Field-Effect Transistor (JFET)

Semiconductor Devices · JFET · Learn

The BJT is a current-controlled device — a small base current controls a large collector current. The JFET (Junction Field-Effect Transistor) is fundamentally different: it is voltage-controlled. The gate draws essentially zero current; it squeezes a conducting channel with an electric field. That high input impedance makes JFETs ideal for instrumentation amplifiers, RF front ends, and anywhere you cannot afford to load the signal source.

Construction

Figure 1. N-channel JFET cross-section. Two p-type gate regions sandwich the n-type silicon channel. Gate–channel junctions are reverse-biased; a more negative V_GS widens the depletion regions and narrows the channel.

An N-channel JFET consists of a bar of n-type silicon (the channel) with two p-type regions diffused into it from opposite sides to form the gate. The two ends of the channel are the drain (at higher potential) and the source (at lower potential). When V_GS = 0 V the depletion regions are thin, the channel is wide open, and the maximum drain current flows — this is called $I_{DSS}$ (drain-to-source current with the gate shorted to source).

Making V_GS more negative reverse-biases the gate–channel pn-junctions. The depletion regions grow inward, pinching the channel narrower and reducing $I_D$ . At the pinch-off voltage $V_P$ (a negative number, also called $V_{GS(off)}$ ) the two depletion regions meet and the channel is completely closed — $I_D = 0$ .

Transfer characteristic

Figure 2. Transfer curve:

I_D

V_{GS}

in saturation. I_DSS = 10 mA, V_P = −5 V. The parabolic relationship means a 2 V swing in V_GS produces a much larger change in I_D near V_GS = 0 than near pinch-off.

In the saturation region (where V_DS ≥ V_GS − V_P), the drain current follows Shockley's square-law:

I_D = I_{DSS}\left(1 - \frac{V_{GS}}{V_P}\right)^2

This parabola is the JFET's "personality" — a plot of $I_D$ vs $V_{GS}$ for a fixed $V_{DS}$ in saturation. Two device parameters fully characterise it: $I_{DSS}$ and $V_P$ . Typical values for a 2N5457 are I_DSS = 1–5 mA, V_P = −0.5 to −6 V (wide spread — always read the datasheet).

Gate current is essentially zero — the reverse-biased pn-junction blocks it. V_GS alone controls I_D, with no loading of the signal source.

Output curves and operating regions

Figure 3. Family of output curves (I_D vs V_DS) for V_GS = 0, −1, −2, −3, −4 V. The dashed knee locus separates the ohmic and saturation regions.

The output curves reveal two distinct regions:

Ohmic region (V_DS < V_GS − V_P): the channel is not yet pinched off. The JFET behaves like a voltage-controlled resistor — I_D rises steeply with V_DS. The on-resistance is roughly $r_{DS} \approx r_{DS(on)} / (1 - V_{GS}/V_P)^2$ .
Saturation region (V_DS ≥ V_GS − V_P): the channel is pinched at the drain end, and I_D is approximately constant at the value set by V_GS. This is the amplifier operating region.

The locus of knee points (dashed line in Fig. 3) is given by $V_{DS} = V_{GS} - V_P$ . Above this line lies saturation; below it is the ohmic region.

Transconductance g_m

The transconductance g_m measures how effectively the gate voltage controls the drain current:

g_m = \frac{\partial I_D}{\partial V_{GS}} = g_{m0}\left(1 - \frac{V_{GS}}{V_P}\right)

where $g_{m0} = -2I_{DSS}/V_P$ is the maximum transconductance at V_GS = 0. Note that g_m decreases as V_GS goes more negative — the JFET becomes less responsive near pinch-off. This is why JFET amplifiers are often biased near V_GS = 0 to maximise gain.

The small-signal voltage gain of a common-source amplifier is:

A_v = -g_m R_D

JFET vs BJT

JFETs and BJTs complement each other. JFETs win on input impedance (G_in > 10⁹ Ω, no gate current), noise (lower 1/f noise — preferred in audio front ends), and thermal stability (I_D has a negative temperature coefficient at the right bias — self-stabilising). BJTs win on transconductance per milliamp (g_m = I_C/V_T ≈ 40 mA/V per mA — much higher than JFETs), and on predictability (β varies 3×, but V_P can vary 10×).

In practice: use a JFET at the input of a low-noise amplifier or when the signal source cannot supply any current. Use a BJT when you need high gain from a low supply current, or when tight part-to-part consistency matters.

The JFET has three regions: ohmic (voltage-controlled resistor), saturation (constant I_D controlled by V_GS), and cutoff (V_GS ≤ V_P, channel fully pinched). Amplifiers operate in saturation; switches use cutoff ↔ ohmic.