 2 from datasheet.
The Vgs on the Gate Threshold Voltage, that is not really ON, when the Vgs reach to 4.5V then it will be fully ON for a 15A current, if you want to get more current then you should increasing the Vgs.

VGS(th),Gate Threshold Voltage,0.7V(min), 1V(typ), 2.2V(max)
RDS(ON),Static Drain-Source On-Resistance,VGS=4.5 V, ID=21 A,0.018~0.02 Ohm
ID(on),On-State Drain Current,VGS=4.5 V, VDS=10 V.

NPD6030L datasheet:
http://www.datasheetcatalog.org/datasheet/fairchild/NDP6030L.pdf

 

Ohhh... So if I my load is 0.5 ohms, and the MOSFET resistance is 0.35 ohms, then the voltage drop across the load is 2.94V, and the drop across the MOSFET is 2.06V. So here the MOSFET is dissipating 10 watts of power...

OK so with the NPD6030L, with a resistance of 0.0135 ohms, the MOSFET would only drop 0.14 volts, and so would only dissipate 1 watt.

Am I understanding this correctly?

 

Ohhh... So if I my load is 0.5 ohms, and the MOSFET resistance is 0.35 ohms, then the voltage drop across the load is 2.94V, and the drop across the MOSFET is 2.06V. So here the MOSFET is dissipating 10 watts of power...

OK so with the NPD6030L, with a resistance of 0.0135 ohms, the MOSFET would only drop 0.14 volts, and so would only dissipate 1 watt.

Am I understanding this correctly?

Yes, something like that.
That's why when we have the load is high current then we have to find the internal impedance of MOSFET is as low as possible.