[SOLVED] How many pins do a common mosfet have?

[Jasper Chow]

Hi, there.
I'm new here and a student in electronics. It seems many textbooks sate a mosfet usually has FOUR connnecting pionts to outside, i.e. gate, drain, source, and substrate(bulk), and for a nmos, the substrate is typically linked to the ground or the source. But as far as I've found out, a real mosfet has only three pins, gate, source, drain. Obviously, there is no substrate pin avaible to plant in the circuit. What's the problem? I am confused. What I am pretty sure is the substrate needn't to be connected to the circuit, cause a mosfet can work well if it is connected to the circuit like a bipolar, with only three terminals G, S, D, just like a biploar's B, E, C respectively.
Any opinion will be appreciated.

 

[akshat_gupta_06]

The substrate is the material on which the entire device (in this case, the MOSFET) is developed. For example, if you were a sculptor, and made a statue with arms and legs, your substrate would be stone or marble or any material you used. In this case too, the material is then given various functional properties to collectively behave alike a MOSFET.

Speaking of number of pins, there are only three; gate, source and drain. In case of a BJT, emitter, collector and base.
Hope that helps!

 

[Jasper Chow]

The substrate is the material on which the entire device (in this case, the MOSFET) is developed. For example, if you were a sculptor, and made a statue with arms and legs, your substrate would be stone or marble or any material you used. In this case too, the material is then given various functional properties to collectively behave alike a MOSFET.

Speaking of number of pins, there are only three; gate, source and drain. In case of a BJT, emitter, collector and base.
Hope that helps!

However, Razavi's Fundamatals of Microelctronics says the substrate CAN have a contact, which bewilders me.
here is the quote:
"6.2.6 Other Second-Order Effects
Body Effect In our study of MOSFETs, we have assumed that both the source and the
substrate (also called the “bulk” or the “body”) are tied to ground. However, this condition
need not hold in all circuits. For example, if the source terminal rises to a positive voltage
while the substrate is at zero, then the source-substrate junction remains reverse-biased
and the device still operates properly."

 

[FvM]

If you are referring to discrete components, the substrate connection is clearly specified in transistor datasheets (bound to source), I don't understand your problem. A few small signal MOSFETs have a separate substrate pin.

In IC design, substrate connection is mostly fixed by the technology.

 

[Jasper Chow]

If you are referring to discrete components, the substrate connection is clearly specified in transistor datasheets (bound to source), I don't understand your problem. A few small signal MOSFETs have a separate substrate pin.

In IC design, substrate connection is mostly fixed by the technology.

you mean most mosfets we can see are cased with a typical configuration that directly connect the souce and substrate, so what we can handle from outside is G, S ,D?

 

[crutschow]

you mean most mosfets we can see are cased with a typical configuration that directly connect the souce and substrate, so what we can handle from outside is G, S ,D?

That is correct.

 

[timof]

Hi, there.
I'm new here and a student in electronics. It seems many textbooks sate a mosfet usually has FOUR connnecting pionts to outside, i.e. gate, drain, source, and substrate(bulk), and for a nmos, the substrate is typically linked to the ground or the source. But as far as I've found out, a real mosfet has only three pins, gate, source, drain. Obviously, there is no substrate pin avaible to plant in the circuit. What's the problem? I am confused. What I am pretty sure is the substrate needn't to be connected to the circuit, cause a mosfet can work well if it is connected to the circuit like a bipolar, with only three terminals G, S, D, just like a biploar's B, E, C respectively.
Any opinion will be appreciated.

A "classical" MOSFET has four terminals, as you mentioned - source, drain, gate, and bulk/body/substrate.

Changing substrate voltage does change the transistor characteristics.
The most notable of these effects is the dependence of Vt (threshold voltage ) on Vbs - source/body voltage.

Very often, substrate voltage is held at the same voltage as source - in which case, you can say that MOSFET has three terminals, and refer to this device (SPICE model, etc.) as three-terminal device.

In some technologies, there is no contact to the substrate/body, and the body is electrically isolated - said to be floating - for example, in SOI (silicon-on-insulator) process.

In some other technologies (for example, in BCD - bipolar-CMOS-DMOS, also known as power management technologies), MOSFET might be placed into one, two, or more wells - and often contacts to these wells are declared as transistor terminals, so that MOSFETs can have 5, 6, or more terminals.

So, the physical structure and operation of the MOSFET define how many terminals the device have - and it can be anywhere form two (for example, when MOS transistor is configures as MOS capacitor), to many (many may be up to 6-7), with typical number being 4 or 3.

Max
-------