Author Topic: FETs are voltage controlled devices?  (Read 13053 times)

0 Members and 1 Guest are viewing this topic.

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
FETs are voltage controlled devices?
« on: May 15, 2015, 03:49:15 am »
I'm trying to replace a buffer used in a power control circuit with a mosfet and I'm confused about something.

The circuit in question is here:
[link removed]

I want to replace the buffer because the buffer's output is not 5V tolerant when it's powered by 3.3V and 3.3V cannot drive the input when it's powered by 5V.  I considered using a translator instead, but it's more expensive and seems like overkill just to switch the voltage source.

I found this circuit, in the section "MOSFET (Pass Transistor) Interface" which I think could replace it:
http://www.savagecircuits.com/content.php?85-Mixed-Voltage-Systems-Interfacing-5V-and-3-3V-Devices

It's also seen here:
http://www.hobbytronics.co.uk/mosfet-voltage-level-converter
http://electronics.stackexchange.com/questions/97889/is-there-any-bidirectional-5v-3-3v-level-shifter

I was going to use 100K resistors for the pullups, to reduce power consumption.  I assume that won't be an issue since this circuit doesn't need to switch fast.

The question I have however is this...  If my microcontroller's pins aren't 5V tolerant, what happens when the FET turns on?  Will the microcontroller pin see 5V?

I asked someone else, and they said no, a mosfet is a voltage controlled device, and if the gate is at 3.3V then the source will never see more than 3.3V.

That seemed to make sense, until I considered using a mosfet as a switch:
http://www.electronics-tutorials.ws/transistor/tran_7.html

Here, an N-FET is used in a configuration which isn't terribly different, yet I can apply 3.3V to the gate and switch a 20V power source, if I have a logic-level FET. 

So what gives?  Was this person wrong?  Were they referring to non-logic level FETs?  I'm pretty sure even a generic NFET can connect a 20V power source to the source with 5V at the gate and that that would certainly blow up a 5V microcontroller pin connected to said source.

And most importantly, will what I have planned work?  This is the part I intended to use in place of the two FETs, with the 100K pullups I mentioned:
http://www.diodes.com/datasheets/DMC2038LVT.pdf

It's a PFET and NFET in a single package with sufficient current carrying capacity and power dissipation for what the PFET's body diode may need to conduct.
« Last Edit: July 14, 2016, 11:57:15 am by Starlord »
 

Offline PSR B1257

  • Regular Contributor
  • *
  • Posts: 243
  • Country: de
Re: FETs are voltage controlled devices?
« Reply #1 on: May 15, 2015, 07:42:37 am »
Well, that's all rather confusing...

All you need is a NPN BJT with it's collector tied to the gate of the MOSFET. You can drive the base of the BJT from either 3.3V or 5V through a 10k resistor (value not that critical) for instance.
In theory, there is no difference between theory and practice. But, in practice, there is.
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #2 on: May 15, 2015, 07:54:52 am »
1. Why did you specify a BJT instead of a FET? 
2. Wouldn't the logic on the USB_HOST_EN pin then need to be reversed from the original circuit?  I cannot change the signal the microcontroller will output.

I also still want to understand how that level translation circuit protects the 3.3V side because there's clearly something I don't understand about mosfets if in fact that will protect the microcontroller on the 3.3V side from overvoltage at its pin due to the 5V pull up.
 

Offline PSR B1257

  • Regular Contributor
  • *
  • Posts: 243
  • Country: de
Re: FETs are voltage controlled devices?
« Reply #3 on: May 15, 2015, 09:02:22 am »
Quote
1. Why did you specify a BJT instead of a FET?
Both works. You just don't have to look for a logic level FET, since a BJT works even with 1V.

Quote
2. Wouldn't the logic on the USB_HOST_EN pin then need to be reversed from the original circuit?
Yes, it would. In this case, a second stage is needed.

Quote
I also still want to understand how that level translation circuit protects the 3.3V side because there's clearly something I don't understand about mosfets if in fact that will protect the microcontroller on the 3.3V side from overvoltage at its pin due to the 5V pull up.
If you replace the buffer IC3 with a transistor (BJT or MOSFET dosen't matter) you are seperating the two voltage rails. Unless the transistor goes smoky, there is no way for the 5V rail to feed through to the 3.3V rail.
In theory, there is no difference between theory and practice. But, in practice, there is.
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #4 on: May 15, 2015, 03:30:41 pm »
If you replace the buffer IC3 with a transistor (BJT or MOSFET dosen't matter) you are seperating the two voltage rails. Unless the transistor goes smoky, there is no way for the 5V rail to feed through to the 3.3V rail.

Okay, but that's the bit I don't understand.  Nothing I have read about mosfets states that Vds cannot exceed the gate voltage.  In fact, the whole point of a logic level mosfet is to allow switching higher voltages with a lower voltage, is it not?

Also when a mosfet is on, it's like a very low value resistor.  If I just stuck a resistor between my 5V rail and my uC pin I would destroy it unless the pin had voltage clamping diodes, which it doesn't.

So how can the mosfet prevent 5V from flowing from one side to the other, when it's on?
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #5 on: May 15, 2015, 03:41:17 pm »
Here's a nice tutorial on mosfets:
http://www.electronics-tutorials.ws/transistor/tran_6.html

Is there anything on that page that supports your claim that 5V will not flow from drain to source when the fet is on, which it will be when the uC pulls the source low?  because looking at this graph:


...what I see is that Vds can be as high as it likes so long as the Vgs is high enough to put the mosfet into it's full saturation region.
 

Offline LvW

  • Frequent Contributor
  • **
  • Posts: 282
  • Country: de
Re: FETs are voltage controlled devices?
« Reply #6 on: May 15, 2015, 04:17:04 pm »
May I give a short comment to the above contribution?

For my opinion it is not very helpful because the shown graph is misleading and wrong.
The question concerns the OHMIC region only (and not the saturation region) - and this region is not correctly shown in the graph.
From the graph one could derive that the ohmic resistance is nearly the same for all VGS values (constant slope) - and this is not true.
In reality, it is possible to control the FET resistance over a wide range with different VGS values.
It seems that the producer of this graph has used BJT characteristics - edited with VGS values.
(By the way - the "tutorial" contained in the mentioned link contains many errors and false statements). 

See Fig.1 in the following link for a correct FET characteristic:
http://www.diodes.com/datasheets/DMC2038LVT.pdf
« Last Edit: May 15, 2015, 04:20:54 pm by LvW »
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #7 on: May 15, 2015, 04:37:52 pm »
Quote
The question concerns the OHMIC region only (and not the saturation region) - and this region is not correctly shown in the graph.

The question?  You mean my question? 

Why does my question concern the OHMIC region only?  The FET I'm using is logic level, and Vth is way below the 3.3V which will be applied at the gate when the FET is on and the uC pin is pulling the source low.  I would expect the FET to be in full saturation.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 13425
  • Country: gb
  • 0999
Re: FETs are voltage controlled devices?
« Reply #8 on: May 15, 2015, 05:25:32 pm »
No, the saturation region is when the current through the FET depends on VGS. In this case, you don't want your FET to be operating in the saturation region.

 

Offline IanB

  • Super Contributor
  • ***
  • Posts: 9618
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #9 on: May 15, 2015, 05:26:33 pm »
No, the saturation region is when the current through the FET depends on VGS. In this case, you don't want your FET to be operating in the saturation region.
What if the FET is being used as a switch?
I'm not an EE--what am I doing here?
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #10 on: May 15, 2015, 05:30:14 pm »
No, the saturation region is when the current through the FET depends on VGS. In this case, you don't want your FET to be operating in the saturation region.

I don't understand.  The gate in this circuit is being pulled well above Vth.  So how could it not be operating in the saturation region?
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 13425
  • Country: gb
  • 0999
Re: FETs are voltage controlled devices?
« Reply #11 on: May 15, 2015, 05:42:42 pm »
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #12 on: May 15, 2015, 06:25:43 pm »
You're confusing FETs with BJTs.

http://en.wikipedia.org/wiki/MOSFET#Modes_of_operation

In what way am I confusing FETs with BJTs?

Quote
Triode mode or linear region (also known as the ohmic mode)
When VGS > Vth and VDS < ( VGS – Vth )


Saturation or active mode
When VGS > Vth and VDS ? ( VGS – Vth )

When the uC pin is pulled down, the source sees 0V, the drain sees 5V, and the gate sees 3.3V.

So, with this NFET:
http://www.diodes.com/datasheets/DMC2038LVT.pdf

Vth = 0.4 ... 1V
Vgs = 3.3V - 0V = 3.3V
Vds = 5V - 0V = 5V
Vgs-Vth = 3.3V - 1V = 2.3V

And therefore:

Vgs > Vth and Vds > Vgs-Vth
 
Which Wikipedia states is saturation or active mode, not ohmic mode.
 

Offline AndreasF

  • Frequent Contributor
  • **
  • Posts: 251
  • Country: gb
    • mind-dump.net
Re: FETs are voltage controlled devices?
« Reply #13 on: May 15, 2015, 07:38:33 pm »
from the second link you provided:
Quote
When the high side transmits a '1' (5V) the MOSFET source pin is pulled up to 3.3V and the MOSFET is OFF.

... so as soon as the source voltage is pulled up high enough by the conducted voltage from the drain side, VGS is below the threshold and the FET no longer conducts.
my random ramblings mind-dump.net
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #14 on: May 15, 2015, 07:48:07 pm »
from the second link you provided:
Quote
When the high side transmits a '1' (5V) the MOSFET source pin is pulled up to 3.3V and the MOSFET is OFF.

... so as soon as the source voltage is pulled up high enough by the conducted voltage from the drain side, VGS is below the threshold and the FET no longer conducts.

I don't understand.

When the mosfet is on, which it is when the uC's pin is low, the uC's pin is connected through the mosfet directly to the 5V pullup.

And connecting that 5V 100K pullup directly to my uC's pin would destroy it right?
(The pin has no overvoltage clamping diode and is not 5V tolerant.)

So how does having the mosfet between them make a difference, if the mosfet is on? 

Even if it's only partially on and has some small resistance, that simply increases that 100K resistance between the pin and the 5V rail slightly.
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 6093
  • Country: gb
Re: FETs are voltage controlled devices?
« Reply #15 on: May 15, 2015, 08:00:56 pm »
from the second link you provided:
Quote
When the high side transmits a '1' (5V) the MOSFET source pin is pulled up to 3.3V and the MOSFET is OFF.

... so as soon as the source voltage is pulled up high enough by the conducted voltage from the drain side, VGS is below the threshold and the FET no longer conducts.

I don't understand.

When the mosfet is on, which it is when the uC's pin is low, the uC's pin is connected through the mosfet directly to the 5V pullup.

And connecting that 5V 100K pullup directly to my uC's pin would destroy it right?
(The pin has no overvoltage clamping diode and is not 5V tolerant.)

So how does having the mosfet between them make a difference, if the mosfet is on? 

Even if it's only partially on and has some small resistance, that simply increases that 100K resistance between the pin and the 5V rail slightly.

You are aware the gate is insulated.. right?

Also, you should stop asking people questions if they say things like "if the gate is at 3.3V then the source will never see more than 3.3V.".
 

Offline AndreasF

  • Frequent Contributor
  • **
  • Posts: 251
  • Country: gb
    • mind-dump.net
Re: FETs are voltage controlled devices?
« Reply #16 on: May 15, 2015, 08:08:14 pm »
The MOSFET is only "on" if the voltage difference between gate and source is above a certain threshold, so if the source pin (the one connected to your uC) is being pulled up by applying a high voltage from the drain side, the voltage difference between gate and source is reduced and thereby effectively turns the MOSFET "off".
my random ramblings mind-dump.net
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #17 on: May 15, 2015, 08:31:47 pm »
You are aware the gate is insulated.. right?

Yes.  But how does that change anything?


Quote
Also, you should stop asking people questions if they say things like "if the gate is at 3.3V then the source will never see more than 3.3V.".

I trust that it will probably work the way I want it to, but if I don't ask why it will work then I'll never understand mosfets properly and the next time I go to use one I'll be just as lost.
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #18 on: May 15, 2015, 08:33:17 pm »
The MOSFET is only "on" if the voltage difference between gate and source is above a certain threshold, so if the source pin (the one connected to your uC) is being pulled up by applying a high voltage from the drain side, the voltage difference between gate and source is reduced and thereby effectively turns the MOSFET "off".

You're still not making sense.

The mosfet IS on when the uC pin is low.  It HAS to be, because if it isn't then it cannot possibly pull the 5V side low and it would therefore not function as a level shifter.
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #19 on: May 15, 2015, 08:38:55 pm »
Replace the mosfet in this picture with a resistor:


When the uC pin is set low, the mosfet turns on and it becomes a hopefully, very low value resistor. 

Now I've got a 5V pullup connected to my 3V microcontroller pin.  And everything I know says that will destroy my microcontroller.  Even with that 3.3V pullup there pulling the voltage down slightly.




 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #20 on: May 15, 2015, 08:46:55 pm »
Also, you should stop asking people questions if they say things like "if the gate is at 3.3V then the source will never see more than 3.3V.".

Also, that statement doesn't make sense.

If that were true, I could not switch a 20V source using logic level mosfet because the gate would never see more than 3-5V.

http://www.electronics-tutorials.ws/transistor/tran_7.html



uC pin goes high, gate sees 5V, source is connected to 20V source powering lamp.  Source sees more than 5V.
« Last Edit: May 15, 2015, 08:50:12 pm by Starlord »
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #21 on: May 15, 2015, 09:19:53 pm »
The MOSFET is only "on" if the voltage difference between gate and source is above a certain threshold, so if the source pin (the one connected to your uC) is being pulled up by applying a high voltage from the drain side, the voltage difference between gate and source is reduced and thereby effectively turns the MOSFET "off".

Wait.  Perhaps you mean it turns it partially off? 

Ie, the 5V side - the drain in this case, pulls the "3.3V" side, ie, the source (and my microcontroller pin) up from 0V to... what?  Some voltage where the mosfet's resistance is high enough that some kind of equilibrium is reached?  I mean it can't reach 3.3V because that would turn the mosfet "off" and if the mosfet is "off" then it can't pull the 5V side down to 0V to do the level translation.

But at the same time, the 5V side can't pull the source up that much, because if the source isn't at 0V then how could the 5V side be at 0V?

So this equlibrium point would have to be slightly more than 0V but still very close to 0V, wouldn't it? And the mosfet's resistance would have to be much less than the pullup on the 5V side.  Because there's a voltage divider formed there between the mosfet's resistance and the voltage at the source and the 5V pullup. 

Ugh, none of this is making any sense.

And I tried to simulate this with Digikey's PartSim and I was unable to ever get the voltage on the 5V side below 5V. :( 
 

Offline edavid

  • Super Contributor
  • ***
  • Posts: 2903
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #22 on: May 15, 2015, 09:23:21 pm »
Also, you should stop asking people questions if they say things like "if the gate is at 3.3V then the source will never see more than 3.3V.".

Also, that statement doesn't make sense.
It does make sense.  The MOSFET will not conduct when VGS is 0V, therefore the source voltage will not be higher than the gate voltage.

Quote
If that were true, I could not switch a 20V source using logic level mosfet because the gate would never see more than 3-5V.
The 20V switch is a common source circuit ("low side switch").  The level translator is a common drain circuit.  Totally different.  In fact, you can't easily use an N MOSFET as a 20V high side switch.

Quote
uC pin goes high, gate sees 5V, source is connected to 20V source powering lamp.  Source sees more than 5V.
Now this does not make sense.  In the 20V switch circuit, the source is grounded.
« Last Edit: May 15, 2015, 09:26:27 pm by edavid »
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 6093
  • Country: gb
Re: FETs are voltage controlled devices?
« Reply #23 on: May 15, 2015, 09:37:47 pm »
Perhaps you could explain in a little more detail what you're doing. What micro, what type of FET you have selected, etc..
 

Offline Starlord

  • Frequent Contributor
  • **
  • Posts: 325
  • Country: us
Re: FETs are voltage controlled devices?
« Reply #24 on: May 15, 2015, 09:47:01 pm »
It does make sense.  The MOSFET will not conduct when VGS is 0V, therefore the source voltage will not be higher than the gate voltage.

Obviously there is some critical concept here I'm having a hard time grasping. 

Let's say I have a 5V power source, a resistor, and my 3V microcontroller pin.  If I connect the 5V source to the resistor and then the resistor to my microcontroller, I will destroy the microcontroller.

Now I do the same with a mosfet in the resistor's place.  What's changed?

We know the mosfet must be on and in a low resistance state, because otherwise the microcontroller could not pull the 5V side down when it's pin goes LOW.  The mosfet's resistance and the pullup form a voltage divider, and the drain pin is the point between them we tap and connect to he PFET.  For that point to be near 0V, the mosfet MUST be almost fully on, with a resistance of less than a few ohms.

And if the mosfet is almost fully on, then that means there's almost no resistance between that 5V pullup and the uC pin.  Which is the same as my example with just the resistor.  And the uC will be destroyed.

Except you say it won't.  That the source can never go above the gate.  That's fine, and I'd accept that, except it directly conflicts which what I just said about how the mosfet's resistance must be low, and therefore effectively a dead short between my uC pin and that 5V pullup... which would destroy the pin.

What am I missing here?


Quote
If that were true, I could not switch a 20V source using logic level mosfet because the gate would never see more than 3-5V.
Quote
The 20V switch is a common source circuit ("low side switch").  The level translator is a common drain circuit.  Totally different.  In fact, you can't easily use an N MOSFET as a 20V high side switch.

I don't understand the difference.  If I took that drawing and put my uC pin where the ground is, and set it low, it would look very much the same, and would the lamp not still light when 3.3V is applied at the gate?  That is, until the pin burnt out.


Quote
uC pin goes high, gate sees 5V, source is connected to 20V source powering lamp.  Source sees more than 5V.
Quote
Now this does not make sense.  In the 20V switch circuit, the source is grounded.

And it is grounded when I set my uC pin low as well, is it not?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf