MOSFET Resistance Problem

[BlackMelon]

Hello guys,

I'm checking my MOSFET cuz I want to know that it's dead or not..

I was checking it without a power supply but only a digital multi-meter. I've got an infinite drain-source resistance with the + probe on drain and the - probe on source but after I swapped the probes, I got a reading of 5.88 Megaohms

Is it uncommon to have different values of resistance when swapped those probes?
Is this the effect from the current from multimeter battery which is injected to the Drain-Source? If so, should the
drain-source and source-drain resistance be different from each other so much?

PS. I was trying to bias her with a power supply too. When it was biased with an 8V Gate-Source voltage, I got 0 ohms reading on drain-source resistance which is still the same after I swapped the probes.
When it was biased with 0V Gate-Source voltage, I got an infinite resistance with + probe on Drain and - probe on Source. After I swapped the probes, I got 5.879Megaohms.

I would be very grateful if you could provide me some theory background too

Thank you
BlackMelon

 

[kpatz]

You're seeing the effect of the MOSFET "body diode". The construction of the FET causes it to act as if a reverse biased diode exists across the source and drain. So, if you measure the FET in the "correct" polarity (drain-to-source voltage positive for N-channel, or negative for P-channel), you should see an infinite resistance when the transistor is off, but you'll get some conduction in the opposite direction.

If you use the diode-test function on your meter (which gives you the forward voltage drop of the diode), you'll see that it behaves similarly to a diode, showing about 0.6V when forward biased (or "backwards" in terms of the drain/source of the FET).

The fact that you got 0 ohms after charging the gate and a high resistance after discharging the gate, you know the FET is working properly.

 

[BlackMelon]

Thank you for the answer...
I still have sth. that I'm curious in MOSFET.

The "some conduction" which you've said on the reply is done by the digital multimeter battery, right?

And about biasing the gate-source voltage..

I've heard that the FET gate-source behaves like a capacitor... it's needed to be bias with 8V and it might need a current source to inject current to the gate pin in order to obtain a high switching rate

My question is how can I do this two job in the same time, same circuit?

BlackMelon

 

[duke37]

The mosfet looks to be OK but may not stay that way. Wih no voltage between the gate and emitter, the mosfet will be off so measres infinity resistance. There is however a diode inside the mosfet so which will pass current with the appropriate polarity. The 5.88M you measure is probably 0.588V diode drop.

The maximum gate source voltage is perhaps 20V. This can be greatly exceeded by static electricity. Pop goes the fet! It needs careful handling.

You have no need to worry about gate capacitance unless you are switching rapidly. You should drive with a voltage source, not a current source. The source should be able to provide enough current for the speed you select. There are gate driver chips or you can make a driver yourself. If you are running under say, 100kHz, you do not need to bother.

 

[kpatz]

When you use your multimeter to measure across drain and source, it's applying a voltage (from the meter's battery) and then measuring how much current flows. If no current flows, it shows an infinite resistance. If current flows, the "resistance" is calculated based on that. The body diode will allow some current to flow if the source is at a higher voltage than the drain (on an N-channel FET). If the FET is on (gate-source voltage higher than the threshold), more current will flow, in either direction (which would show a low resistance, 0 ohms ideally if the G-S voltage is high enough, on your multimeter).

The gate of a MOSFET doesn't draw any current, except during switching when it charges or discharges (it acts much like a capacitor in this regard). As a result, the gate capacitance combined with any series resistance/impedance in the circuit driving the gate acts as a low-pass filter that limits how fast the gate voltage will change when switching. The more current the circuit controlling the gate can supply, the faster the switching transition. This isn't a problem at low frequencies, but when you get into the kilohertz range (such as when controlling motors with PWM etc). it can become a factor since the FET will dissipate power (and heat) while it's in the linear region between full-on and full-off.

And like duke37 said, the gate is sensitive to ESD, so it's important to consider when handling MOSFETs. It's a good idea to store leaded parts in a piece of conductive foam, and when I build circuits using them, I include a gate-source resistor of say 100K or so, to ensure any gate charge is discharged to source. Wearing a grounding strap is ideal, but if you don't have one, touching something grounded before handling the FET works too. My hot-air soldering station has a metal case that's grounded so I'll touch that before handling ESD-sensitive parts.