linearity of mosfets

Status
Not open for further replies.

Thunderchild

New Member
I was just curious and did a little setup with a mosfet and a potentiometer checking the resistance of the drain/source channel at different gate voltages, hm interesting I only did a rough check and could do a proper setup and take down v/R details and plot them on a graph but usually are mosfets "linear" ? are their carachteristics repeatable or are they quire variable like BJT's (for example the 2N2222 will from from a beta of 50 to 300 !)
 
Above the threshold voltage the drain-source resistance is a non-linear function of gate voltage. See **broken link removed** for more info. It refers to J-FETs, but the same is generally true for MOSFETs. Edit: The reference does show a technique for making the resistance linear.

This will also vary between transistors, depending upon both the threshold voltage and gain of a particular transistor. Thus ii you want to use a MOSFET as a variable resistor, it is usually in some sort of circuit where the effect of the variation is minimized (by having it in a negative feedback loop for example).
 
Last edited:
thats interesting, so basically I put a resistor in series with the gate and a resistor from gate to drain of the same value and I get a linear V/R carachteristics ? but the two resistors must be sufficiently high to not interfere with the circuit and the desired resistance of the source/drain path
 
True. But since a MOSFET has a very high input impedance, you can use large values of resistors without affecting the transistor operation.
 
1M is probably OK. Look at the gate leakage value in the spec sheet and determine the maximum drop across the resistor from that. You generally don't want the drop to be more than a few mV.
 
so do mosfets/fets of the ame model number have pretty much identical properties or do they vary like in BJT's ?
 
They vary like BJT's but probably not as severely, also the beta of a transistor doesn't have anything to do with it's linearity, just it's gain. I think the equivalent to beta in a Mosfet is transconductance, so you should be able to look up the max/min and nominal values if the datasheet for your fet lists it.
 
Last edited:
right, well for a BJT if I were to create a current source in a very simple way the current could change by up to 5 times if I were to change the BJT, so basically for FETs/MOSFETs the resistance that coresponds to a certain gate voltage will differ from part to part ?
 
so what does it translate to in practice ? the current consumed by the gate ? although it in theory nil
The gate itself looks like an oxide dielectric capacitor, so it draws virtually no current. But most MOSFETs gates are protected by diodes, and it is the diode leakage that is the predominate "gate" current. So that leakage current will generate some voltage drop across any gate resistance. You just want to keep the drop low (probably no more than a few mV for typical applications) so that determines the maximum gate resistance you can have.
 
right I see so basically it will create a potential divider with the input resistor and cause unpredictable behaviour
 
Thunderchild, no circuit should EVER depend on the specific beta of a transistor, this is a sometimes touched upon unwritten law of electronics. Use more gain than you need and use negative feedback to cancel out what you don't need. With temperature and possible voltage variations beta can vary dramatially, let alone static manufactoring differences.
 
Well my question is really quite hypothetical at the moment, I was just curious about the use of a mosfet as a voltage controlled resistor. If I were to build an amp myself I'd go for a class D using the losfet as a switch
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…