MOSFET - momentary on?

[MrNaz]

So MOSFETs. I followed the tutorial here:
http://www.youtube.com/watch?v=Te5YYVZiOKs
but I found that the MOSFET stayed on when I pulled the voltage from the gate. A little research led me to the fact that I need to short the gate to the source so that the electrostatic field on the gate can be neutralized when the gate is disconnected from the V+. This makes sense to me.

Now, I want *all* my MOSFETs to be momentary on when the gate sees a voltage. I intend to modify my MOSFETs by soldering the necessary 10kO resistor to them like so:

Is this a good idea? Will it work? Am I just breaking all my MOSFETs by doing this? Is there a better way to achieve what I need to do?

Thanks all!

 

[HKJ]

It is a very good idea, you might blow you mosfet if you do not do it.

 

[lolzertank]

That's a great idea! As HKJ said, not having one can destroy your MOSFET from excessive heat. Stupid linear zone... :green: :mecry:

 

[MrNaz]

Out of interest, why don't MOSFETs just integrate the resistor into the package? Surely MOSFETs are supposed to be momentary on, and not stay on when the gate is charged and then isolated.

Even if you wanted momentary on this would be a poor way to achieve it, surely grounding the gate and then using a proper momentary on circuit or switch would be the proper way to do it.

So, why isn't the resistor just built into the MOSFET?

 

[Mr Happy]

Out of interest, why don't MOSFETs just integrate the resistor into the package? Surely MOSFETs are supposed to be momentary on, and not stay on when the gate is charged and then isolated.

Even if you wanted momentary on this would be a poor way to achieve it, surely grounding the gate and then using a proper momentary on circuit or switch would be the proper way to do it.

So, why isn't the resistor just built into the MOSFET?

The resistor is normally designed into the full circuit where the FET will be used. The exact value of the resistor or other component is at the discretion of the circuit designer, so leaving it out of the FET package is more useful overall. You can always add a resistor, but you can never take it away if it is built in by the manufacturer.

 

[MrNaz]

The resistor is normally designed into the full circuit where the FET will be used. The exact value of the resistor or other component is at the discretion of the circuit designer, so leaving it out of the FET package is more useful overall. You can always add a resistor, but you can never take it away if it is built in by the manufacturer.

I understand, however, on that line of thinking you can take every component out of ICs. What I'm asking (in a round about fashion) is what applications are there for MOSFETs with anything other than a resistor between gate and source? Also, how does the resistor value alter the MOSFET's behavior? I know that non-logic MOSFETs can use a pot between G and S to alter the resistance of the D-S path, but in a logic FET, how is this useful?

Thanks for your help!

 

[HKJ]

I understand, however, on that line of thinking you can take every component out of ICs. What I'm asking (in a round about fashion) is what applications are there for MOSFETs with anything other than a resistor between gate and source? Also, how does the resistor value alter the MOSFET's behavior? I know that non-logic MOSFETs can use a pot between G and S to alter the resistance of the D-S path, but in a logic FET, how is this useful?

Thanks for your help!

This depending on what kind of circuit you make, for low power circuit you either what a very high value resistor or directly drive it from a digital output, for high speed circuit you need a rather low resistance or you are using a real driver with low output impedance and they again you do not need the resistor.

I do not know about mosfet/ic design, but I believe that you need another type of design if you wish to incorporate components together with the mosfet.

 

[phalanx]

I understand, however, on that line of thinking you can take every component out of ICs. What I'm asking (in a round about fashion) is what applications are there for MOSFETs with anything other than a resistor between gate and source? Also, how does the resistor value alter the MOSFET's behavior? I know that non-logic MOSFETs can use a pot between G and S to alter the resistance of the D-S path, but in a logic FET, how is this useful?

Thanks for your help!

Let's see... You asked a few questions there!

Unlike a BJT which is a current controlled device, MOSFETs are voltage controlled. The gate on a MOSFET behaves like a capacitor in terms of circuit interaction. This means that when you apply a voltage to it, there will be short period of time where the gate is charging and current is actually flowing into it. During this period the circuit is actually consuming power and draining your batteries. Once the gate is fully charged, current stops flowing (except for a very small leakage current) and the MOSFET stops consuming power from the circuit. The ability to remain in an "ON" state while drawing a nearly negligible amount of power is one of the biggest benefits of working with MOSFETS.

You have probably heard people say that when working on high power electronics, you should short or ground all large capacitors because they could contain dangerous voltage levels in them. Since the original circuit did not provide a path to drain the charge out, they can remain energized for a long period of time. Since the gate on a MOSFET is essentially a capacitor, when you remove a voltage source from it and you don't provide a path to ground for it to discharge, enough charge will remain in the gate to leave the MOSFET in an "on" state. The resistor that you add between the Gate and Source provides a path for the charge to flow when you remove your voltage source from the Gate. A resistor in this configuration is called a "pull-down resistor" since it is pulling the gate to a lower voltage level. Once the charge has drained, the voltage of the Gate will equal the voltage of the source and your MOSFET will be in an "off" state.

The reason why resistors are not designed into the MOSFET package is it defeats the low power benefits that it provides. Without a resistor the leakage current of the gate is usually a good deal less than 100nA (.0000001A) but it will vary by device and gate voltage. In a 5V system with a 10K gate to source resistor you will be using 500uA (plus the original leakage) which is 5000 times more current than you were drawing before. While 500uA doesn't sound like a lot, for low power applications it can be huge. If your device has a lot of of MOSFETS, the power consumption can rise rather quickly.

In your application where you do not directly short the gate to ground, the pull-down resistor will change the time it takes for MOSFET to shut off (fall time). The more resistance between your gate and source, the longer it will take for the gate to discharge. Likewise, when you decrease the resistance, it will shut off faster but with the down side of drawing more current when in normal operation. Shut off time is important especially in higher current applications because while the MOSFET is transitioning from "on" to "off" it is operating in its linear zone which can cause it to dissipate a lot of heat from the load you are connected to. I could elaborate on that a lot more but it would be best to save it for the next lesson. More often than not, it is favorable to turn off the MOSFET as quickly as possible.

Lastly you asked about logic vs. non-logic MOSFETS. The only difference between them is logic MOSFETS require a lower gate voltage to reach their full "on" state. When people use a pot between gate and source, they are usually trying to set up a voltage divider for the gate which would force the MOSFET into its linear region. In the linear region the MOSFET behaves like a resistor. The pot allows it to behave like a variable resistor with current handling capabilities much greater than the pot alone.

I hope that helps to answer some of your questions!

 

[wquiles]

Let's see... You asked a few questions there!
...

Very good explanation. I would like to contribute that you need to know if the circuit calls for an N-channel of P-channel MOSFET, so that you can wire it properly, since you can have the MOSFET on the HIGH side or the LOW side of a circuit. To be specific, this image shows how a typical (N-channel) MOSFET is wired to switch on the LOW side of the circuit, meaning that it completes the ground path and therefore allows current to flow (the reed switch would be the equivalent momentary switch):

Will