Gate Current in MOSFET

[Sceadwian]

Calm down now Claude
I ment electric field not voltage field, my mistake.
And no I'm not contradicting myself when I say a FET needs charge not current. It needs current to gain the charge to switch from one state to the other, but once that charge is obtained no current is required to retain that switched state, that's all I was saying, just look at flash memory, the data is stored in electric charges that are sitting on the floating gate capacitance of a FET which can be read via an extra gate.

 

[MrAl]

Hi there,


I think maybe you two are arguing the same thing except looking at it in a different way

The MOSFET needs 'current' to operate dynamically, but only needs 'charge' to operate statically.

Normally we would be switching the MOSFET and we would have to do so within a given time period, so we would have to
choose a gate driver that could handle the required current level so that the gate charge could in fact charge up fast enough
to operate the device as fast as we need it to switch. Once the gate charges up however, no more current is needed, at
least in theory.

 

[Sceadwian]

Let me use this post to replace The #14 post, that Claude responded to with the #20 post with the following single sentence. Claude, I believe you should say instead that MOSFETs are charge controlled devices typically driven by voltage sources.

Yes MrAl, that's why I said calm down to Claude, cause I didn't mean to argue anything at all =P Claude didn't however use the word charge controlled device in his post. I know he's argued for them being charge controlled devices in other threads because I was part of one of them that turned into a big deal (Much like yours and my arguement in the Ohm's law thread MrAl)

However I still side with description of the semi-conductor effect as being defined by the electric field, in many devices current is required to be moving for those fields to exist, however it is the field itself which causes the change in the conduction state of the material, these fields are VERY closely bound to atomic structure so it's difficult to describe them without the currents, and in the real world the currents are actually the important part of what's going on, they are the initiator of the switching action (please read switching action as change in conductance), but they are not the cause of it. That's where me and Claude disagree. Because as modeled (and believe me the models are damn accurate) a transistor can be completely described by the currents flowing through the junctions and their discrete properties. However without the electric fields that are produced the actual PN junctions, those currents can't flow.

In slightly different words a difference of potential can not CHANGE without the flow of a current, however the electric field itself changes the possible electron configurations in the atoms near it to allow current to flow in the first place. The manipulation of those boundaries is the entire foundation of nearly all modern commercial semi conductor devices.


A field can exist without a current, the field causes the state in a semi conductor material. To change states the field has to change which means a current needs to flow. It's a symbiotic relationship for switching, it is however the field that determine the actual state.

 

[Claude Abraham]

Let me use this post to replace The #14 post, that Claude responded to with the #20 post with the following single sentence. Claude, I believe you should say instead that MOSFETs are charge controlled devices typically driven by voltage sources.Yes MrAl, that's why I said calm down to Claude, cause I didn't mean to argue anything at all =P Claude didn't however use the word charge controlled device in his post. I know he's argued for them being charge controlled devices in other threads because I was part of one of them that turned into a big deal (Much like yours and my arguement in the Ohm's law thread MrAl)

However I still side with description of the semi-conductor effect as being defined by the electric field, in many devices current is required to be moving for those fields to exist, however it is the field itself which causes the change in the conduction state of the material, these fields are VERY closely bound to atomic structure so it's difficult to describe them without the currents, and in the real world the currents are actually the important part of what's going on, they are the initiator of the switching action (please read switching action as change in conductance), but they are not the cause of it. That's where me and Claude disagree. Because as modeled (and believe me the models are damn accurate) a transistor can be completely described by the currents flowing through the junctions and their discrete properties. However without the electric fields that are produced the actual PN junctions, those currents can't flow.

In slightly different words a difference of potential can not CHANGE without the flow of a current, however the electric field itself changes the possible electron configurations in the atoms near it to allow current to flow in the first place. The manipulation of those boundaries is the entire foundation of nearly all modern commercial semi conductor devices.


A field can exist without a current, the field causes the state in a semi conductor material. To change states the field has to change which means a current needs to flow. It's a symbiotic relationship for switching, it is however the field that determine the actual state.

1st bold highlight. "MOSFETs are charge controlled devices best driven from a *constant voltage* source"! But I've stated the same since forever. Remember the OP topic, "gate current in MOSFET". I was explaining its significance. Sure a FET needs current to charge the gate & establish the E field. I've argued that for years. Once the FET is in a given state, no gate current is needed to hold that state. I agree with Mr. Al that FETs need current in the dynamic mode, but not in the static mode. We have universal agreement there.

2nd highlight. Sure, fields are all important. But fields exist due to the presence of charges. A charged particle is surrounded by an E field. I know about time changing magnetic fields inducing E fields, but that is not how a bjt or FET operates. Any E field is due to charges. The quantity & location/distribution of said charges ultimately controls device behavior. Voltage is defined as work per unit charge along a specific path. But what is the work done? It is the line integral of force over distance. But what defines said force? Answer, Coulomb force, an inherent property of charges. First, charges exerting forces against one another over distance give meaning to potential or voltage. E fields are due to charge, and voltage is a derived quantity. Thus the bjt & the FET are ultimately charge controlled when viewing them at the mIcroscopic semiconductor physics level.

3rd highlight. "A difference of potential cannot change without the flow of current"!!! Those are my words exactly for the last 30 years of my EE career. We agree. How about that!

As far as a field existing w/o a current goes, it depends. For a FET in the static mode, charge flow from drain to source does not expend energy from the gate to source E field. The E field from gate to source requires no gate current to sustain it, only to change it. But in a bjt, the E field from base to emiter would be neutralized by charge flowing from base to emitter, holes, and emitter to base recombination, electrons, ref npn. Thus base & emitter current are needed to sustain the E field, even in static conditions. So a bjt needs current just to stay in the same state. Likewise for diodes, LEDs, SCRs, etc. The E field requires I & V to switch, and in the case of non-insulated semiconductors, to sustain a static state as well. You earlier stated that I & V are basically co-dependent. I've never said otherwise.

"The field determines the state" is just a roundabout way of saying that bjt & FET are *charge controlled*. I & V are important & necessary, but charge is the most basic quantity, and devices are said to be charge controlled.

In a nutshell, what is the significance of gate current in a FET per the OP original question? Answer, in the analog/amplifying mode, the gate needs current to amplify an analog signal. At low frequencies, said gate current is so small, one can inadvertently ignore it and not even know its significance. But although it is small, it's still important. At high frequencies, gate current is too large to ignore. In switching mode, gate current is needed to switch states, but not to sustain a state.

 

[Sceadwian]

Claude, I think at this point if you have anything further to say to me please do so in private messages, I've stated what I was trying to say which is not nearly as complicated as you've made it so far, however you're reading my poor choice of words in a manner which I didn't intend and it's turning into a perfect storm of non-communication nothing is going to be gained by arguing further.

The only intent in my posting on the topic at all was to add the simple caveat that regardless of the currents flowing in a semi-conductor device it's the electric fields that exist that are actually responsible for its conduction state, in the basic example of an ideal mosfet it's the potential difference between the gate and source and nothing else that determines the semi conductors state, without a current flowing this state will remain indefinitly based only on the electric fields present. In the real world there is a lot more actually going on which causes charge carriers to be exchanged (even in the best mosfets) even with a static gate drive and real currents flow in both DC and AC modes, this does not change the absolute fact that the semi conductor state itself is determined by the electric field. As an EE you deal with the real world more than the purely abstract theoretical and so you know a whole lot more about the interrelation in real world devices between current and voltage and the descriptions and you know the interdependency of the two on a macro scale. This does not change the simple fact that the field causes the actual conduction state.

SCIENCE HOBBYIST: how transistor works, an alternate viewpoint
This one particular link gave me the alternate viewpoint from what is common among even professors of electrical engineering, because the models are already well established and work fine, there is no need to understand or even acknowledge this basic fact to be the best electrical engineer in the world with a lifetime of experience.

I was simply trying to provide a glimpse to the original poster of the intricacies of terminology and the actual truths behind the actual physical phenomenon of the basic semi conductor effect. You can provide a degree, list every job reference you've ever had and post an encyclopedias worth of math to show the requirements of charge carrier mobility for a thousand semi conductor devices. The electric field will go on about it's buisness actually making it work =)

We've both said all that we can say in even a partially civilized manner to each other at this point. If you still disagree with me then again please private message the response, so we can discuss and appropriate public response as us arguing is benefiting anyone and until you and I can reconile our differences we shouldn't be discussing it in the posters thread.

 

[Hero999]

In a common source amplifier there's lots of negative feedback which reduces the gate capacitance.