N-channel mosfet Vgs(th) meter

This sketch measures gate threshold voltage of an N-channel mosfet with the help of a 2k2 or 10k resistor and a breadboard. It might be useful to test/select fets if they will work with Arduino's 5volt logic. For that, threshold voltage should be half of Arduino's VCC or less (~2.5volt or less). Post your comments and results. Leo..

/* Measures threshold voltage of an N-channel mosfet with an Uno or Mega
  Source connected to ground
  Gate and drain connected to A0
  2k2 (~1mA) or 10k (~250uA) pull up resistor between 5volt and A0
*/

void setup() {
  Serial.begin(9600);
  float voltage = (analogRead(A0) * 5.0) / 1024; // assumes VCC is 5.0volt
  Serial.print("Gate threshold is ");
  Serial.print(voltage);
  Serial.println(" volt");
  Serial.println("Press reset to measure again");
  Serial.println("");
}

void loop() {
  // nothing here
}

Edit: changed pull up to 2k2 for 1mA threshold current. Use 10k for 250uA threshold current.

P-channel mosfets can be measured by connecting gate and drain to ground and source to A0.

Interesting. How can you add in Rds ON? I figure some code brainiacs can come up with something to determine if RDS ON is reached within the 5V gate voltage. As has been written many times (by you as well), it's not the threshold, it's the ON state that matters when using them as a switch

This simple setup can only select non-logic level fets from logic level fets. I guess that for Rds(on) you need to send a significant current through the fet to be able to measure that low "ON" resistance. That would require an external supply. Threshold voltage has a wide range in the datasheets, "ON" has resistance not. Leo..

I never figured it would be easy.

I did just try it out with a FDP8880 on a pro-mini and it came up within spec--Vth 1.95V

Good job!

Nice idea - but a DMM would do as well.

Replace the 2k2 with a variable supply voltage , current meter and variable resistor and you have the beginnings of a curve tracer.

Or you could always read the datasheet.

regards

Allan.

allanhurst: Nice idea - but a DMM would do as well.

Replace the 2k2 with a variable supply voltage , current meter and variable resistor and you have the beginnings of a curve tracer.

Or you could always read the datasheet.

Do you mean a DMM on its own. Like measuring a diode. My DMM works with a measuring voltage of 3volt. Not enough for non-logic level fets. Some non-logic fets I measured were 3volt.

Was thinking of using digital output pins for different value pull up resistors. But hey, this is a simple go/nogo gadget.

Tested a batch of 2N7000 fets. They varied between 2.55 and 2.65 volt. Datasheet states 0.8-3volt, with 2.1volt as typical. This is a nice and easy way to make sure. Leo..

I take it you checked the arduino's Vdd was really 5.0v - as you know it ain't necessarily so.

But a neat trick all the same...

Allan.

Wawa:
This sketch measures gate threshold voltage of an N-channel mosfet with the help of a 2k2 or 10k resistor and a breadboard.
It might be useful to test/select fets if they will work with Arduino’s 5volt logic.
For that, threshold voltage should be half of Arduino’s VCC or less (~2.5volt or less).

Logic level devices have threshold of 0.5 to 1.0V, typically.

In general the gate drive should be substantially more than the plateau voltage, but measuring
the plateau is harder as it requires a 'scope typically. The plateau voltage indicates at what
gate-source voltage the bulk of the conducting channel is forming between source and drain - before that
point the channel exists, but has a significantly greater resistance.

As a rule of thumb I’d say ensure the Vthr is < 1/3rd of your gate drive voltage for a good chance
of switching high currents reliably.

Non-logic-level devices usually have Vthr specified as 2–4V (large manufacturing spread is always
the case with FET voltages). Even a device at the bottom of that range wants 6V drive for a good
chance of working well. And because of the tendancy for gate voltages to drift with time due
to ion migration you cannot assume a measurement at one point in time will hold for the device
lifetime…