Calculate resistor from uC pin to MOSFET gate, and saturation question

Discussion in 'General Electronics Chat' started by ke5nnt, Apr 22, 2011.

  1. ke5nnt

    Thread Starter Active Member

    Mar 1, 2009
    I'd like to understand a couple of things better than I do regarding MOSFETs. A lot of my projects use a MOSFET as a switching device driven from the output of a microcontroller, so the MOSFET would be logic level.

    A resistor is ideal between the uC and the Gate of the MOSFET to limit current out of the uC. In calculating a value for this resistor, I believe the formula for doing that is Vcc @ 5V, Vgs, and uC pin current @20mA max... so 5V - Vgs = (5V-Vgs)/0.02 = Resistance.

    This brings 2 question to mind for me, which may seem like simple questions but I need clarification regardless. As an example, I'm using this datasheet.

    Figure 3 shows typical transfer characteristics. Using the 25°C plot line, my drain current is about 150mA. Does this mean that it takes about 2.2V into the gate for the MOSFET to switch on? Also, I hear the term "saturation" a lot. Is saturation considered the point where the MOSFET starts to conduct from Drain to Source, which if the above is correct would be when the gate reaches 2.2V? Maybe I'm way off on this.

    I'm not the best at picking information out of datasheets. I'd like to have a better understanding of what voltage into the gate starts conduction from Drain to Source, and also how to determine what current into the gate from the uC pin is required, which I believe is called HFE. I believe HFE is determined by how much the load current into the Drain is, but perhaps I'm mistaken?

  2. Jony130

    AAC Fanatic!

    Feb 17, 2009
    I think that you need "back to basics".
    When you driven the constant load (DC load) you don't need any resistor in series with the gate.
    And the HFE is for BJT not for MOSFET.
    MOSFET start conduction when Vgs> Vgs(th)
    • 112.PNG
      File size:
      6.2 KB
    Last edited: Apr 22, 2011
  3. ke5nnt

    Thread Starter Active Member

    Mar 1, 2009
    Ok, thanks for the link. Here's what I'm understanding, please point me in the right direction if I'm heading off course...

    See this

    You are saying that I do not need a resistor in series with the gate from the uC output. I don't understand why? Is it because of the uC resistance that would prevent any excess current surge into the gate, and also prevent accidental reverse voltage into the pin?

    Would a resistor still be appropriate from gate to source to allow the discharge of capacitance that builds up in the MOSFET? I have typically used both a current limiting resistor in series with the gate, and a discharge resistor from gate to ground in projects like this, and have never had a problem. However, if I'm able to eliminate unnecessary components from my designs, all the better (saves room on the board right?)

    So, ~5V from the uC output into the gate is greater than the 1-3V Vth, therefore, the MOSFET is saturated (on) and is conducting from source to drain. That's what I'm understanding. Also, looking at this datasheet at the plot for Transfer Characteristics (page 4, figure 8), I'm gathering that with a Drain Current of 150mA (my load), that my Vth will be roughly 2.6V.

    Ok, let me know if I screwed anything up :rolleyes:
  4. ronv

    AAC Fanatic!

    Nov 12, 2008
    The 10K in your schematic will make sure the FET is off if the input is disconnected or if the output pin is made into an input pin.
    The 270 ohm is to limit the current for your micro. Without it the peak current will be as high as the output of the micro will support. If you switch at a high frequency you might exceed the micros spec as you charge and discharge the gate capacitance. It also helps to damp any oscillations due to fast switching and inductance in the layout.

    The threshold voltage is the voltage where the FET just begins to turn on. The curves are typical values and will vary. It is best to operate well above it. If you look early in your data sheet you will see it can vary from 1 to 3 volts for only 250 ua.
  5. ke5nnt

    Thread Starter Active Member

    Mar 1, 2009
    Ok, so the threshold voltage is 1-3V, which I saw in the datasheet. You mention operating well above it. How do you ensure you're doing this?

    Output from the uC pin should be roughly Vss, or 5V. As far as the test conditions for Vth, where you reference 1-3V at 250uA, is the 250uA the Gate-Source current, or Drain-Source current?

    Regarding the 270Ω resistor there, there was a formula I used a while back that someone had explained to me once as being (I believe) (Vss-Vth)/uC pin max current (or 20mA). If I were to use the 2.6V value I'd gotten from that plot as Vth, I would get (5-2.6)/0.02 = 120Ohms. Do you agree with my assessment?
  6. Jony130

    AAC Fanatic!

    Feb 17, 2009
    You really don't need any resistor for MOSFET switch for DC and low frequency signals. For high frequency you need a driver circuit because the MOSFET gates have considerable capacitance, which need to be quickly charge / discharge.
    The output pin of the PIC has MOSFET inverter.
    So for sure there is a low resistance path for MOSFET charge and discharge current.
    Sometimes we add a pull-down resistor so that the MOSFET is held off while the PIC is in reset.
    Of course this is a drain current (Drain-Source current)
    The Gate current flow only in very limited amount of time (T_on/ T_off times). When the MOSFET capacitor is charging or discharging.
    Last edited: Apr 23, 2011
  7. ke5nnt

    Thread Starter Active Member

    Mar 1, 2009
    So you are suggesting I use neither the resistor from the uC pin to the gate, nor the 10K resistor from gate to source? See .pdf attachment.

    I feel like this question was unanswered. I'd still like to know about this. With a 5V Vss from uC, I will be well above Vth?

    Thank you.
  8. Jony130

    AAC Fanatic!

    Feb 17, 2009
    Yes, you don't need any resistors. You can connect the MOSFET gate directly to the output pin of the uC

    I don't see any problem with that.
    Are you worried that Vth ( Vgs(th) ) is < then Vdd??
  9. SgtWookie


    Jul 17, 2007
    I'll suggest that keeping SOME resistance between the uC output pin and the MOSFET gate is a good idea, if only to keep the gate from "ringing" due to the C of the gate and the L of the wiring. You'd probably be OK with anywhere from 100 Ohms to 270 Ohms resistance.
    Keep the 10k resistor from the gates to ground. That will also help prevent damage to the MOSFETs when you're assembling the board - if you install the resistors before the MOSFETs. It will also keep the MOSFETs off if you remove the uC for service/troubleshooting.

    Ignore the threshold voltage. Look at the Rds(on) specifications for resistance, Id and Vgs.
    When Vgs=4v, Id=0.85A, Rds(on) will be a max of 0.28 Ohms - so you need Vgs to be >= 4v.
  10. ifixit

    Distinguished Member

    Nov 20, 2008
    Yes. Refer to Fig 12. This shows you how the Vgs is effected by temperature. Even at -55°C the Vgs is only 3.45V worst case. At high temperatures the Vgs is lower, 2V at +160°C.

    Refer to Figure 8. With a Vgs of 5V and Id of 150mA you'll be driving the socks of your load. When a increase in gate voltage doesn't yield an increase in Id then it is saturated.

    Leave the 10K in for reasons allready given.

  11. ke5nnt

    Thread Starter Active Member

    Mar 1, 2009
    Ok, thank you. I thought I'd remembered you telling me that at some point in the past. I assume you get this 100-270 range from somewhere? What would be the factors at play in determining this range, I'd like to know so I can make a note of it.

    Also, it may be in the datasheet and I'm just missing it, but is there a way to determine what current from the uC pin into the Gate will be under a certain Drain-Source load?

    Thank you for the help. I knew I'd read in more than one place that a resistor from gate to source was a good idea, that's why I always designed one there.