1
\$\begingroup\$

I keep seeing different and contradictory explanations of how MOSFETs are supposed to operate when they are P/N channel and depletion/enhancement mode.

When it comes to the operation of a MOSFET, aren't these contradictory? Online it's said that an N-channel MOSFET conducts when there is a positive voltage at the gate, but also that when a MOSFET is in depletion mode then it conducts when there is no voltage on the gate and becomes "open" when a positive voltage is applied.

How can a MOSFET be both N-channel and depletion mode then? Or indeed a P-channel be enhancement mode?

\$\endgroup\$

2 Answers 2

2
\$\begingroup\$

This is often discussed in a confusing way, sadly. It's actually very simple.

Enhancement vs. depletion is just a shift of Vgs. Enhancement (for N channel) is positive, i.e. at Vgs = 0 it's off, and Vgs > Vgs(th) it's on.

You can make your own depletion MOS by putting a battery in series with the gate of an enhancement type, or vice versa.

The actual device characteristics vary a little bit, because they don't make them with batteries, there's a physical difference in construction and materials. But it has the same general effect.

As for N vs. P, simply take everything that N-channel does, and reverse it. Just insert a minus sign everywhere. Flip the diodes in your circuit, flip the supplies, and, meter readings will be negated. Now the gate needs negative voltage to turn more on, and the body diode conducts for positive drain voltage (Vds). Or if you prefer positive readings, swap the order: Vsd, Vsg, etc.

It probably doesn't help that datasheets aren't very consistent on this. Some P-ch. datasheets are written entirely in negative values, which is correct when labeled in the usual way (i.e. "Vgs"). Some have a note indicating all signs are implied, so just don't worry about it it's fine. Others omit that note and you're left to assume by yourself!

Even worse, a lot of datasheets confuse "greater" by magnitude versus by sign. If a parameter ranges from -20 to -10, which extreme value is "min" and which is "max"? You will see datasheets written both ways!

Note that "off" and "on" are not absolute, but are measured at some margin of current flow. Usually "off" is microamperes or less, and "on" is maybe 1mA or more. Exact values are given by Idss (or Id(off)) and Vgs(th) (or Vgs(off)) test conditions in the datasheet.

MOSFETs are never non-inverting. That is, as a common-source linear amplifier, the direction of voltage change on gate and drain will always be opposite. You (or one of the sources you were reading) have confused P-depl. with N type, it seems.

I never liked the usual descriptions, myself, because the phrase "no voltage" is disingenuous. It's not that the voltage doesn't exist (which can happen in mathematical terms, and rarely in certain systems; but certainly not here: there are definitely two terminals and a voltage between them at all times!), it's that it is [approximately] zero. They don't mean "no" voltage, they mean they are applying a voltage of 0V.

Put another way: it's sort of a nonlinearity due to figures of speech. There might be circuit differences in how we apply a voltage (say by leaving the terminals open, or shunting them with a resistor, or by touching a battery to the leads one way around or the other), and often we describe things we are doing -- which is important by itself. But those differences are irrelevant here, so long as the voltage on the device is what we say it is. So, it often happens that we say what we're doing (descriptive), but we mean to convey its effect (prescriptive). And often that difference isn't well defined, and is easily misinterpreted.

Also, to be a little pedantic, voltages near 0V are never perfectly zero, anyway. There is always thermal noise, both at the terminals themselves (even when shorted with a low resistance, some ~nV of noise remains), and within the device (which has higher resistance: RG "gate spreading resistance"). Usually when we say "zero", we mean the expected value, or within measurement error.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ "It's actually very simple. <wall of text follows>" 😉 \$\endgroup\$
    – tobalt
    Jun 9, 2023 at 18:48
  • \$\begingroup\$ Yeah, I know... The basic idea is simple, and contained in the first four paragraphs. The rest are to be very precise about that, and to hopefully counter confusion from those poorly-written sources or the interpretation of them. \$\endgroup\$ Jun 9, 2023 at 18:53
1
\$\begingroup\$

P or N channel in FETs tell which direction you have to charge the gate to turn the channel more on or more off.

  • Gate more positive to turn N-channels more on
  • Gate more negative to turn P-channels more on

Enhancement is the type whose channel is off when there is no voltage between gate and channel/body.

Depletion-mode FETs have a channel that is already a bit on with 0 V between gate and channel.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.