BJT as a switch — Hardware notes

WHAT is switching mode?

The load is placed between the supply $V_{CC}$ and the collector; the emitter goes to ground (common-emitter configuration).

HOW to derive the design equations (from scratch)

We want a guaranteed hard ON. Let's build the equations from Kirchhoff's laws, not memorize them.

Step 1 — Collector current when ON

When ON, the transistor pulls the collector node down to $V_{CE(sat)}$ . Apply KVL around the collector loop (supply → load resistor $R_L$ → collector → emitter → ground):

$V_{CC} = I_C R_L + V_{CE}$

Why this step? It's just voltage adding up around a loop. Solving for the current that flows through the load:

$I_{C(\text{on})} = \frac{V_{CC} - V_{CE(sat)}}{R_L}$

This is the current we need the transistor to sink.

Step 2 — How much base current does that need?

In the active region, $I_C = \beta I_B$ . But at the edge of saturation, the base current needed is exactly:

$I_{B(\text{edge})} = \frac{I_{C(\text{on})}}{\beta}$

Why this step? $\beta$ (current gain, also written $h_{FE}$ ) tells you how much collector current one unit of base current can support. If we supply exactly this, we're teetering on the boundary — risky.

Step 3 — Overdrive for a solid switch

$\beta$ varies wildly with temperature and part-to-part (a " $\beta=100$ " transistor might really be 50). To be safe we deliberately push in more base current than the minimum:

$I_B = k \cdot I_{B(\text{edge})}, \quad k \approx 2\text{–}10 \; (\text{overdrive factor})$

Why this step? Extra base current forces the transistor deep into saturation, keeping $V_{CE(sat)}$ low and the switch reliably closed even if $\beta$ is smaller than the datasheet's "typical."

The saturation condition (the acid test)

The transistor is truly saturated when:

$\beta I_B \; \ge \; I_{C(\text{on})} \quad\Longleftrightarrow\quad \frac{I_C}{I_B} \le \beta$

The ratio $I_C/I_B$ actually flowing is called $\beta_{\text{forced}}$ . If $\beta_{\text{forced}} < \beta$ , you are saturated.

Worked Example 1 — Driving an LED-like load

Drive a load needing $I_C = 100$ mA from $V_{CC} = 5$ V, controlled by a $3.3$ V logic pin. Transistor $\beta_{min} = 100$ .

Collector current (given/target): $I_C = 100$ mA. Why? This is set by the load, not us.
Edge base current: $I_{B(edge)} = 100\text{ mA}/100 = 1$ mA. Why? Minimum base current $\beta$ can amplify to 100 mA.
Overdrive $k=5$ : $I_B = 5\text{ mA}$ . Why? Guarantees saturation despite $\beta$ spread.
Base resistor: $R_B = \frac{3.3 - 0.7}{5\text{ mA}} = \frac{2.6}{0.005} = 520\ \Omega$ Pick nearest standard 470 Ω (slightly more overdrive — fine). Why 470 not 560? Lower $R_B$ = more base current = deeper saturation, safer for a switch.

Worked Example 2 — Check if it's really saturated

Same circuit, but suppose the real transistor has $\beta = 60$ (worse than datasheet).

We supply $I_B = 5$ mA (from the 470 Ω design). Actual $I_B = (3.3-0.7)/470 \approx 5.5$ mA.
Max collector current transistor could pass in active mode $= \beta I_B = 60 \times 5.5\text{ mA} = 330$ mA.
Load only demands 100 mA.
Since $330 > 100$ , the transistor can't push all that current → the collector voltage collapses → saturated. ✅

Why this step? The collector current is limited by the load, not by $\beta I_B$ , whenever $\beta I_B$ exceeds the load demand. That surplus is exactly what "hard ON" means.

Worked Example 3 — Power dissipation sanity check

While ON, the transistor dissipates $P = V_{CE(sat)}\cdot I_C$ .

$P = 0.2\text{ V} \times 0.1\text{ A} = 0.02\text{ W} = 20\text{ mW}$

Why so low? Because $V_{CE(sat)}$ is tiny. Compare to linear operation at mid-supply: $2.5\text{ V}\times0.1\text{ A}=250$ mW — 12.5× hotter. This is why switching is efficient.

Common mistake Steel-manned common errors

1. "Just use $I_B = I_C/\beta$ exactly." Why it feels right: that's the textbook active-region relation. The fix: At the exact edge, any drop in $\beta$ (heat, tolerance) throws you back into active region → high $V_{CE}$ → the transistor overheats. Always overdrive by 2–10×.

2. Forgetting to subtract $V_{BE}$ when sizing $R_B$ . Why it feels right: $0.7$ V seems negligible next to $3.3$ V. The fix: It changes $R_B$ by ~20% and, near threshold voltages, can be huge. Use $V_{in}-V_{BE(on)}$ .

3. Putting the load on the emitter (emitter follower) expecting a good switch. Why it feels right: current still flows through the load. The fix: Then the emitter can't sit at 0 V, so $V_{BE}$ shrinks and you lose control ("emitter degeneration"). For a low-side switch, load on collector, emitter grounded.

4. Assuming $V_{CE(sat)} = 0$ . Why it feels right: "closed switch = 0 volts." The fix: It's ~0.1–0.2 V. For high currents this drop and its power matter.

Recall Feynman: explain to a 12-year-old

Imagine a garden hose (the big water = collector current) with a stiff valve. A grown-up can barely turn the valve with one finger. But you attach a little lever (the base current): now a tiny push on the lever fully opens or shuts the big valve. The BJT is that lever — a weak signal from a computer chip opens or closes a big flow of electricity to a motor or lamp. And you push the lever a bit harder than needed so it never slips halfway open (which would make it get hot).

Active-Recall Flashcards

#flashcards/hardware

What are the two regions used for BJT switching?

Cutoff (OFF, open switch) and Saturation (ON, closed switch).

In saturation, what is

V_{CE}

roughly equal to?

V_{CE(sat)} \approx 0.1

–

0.2

Why do we overdrive the base current instead of using

I_B=I_C/\beta

exactly?

To guarantee hard saturation despite

\beta

variation with temperature and tolerance.

Formula for the collector ON-current with a load resistor

R_L

I_{C(on)} = (V_{CC}-V_{CE(sat)})/R_L

Formula for the base resistor?

R_B = (V_{in}-V_{BE(on)})/I_B

What is the saturation condition in terms of

\beta

and currents?

\beta I_B \ge I_C

, i.e.

\beta_{forced} = I_C/I_B \le \beta

Why is switching more power-efficient than linear operation?

Because

V_{CE(sat)}

is tiny, so

P=V_{CE(sat)}I_C

is small; in active region

V_{CE}

is large and dissipates more.

For a low-side NPN switch, where do the load and emitter go?

Load between

V_{CC}

and collector; emitter to ground.

What is the overdrive factor

k

typically?

About 2 to 10.

\beta I_B

(330 mA) far exceeds load demand (100 mA), what limits

I_C

The load/

R_L

limits it; the surplus capability just means deep saturation.

Connections

BJT as an amplifier — same device, but biased in the active region we avoid here.
MOSFET as a switch — voltage-controlled alternative, no steady gate current, lower $R_{ON}$ loss at high current.
Transistor beta (hFE) — the gain parameter whose variability forces overdrive.
Saturation and Cutoff regions — the operating-region map on the output characteristics.
Flyback diode — needed across inductive loads (relays/motors) switched by a BJT.
Kirchhoff's Voltage Law — the tool used to derive every equation above.
Power dissipation in switches — why $V_{CE(sat)}$ and $I_C$ set heating.

Concept Map

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Bhai, BJT ko switch ki tarah use karna bahut simple idea hai: thoda sa base current dedo, aur transistor bade collector current ko ON kar deta hai — jaise ek chhoti si chaabi se bada gate khulta hai. Do hi states matter karti hain: cutoff (base pe kuch nahi, transistor OFF = open switch) aur saturation (base pe kaafi current, transistor fully ON, $V_{CE}$ sirf 0.1–0.2 V bachta hai = closed switch). Beech ka "active" region amplifier ke liye hai, switch ke liye nahi.

Design ka funda: pehle load se $I_C$ nikalo, phir $I_B = I_C/\beta$ nikalo. Lekin yahin pe log galti karte hain — sirf itna base current mat do. $\beta$ garmi aur part ke hisaab se badalta rehta hai, isliye overdrive karo, matlab 2 se 10 guna zyada base current. Isse transistor pakka deep saturation mein rahega. Base resistor ka formula bhi KVL se banta hai: $R_B = (V_{in} - 0.7)/I_B$ . Woh $0.7$ V ka $V_{BE}$ mat bhoolna, chhota lagta hai par matter karta hai.

Yeh important kyun hai? Microcontroller ka pin sirf thoda sa current de sakta hai, motor ya relay ya bright LED chala nahi sakta. BJT beech mein aakar weak signal se strong load ko command karta hai. Aur switching efficient hai kyunki ON state mein $V_{CE(sat)}$ chhota hai, toh $P = V_{CE(sat)}\times I_C$ kam banta hai — transistor thanda rehta hai, linear mode ke mukable bahut kam garmi.