MOSFET ohmic region, needs to be explained

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
Can anyone explain to me how to understand the ohmic region of a MOSFET from its datasheet, or any formula to calculate the ohmic region according to the MOSFET datasheet, let's say IRFZ44n vs IRF3710, which one has a bigger ohmic region and how to analyze them accordingly?
I am not an expert please explain to me as simply as possible.
 

Attachments

BobTPH

Joined Jun 5, 2013
8,804
Look at fig. 1. In the IRF3710 and fig. 5. In the IRFZ44N datasheets.

See how the curves look like a straight line going up starting at 0,0 on the left, then they turn downward and become more or less flat?

The part on the left is the Ohmic region. In that region, it acts like a resistor, with its resistance called the on resistance.

Why are you interested in a wide Ohmic region? You generally do not vary the operation in that region, you supply enough gate voltage to allow the current you need with too high a Vds

Bob
 

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
Look at fig. 1. In the IRF3710 and fig. 5. In the IRFZ44N datasheets.

See how the curves look like a straight line going up starting at 0,0 on the left, then they turn downward and become more or less flat?

The part on the left is the Ohmic region. In that region, it acts like a resistor, with its resistance called the on resistance.

Why are you interested in a wide Ohmic region? You generally do not vary the operation in that region, you supply enough gate voltage to allow the current you need with too high a Vds

Bob
I want to make a variable resistor as a hobby, by tuning the pwm duty cycle, the amount of resistance changes, please take a look at the picture, can you explain me what is happening in this circuit and why it is not as what Eric send me as an equation.
why there is high amount of current passing through MOSFET while the gate voltage is too far from the threshold voltage?
123.jpg
 
Last edited:

jpanhalt

Joined Jan 18, 2008
11,087
If you want to mimic a voltage controlled variable resistor, I would not send the PWM directly to the gate. That will produce a pulsing output. Instead, pass the pwm through a low-pass filter (integrator) to give a variable DC voltage. Then apply that voltage to the gate. For the IRF3710, it looks like a gate voltage of 3.5V to 5V will be in the linear/ohmic region. For the IRFZ44, it looks like 5V to 8V will do the same. Again, that is pure DC, not pulses. A potentiometer as a voltage divider may be easier to experiment with. Be sure to have a common ground.
 

BobTPH

Joined Jun 5, 2013
8,804
The resistance of power MOSFETs is in the mOhm region, not likely the range of resistance you want, JFETs would be more suitable for this.

Bob
 

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
If you want to mimic a voltage controlled variable resistor, I would not send the PWM directly to the gate. That will produce a pulsing output. Instead, pass the pwm through a low-pass filter (integrator) to give a variable DC voltage. Then apply that voltage to the gate. For the IRF3710, it looks like a gate voltage of 3.5V to 5V will be in the linear/ohmic region. For the IRFZ44, it looks like 5V to 8V will do the same. Again, that is pure DC, not pulses. A potentiometer as a voltage divider may be easier to experiment with. Be sure to have a common ground.
thanks for the advice, but why not pulses? if the frequency is high enough then the devices connected to the MOSFET never get realized it.
 

jpanhalt

Joined Jan 18, 2008
11,087
The equivalent on resistance drops quickly once one exceeds the Vgs(th). For the IRF3710, Fig.1:
Vgs = 3.5V
Vds = 0.3V
Ids = 0.5A

Equivalent Rds = 0.3V/0.5A = 0.6 Ω
 
Last edited:

BobTPH

Joined Jun 5, 2013
8,804
That is not usunf
thanks for the advice, but why not pulses? if the frequency is high enough then the devices connected to the MOSFET never get realized it.
That is not using the MOSFET as a variable resistor. That is PWM to control the power supplied to a device, which is used all the time for certain devices, including motors and lights.

Power MOSFETs are the appropriate device for that. And the width of the ohmic region is not something I have ever heard enter the discussion. You simply need a MOSFET that can handle the max current with a low enough voltage drop and you need to supply the appropriate gate voltage to turn it on.

Bob
 

jpanhalt

Joined Jan 18, 2008
11,087
1) Maybe your observations (post 9) reflect the average current rather than the peak on current. That is, the effect of PWM and meter response speed.

2) Whether it makes a difference depends on what is being powered. If you are dimming an LED at a fairly high PWM frequency, it won't matter. If your PWM frequency is 10 Hz to 20 Hz, it will drive most people crazy.

3) While maximum current is often related to heat (i.e., many components will survive short excesses in current, but not a steady current of the same value), that doesn't hold true for all.

4) I agree in general with what you say about frequency. However, consider an inductive load being turned on and off. That creates spikes and EMI. Pass the same average current as a steady current and that problem doesn't exist.
 
Last edited:

crutschow

Joined Mar 14, 2008
34,280
The high transconductance of a MOSFET means that, once you reach the MOSFET Vgs threshold voltage it takes only a very small voltage change above that to cause a rapid change in the MOSFET resistance.

This is shown below in the LTspice simulation of a typical small MOSFET:
As can be seen, the equivalent MOSFET drain-source resistance (green trace) varies rapidly for a small change in the gate-source voltage (yellow trace, horizontal axis).

1612796837331.png
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,672
IRFxxx Mosfets need a gate-source voltage of 10V to fully turn on. You are using 4V pulses and the datasheets for IRFxxx Mosfets show that some of them conduct only 0.25mA (look at Threshold Voltage).

You need IRLxxx Mosets that are fully turned on when the gate-source voltage is 5V.
 

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
IRFxxx Mosfets need a gate-source voltage of 10V to fully turn on. You are using 4V pulses and the datasheets for IRFxxx Mosfets show that some of them conduct only 0.25mA (look at Threshold Voltage).

You need IRLxxx Mosets that are fully turned on when the gate-source voltage is 5V.
Actually, I don't need more than 500mA in my project, but thanks for the tip anyway, something very interesting about this topic I found out today is the temperature dependence of the Ohmic region, when the MOSFET gets hot, the resistance dramatically decrease.
 

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
If you want to mimic a voltage controlled variable resistor, I would not send the PWM directly to the gate. That will produce a pulsing output. Instead, pass the pwm through a low-pass filter (integrator) to give a variable DC voltage. Then apply that voltage to the gate. For the IRF3710, it looks like a gate voltage of 3.5V to 5V will be in the linear/ohmic region. For the IRFZ44, it looks like 5V to 8V will do the same. Again, that is pure DC, not pulses. A potentiometer as a voltage divider may be easier to experiment with. Be sure to have a common ground.
I used a low-pass filter and it didn't work as I want it to be, it seems the duty cycle of the PWM on the Gate is exactly the thing that makes it act like a resistor, thanks anyway for the information
 

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
Look at fig. 1. In the IRF3710 and fig. 5. In the IRFZ44N datasheets.

See how the curves look like a straight line going up starting at 0,0 on the left, then they turn downward and become more or less flat?

The part on the left is the Ohmic region. In that region, it acts as a resistor, with its resistance called the on-resistance.

Bob
in those Figs, after the curve turns almost flat line then with increasing in the Vds, why the current (Id) doesn't increase though? so Rds also increases with increasing the Vds? is that correct?
but before the curve turns almost flat, the ohmic region I mean, the increase in the Vds results increase in the Id with a constant slope so it means the resistance of the MOSFET in that region is almost constant, like 650mOhm for 3710 accordingly, but my experience is a variable resistance with a change in the duty cycle of the Gate!!! it is not the ohmic region though!
any idea?
 

BobTPH

Joined Jun 5, 2013
8,804
You need to get it out of your head that the MOSFET used in PWM control is acting like a variable resistor. It is not. It controls the power to load by controlling the percentage of time that the load is powered.

The resistance of the MOSFET has nothing to do with it. Ideally it would have zero resistance and still work as a PWM controller.

Bob
 
Top