FET

• How FET got its Name
• JFET (Junction FET)
  • I-V Characteristics
    • Questions
  • Pinch-off voltage
  • Channel Current
  • Gain (Transconductance)
• MESFET (Metal Semiconductor FET)
  • Metal-Semiconductor (MS) Contact
  • Schottky Diode
  • Heavily Doped (Degenerately Doped) Semiconductor
  • Schottky Diode Calculations
• MOSFET (Metal-Oxide FET)
  • Composition
  • Terminal Naming
  • Functionality (NMOS)
  • MOS Energy Band Diagram
  • MOS Calculations
    • KaTeX parse error: Undefined control sequence: \PhiS at position 1: \̲P̲h̲i̲S̲: Surface Potential
    • KaTeX parse error: Undefined control sequence: \PhiF at position 1: \̲P̲h̲i̲F̲
    • Depletion Width $W$
    • Threshold Voltage $VT$
    • Cap-Voltage Characteristics
  • Supplement
  • JFET pinch-off
    • Why doesn’t current goto 0
    • How does current flow after pinch-off
  • Gradual Channel Approx
  • JFET Transconductance

How FET got its Name

A voltage applied to the metallic plate modalated the conductance of the underlying semiconductor, which in turn modulated the carrent flowing between ohmic contacts A and B. This phenomenon, where the conductivity of a semiconductor is modulated by an electric field applied normal to the surface of the semiconductor, has been named the field effect.

JFET (Junction FET)

Suppose we connect S to ground, reverse bias at S is $V_G$.

I-V Characteristics

1. $V_G$ = 0. When VD is small, ID is small. Linear I-V. No change in depletion width across channel.
   
   
2. $V_G$ increases. Channel pinches-off.
   
   
   

Questions

1. Should/Should not $I_D$ = 0 beyond pinch-off?
   No. Carriers can also pass depletion region, but see a much higher resisitance.
2. Why does $V_D$ > $V_D^{Sat}$ have no effect on ID?
   Increasing $V_D$ will also increase the length of pinched-off region. These two effects cancel out.

Pinch-off voltage

$V_p$ = Reverse bias between n-channel and $p^{+}$ gate at the drain end $(x=0)$.
$h(x)$ = Channel half-width at any $x$
$a$ = half width of channel
Assumptions:

1. Channel with at $x=0$ decreases uniformly as the reverse bias increases to pinch-off.
2. $V_{bi}$ neglected.
3. $p^{+}-n$ gate junction.

$$V_p=\frac{q{a}^2{N}_D}{2\epsilon }$$
$a$: half thickness

Channel Current

$L$: length
$Z$: depth
$2a$: thickness

$$I_D=G_o{V}_P\left[\frac{V_D}{V_P}+\frac{2}{3}{\left(-\frac{V_G}{V_P}\right)}^{3/2}-\frac{2}{3}{\left(\frac{V_D-V_G}{V_P}\right)}^{3/2}\right]$$
$$G_o=\frac{2aZ}{\rho L}$$
$$I_D(sat)=G_o{V}_P\left[\frac{V_D}{V_P}+\frac{2}{3}{\left(-\frac{V_G}{V_P}\right)}^{3/2}-\frac{2}{3}\right]$$

Gain (Transconductance)

$$g_m(sat)=\frac{\partial I_D(sat)}{\partial V_G}=G_o\left[1-{\left(-\frac{V_G}{V_P}\right)}^{1/2}\right]$$

MESFET (Metal Semiconductor FET)

Metal-Semiconductor (MS) Contact

• ${\Phi }_M$: Metal Work Function (the one in photoelectric effect)
• ${\Phi }_S$: Semiconductor Work Function
• $\chi$: Electron Affinity. $\chi =(E_0-E_C){|}_{surface}$

$${\Phi }_S=\chi +(E_c-E_F{)}_{FB}$$

• $(E_c-E_F{)}_{FB}$: Energy difference between $E_c$ and $E_F$ at flat band (i.e.) zero bias condition.

$${\Phi }_B={\Phi }_M-\chi$$

• ${\Phi }_B$: surface potential-energy barrier encountered by electrons with $E=E_F$ in the metal.

Schottky Diode

• ${\Phi }_M>{\Phi }_S$: Applying $V_A>0$ lowers $E_{FM}$ below $E_{FS}$, reduces the barrier seen by electrons in the semiconductor.

• ${\Phi }_M<{\Phi }_S$: Non-rectifying, Ohmic.

Heavily Doped (Degenerately Doped) Semiconductor

When the barrier is thin enough, the carriers can tunnel through.
Upper: forward bias. Below: rev bias
Additional component of current

Schottky Diode Calculations

• Built-in Voltage

$$V_{bi}=\frac{1}{q}\left[{\Phi }_B-(E_c-E_F{)}_{FB}\right]$$

• $\rho$
  • Metal: delta function (charge only on surface)
  • Semiconductor: $\rho =q{N}_D$
• $E$:

$$E(x)=-\frac{q{N}_D}{{\epsilon }_{Si}}(W-x)\dots 0\le x\le W$$

• $V$:

$$V(x)=-\frac{q{N}_D}{2{\epsilon }_{Si}}(W-x{)}^2\dots 0\le x\le W$$

• Depletion Width

$$W=\sqrt{\frac{2{\epsilon }_{Si}}{q{N}_D}(V_{bi}-V_A)}$$

• $\Phi (x)$

$$\Phi (x)=\frac{q{N}_D{x}^2}{2{ϵ}_S}$$

• Current Density

$$J=J_S(e^{q{V}_a/kT}-1)$$
$$J_S=A^{\ast }{T}^2\exp \left(-\frac{q{\Phi }_B}{kT}\right)$$
$A^{\ast }$ is the Effective Richardson Constant

MOSFET (Metal-Oxide FET)

Composition

• MOS Capacitor
• Two pn juncitons

Terminal Naming

• Carriers enter the structure through Source (S)
• Leave through the Drain (D)
• Subject to the control of the Gate (G)

Functionality (NMOS)

• When $V_G\le V_T$, i.e. $V_G$ is in accumulation or depletion biased, the gated region contains mostly holes and few electrons, an open circuit is formed.
• When $V_G>V_T$, i.e. $V_G$ is inversion biased, an inversion layer (channel) containing mobile electrons is formed.
• As $V_D$ increases, the channel finally pinches-off, the current saturates.

MOS Energy Band Diagram

• Accumulation ($V_G<0$) holes accumulate on the semiconductor side of the gate
• Depletion ($0<V_G<V_T$) holes repelled away, leaving Ionized acceptors atoms
• Inversion ($V_G>V_T$) electron density increase
  • Initially, $n<n_i$.
  • $n=n_i$ when $E_i=E_F$
• When $V_G=V_T$, $n=N_A$, the semiconductor seems no longer to be depleted. Instead, it now behave similar to n-type. The channel has formed.

MOS Calculations

${\Phi }_S$: Surface Potential

$${\Phi }_S=\frac{1}{q}[E_i(bulk)-E_i(surface)]$$

${\Phi }_F$

$${\Phi }_F=\frac{1}{q}[E_i(bulk)-E_F]$$
In p-type, $N_A\gg N_D$, $p_{bulk}=n_i\exp ([E_i(bulk)-E_F]/kT)=N_A$
$${\Phi }_F=\frac{kT}{q}\ln \left(\frac{N_A}{n_i}\right)$$
In n-type, $N_D\gg N_A$, $n_{bulk}=n_i\exp ([E_F-E_i(bulk)])=N_D$
$${\Phi }_F=-\frac{kT}{q}\ln \left(\frac{N_D}{n_i}\right)$$
When $V_G=V_T$,
$${\Phi }_S=2{\Phi }_F$$

Depletion Width $W$

Valid before strong inversion:

$$W=\sqrt{\frac{2{\epsilon }_{Si}}{q{N}_A}{\Phi }_S}$$
At strong inversion:

$$W_m=2\sqrt{\frac{ϵkT}{q^2{N}_A}\ln \left(\frac{N_A}{n_i}\right)}$$
When $V_G=V_T$, ${\Phi }_S=2{\Phi }_F$, the depletion width
$$W_T=\sqrt{\frac{4{\epsilon }_{Si}}{q{N}_A}{\Phi }_F}$$

Threshold Voltage $V_T$

For N(-channel)MOS (P-bulk)

$$V_T=2{\Phi }_F+\frac{{ϵ}_{OXIDE}{x}_O}{{ϵ}_{Si}}\sqrt{\frac{4q{N}_A}{{ϵ}_{Si}}{\Phi }_F}$$
For PMOS (N-bulk)
$$V_T=2{\Phi }_F-\frac{{ϵ}_{OXIDE}{x}_O}{{ϵ}_{Si}}\sqrt{\frac{4q{N}_D}{{ϵ}_{Si}}(-{\Phi }_F)}$$
$x_O$ is the thickness of the OXIDE

Cap-Voltage Characteristics

Supplement

JFET pinch-off

Widening everywhere as $V_{SD}$ grows

Why doesn’t current goto 0

If current is 0, the pinch-off will disappear. To maintain pinch-off, a non-zero current must be present.

How does current flow after pinch-off

Gradual Channel Approx

Formula for depetion layer width remainis same at every point and edge of depletion layer is not a multi-valued function at any point

JFET Transconductance

$$g_m=\frac{\partial I_D(sat)}{\partial V_G}$$
Proportional to $\sqrt{V_G}$
e at every point and edge of depletion layer is not a multi-valued function at any point

JFET Transconductance

$$g_m=\frac{\partial I_D(sat)}{\partial V_G}$$
Proportional to $\sqrt{V_G}$