[answered] What number tells you a MOSFET is Logic Level

I've seen MOSFET datasheets that say in the summary at the top, "Logic Level MOSFET" which is kind of the makers. But I suppose one could still be a logic level mosfet even if it doesn't say so in words.

So which of the many number in a datasheet is the one that says it is?

It is the drain source resistance, this parameter is quoted as being tested at some gate voltage. If it says 5V Gate to Source then you know it is a logic level device.

Thanks.

If it says 5V Gate to Source then you know it is a logic level device.

assuming that you have 5V logic. There are two things to look at. One is the "conditions" of the "rated" Rdson resistance, and Grumpy said. I've attached a piece of datasheet with that part highlighted in red. The other one, the ACTUAL parameter, is the Vgs(th) (Gate/source threshold voltage), which is the voltage at which the mosfet starts to turn on. Highlighted in Yellow in the picture.. You want this to be significantly smaller than your logic voltages.

Most mosfets that are NOT "logic level" will show Rdson measurement conditions of Vgs = 10V or so, so it's not going to be particularly subtle... (This happened to be a P-channel MOSFET, so the voltages show are negative, indicating that it turns on when the gate is at a lower voltage than the source. It still works with regular logic...)

westfw: Most mosfets that are NOT "logic level" will show Rdson measurement conditions of Vgs = 10V or so,

Yes I saw that when I compared two from International Rectifier, an IRLxxx and an IRF.

(Cunning of them to use the "L" in the part number ;) )

Hi there,

I like to see the words "Logic Level Mosfet" right on the data sheet or i do not assume it is logic level right off the bat, but i also check the threshold voltage to make sure it is very low because logic level mosfets have a very low threshold voltage so that by the time you get the gate up to maybe 3v it is turned fully on. I do not trust them if they show Vth=5v because many logic families do not actually put out 5 volts for the logic high, it can be 3.5v or even lower.

We have to be careful because CMOS high logic level can be as high as 15v or maybe 12v anyway. So if i say "logic level" maybe i am referring to the CMOS logic level with a 15v supply. A mosfet that works under that condition may never work with 5v logic of any kind.

Vth is nothing to do with a mosfet being switch on. A Vth of 5V means it is fully off at 5V or below.

MarkT: Vth is nothing to do with a mosfet being switch on. A Vth of 5V means it is fully off at 5V or below.

Hello there Mark,

I dont understand the reason for your statement. The states 'on' and 'off' are binary in nature, so if you have one you dont have the other, and if you have the other than you dont have the first. In other words, if something turns off at 5v then it also turns on at 5v. If the voltage is falling through 5v then it turns off, but if the voltage is rising through 5v then it turns on.

For your example, would you use that transistor in a circuit with 5v logic? And how do you know that?

Sometimes Vth is used to show when the device will turn on. In fact, it is sometimes used to show the minimum voltage that will turn the device on at least a little. Yes, it is better to know the typical Vgs and 'on' current level but Vth does tell you something interesting too. If you prefer though, then just look at Vgs vs Id or the like.

The states 'on' and 'off' are binary in nature, so if you have one you dont have the other, and if you have the other than you dont have the first.

That is true, but it doesn't apply to a FET which is not a binary device.

In other words, if something turns off at 5v then it also turns on at 5v.

No, that doesn't happen even with a binary device, and as I said the FET is not a binary device.

Sometimes Vth is used to show when the device will turn on.

No it is never used like this.

it is sometimes used to show the minimum voltage that will turn the device on at least a little.

Yes, it is ALWAYS used to show that. It is where the device STARTS to turn on.

So if i say "logic level" maybe i am referring to the CMOS logic level with a 15v supply.

You might be but the rest of the world does not use that term to mean that. When used in isolation the words "logic levels" referrer simply to TTL logic levels. Any other meaning of "logic levels" has to be accompanied by a qualifier.

Grumpy_Mike: That is true, but it doesn't apply to a FET which is not a binary device. No, that doesn't happen even with a binary device, and as I said the FET is not a binary device. No it is never used like this. Yes, it is ALWAYS used to show that. It is where the device STARTS to turn on. You might be but the rest of the world does not use that term to mean that. When used in isolation the words "logic levels" referrer simply to TTL logic levels. Any other meaning of "logic levels" has to be accompanied by a qualifier.

HEY THERE Grumpy, havent heard from you in a while. Did Snow White let you out of your cage again? He he, just kiddin'.

It's actually good to hear from you, and i mean that sincerely. I thought you may have left the board or something. I dont like to see people leaving a place where things can be discussed that are usually interesting.

Anyway, the point of that 5v being a 'binary' threshold does not mean that the FET (or MOSFET) is a turn on or turn off device. It means that that 5v threshold is USED as a binary set. It doesnt matter that it really is or not, it matters what we take it to be in theory. For example, if i say to you that i typed this in the 'day' or 'daytime' you can assume that i did not type it at night. You dont have to delve into the fact that the sun appears to rise slowly as the earth revolves. Day and Night form a binary pair that we can use in reasoning, and do not have to go into detail about how we go from Day to Night or vice versa. The same is true for this 5v. We dont have to follow the action down to 4.99v, 4.98v, 4.97v, etc., because we can see that 5v clearly and that tells us something about the operation of the device.

But here is what Wikipedia has to say about it: START_QUOTE The threshold voltage, commonly abbreviated as Vth or VGS (th), of a field-effect transistor (FET) is the minimum gate-to-source voltage differential that is needed to create a conducting path between the source and drain terminals. END_QUOTE

I do agree however that most of the time when we say "Logic Level" we mean some sort of 5v system like TTL or in more recent times CMOS. Interestingly though, 3.3v systems are becoming more and more popular.

It means that that 5v threshold is USED as a binary set.

It doesnt matter that it really is or not, it matters what we take it to be in theory.

And that is crap.

But here is what Wikipedia has to say about it: START_QUOTE The threshold voltage, commonly abbreviated as Vth or VGS (th), of a field-effect transistor (FET) is the minimum gate-to-source voltage differential that is needed to create a conducting path between the source and drain terminals. END_QUOTE

And who could find fault with that? However your leap that this means the FET is in any way "on" is something I would argue with. That "conducting path" could very well have a resistance of say 1M, and that would not imply by any means the FET was on.

The threshold voltage parameter plays no part in being able to determine if a FET can be used at TTL logic levels as a switch. Go and ruin your own devices if you want but don't peddle crap to beginners on this forum.

Grumpy_Mike: Twaddle. And that is crap. And who could find fault with that? However your leap that this means the FET is in any way "on" is something I would argue with. That "conducting path" could very well have a resistance of say 1M, and that would not imply by any means the FET was on.

The threshold voltage parameter plays no part in being able to determine if a FET can be used at TTL logic levels as a switch. Go and ruin your own devices if you want but don't peddle crap to beginners on this forum.

I guess we cant have a reasonable conversation here as you think you know everything so go ahead, but dont tell me what to do because your wasting your time.

I've had no trouble with this explanation since the late 1980's. If you want to knock it that's up to you, but dont tell me what is crap and what isnt because i wont believe you now, and i wont be able to believe you in the future either.

By your own statements, if you have a Vth of 5v then you cant use it for 'logic level' applications because it may only be 'on' by 1MOhm.. Funny how can we know this when Vth cant be used for anything, according to you.

I guess we cant have a reasonable conversation

I guess not if you are wrong and will not admit it.

I've had no trouble with this explanation since the late 1980's.

Not done much with it then have you?

and i wont be able to believe you in the future either.

Fine, this forum is not aimed at people like you, it is here to help people learn stuff, not to propagate misunderstandings no matter how long they have been held.

Funny how can we know this when Vth cant be used for anything, according to you.

Now was that a deliberate misunderstanding of what I said or do you not understand English. I said

The threshold voltage parameter plays no part in being able to determine if a FET can be used at TTL logic levels as a switch.

Note this is not saying it cant be used for anything. Do not lie.

MrAl: Day and Night form a binary pair that we can use in reasoning

Only if you ignore the time that sunrise and sunset take. The transition between "unambiguous day" to "unambiguous night" actually takes time, and between those states there will be a period of uncertainty when some people might say it's day, and others say it's night.

Digital is an abstraction that we can use to simplify reasoning about certain kinds of systems that operate with specific constraints, but the underlying reality is always stubbornly and messily analog.

Transistors are analog devices. It is possible to use them in a digital manner by forcing them to always operate in saturation or cutoff, but in order to do that you need to know the analog boundaries between these modes and the linear region. The Threshold Voltage only gives you the upper bound of the cutoff region, it doesn't tell you the lower bound of saturation.

Stop saying stupid and wrong crap. MOSFETs do not suddenly transition from OFF to ON as the gate voltage rises, and teaching that to someone who doesn't know any better is just going to make them drive it incorrectly and needlessly destroy their parts.

I guess we cant have a reasonable conversation here as you think you know everything so go ahead, but dont tell me what to do because your wasting your time.

I've had no trouble with this explanation since the late 1980's. If you want to knock it that's up to you, but dont tell me what is crap and what isnt because i wont believe you now, and i wont be able to believe you in the future either.

By your own statements, if you have a Vth of 5v then you cant use it for 'logic level' applications because it may only be 'on' by 1MOhm.. Funny how can we know this when Vth cant be used for anything, according to you.

It's like you're not even trying to understand what Mike is writing.

A FET with a 5V threshold would indeed be completely unsuitable for Arduino use. I believe most logic level FETs actually quote a 4.5V theshold, and the graphs will show how much current they can actually handle with the extra 0.5V gate voltage.

Parts like this are nice http://aosmd.com/res/data_sheets/AOI510.pdf Even a 74HC595 with its miserly 6mA output can drive the gate sufficiently to fully turn on & off.

Which is why they invented the Vgs(on) spec which specifies (at a certain drain voltage and current) the voltage when the MOSFET can be considered "on".

Vgs(th) is more suited to SPICE modeling.

KeithRB: Which is why they invented the Vgs(on) spec which specifies (at a certain drain voltage and current) the voltage when the MOSFET can be considered "on".

Vgs(th) is more suited to SPICE modeling.

I just checked all of my MOSFET datasheets* and none of them have that spec. Are you confusing it with RDS(on)?

*2N7000, DMG6968, DMG3415, BSS84, BSS138, FQP27P06 and FQP30N06L

Jiggy-Ninja: Only if you ignore the time that sunrise and sunset take. The transition between "unambiguous day" to "unambiguous night" actually takes time, and between those states there will be a period of uncertainty when some people might say it's day, and others say it's night.

Digital is an abstraction that we can use to simplify reasoning about certain kinds of systems that operate with specific constraints, but the underlying reality is always stubbornly and messily analog.

Transistors are analog devices. It is possible to use them in a digital manner by forcing them to always operate in saturation or cutoff, but in order to do that you need to know the analog boundaries between these modes and the linear region. The Threshold Voltage only gives you the upper bound of the cutoff region, it doesn't tell you the lower bound of saturation.

Stop saying stupid and wrong crap. MOSFETs do not suddenly transition from OFF to ON as the gate voltage rises, and teaching that to someone who doesn't know any better is just going to make them drive it incorrectly and needlessly destroy their parts.It's like you're not even trying to understand what Mike is writing.

A FET with a 5V threshold would indeed be completely unsuitable for Arduino use. I believe most logic level FETs actually quote a 4.5V theshold, and the graphs will show how much current they can actually handle with the extra 0.5V gate voltage.

Hello,

Some of your statements are not very reasonable.

In your first paragraph you state that you sometimes cant tell the difference between night and day. So you are willing to NEVER be able to tell the difference between night and day just because in the morning we might have 1/2 hour where we might call it either and in the evening for 1/2 hour we might be able to call it either. So yeah, let's say that we can never tell the difference because of those two times which total about 1/24 which is less than 5 percent of the time. You see how ridiculous that looks now?

In your last paragraph you void what you said in the first paragraph by ACTUALLY BEING ABLE TO DETERMINE EXACTLY WHAT I HAD SAID TO START WITH while saying you cant do that. Either you can tell the difference, or you cant, which is it? You looked at Vth and determined that the device was unsuitable, yet you and Mr Grumps state that you CANT DO THAT FROM Vth. So you are stating that you CANT do that, then you actually DO THAT. So make up your mind, can you or cant you do that?

You see why i have to chuckle a little.

Also it would be nice if Mr Grumps would be a little more civil in his conversations. If someone does not have respect for me FOR ANY REASON including but not limited to being wrong on some topic, then i cant have respect for them either.

What else you can do after you look at Vth is look at the other spec we had been talking about, which is the Vgs at various current levels. I dont think anyone would disagree with that, but maybe you can find something wrong with that too :-)

You two better study this well because there will be a test on this on Friday :-)