If you've ever taken an
electrical engineering course, you've probably seen it: a dense web of
overlapping circles that looks like a wormhole from a sci-fi movie. Some
textbooks even call it “black magic.” It's called the Smith chart, and despite
the intimidating appearance, it solves a very real, very practical problem —
and once you see the problem it solves, the chart stops looking like magic and
starts looking like a clever shortcut.
This article walks through
where the chart came from, the physics problem that forced it into existence,
and the math — including complex (“imaginary”) numbers — that makes it work.
We'll build up the ideas one at a time, so nothing needs to be taken on faith.
1. The problem: signals that bounce back
In 1928, Philip H. Smith took a
job at Bell Labs. At the time, there was no way to place a phone call across
the Atlantic Ocean — copper cables couldn't do it, so the only option was
radio. Smith's job was to help send a radio signal from New Jersey to receiving
stations in England and Argentina.
To boost the signal strength,
Smith's team didn't use one antenna — they used more than twenty, wired
together through over 2 km of transmission line, all aimed to focus the radio
energy into a narrow beam pointed at the target.
But when Smith tested the
system, he found that part of the signal sent down the line was bouncing back
instead of reaching the antennas. That reflected energy was wasted — it never
made it to its destination. If he wanted a signal to travel across the planet,
he had to figure out how to stop that reflection.
2. Why the signal bounces: AC, wavelengths, and mismatches
A battery-and-lightbulb circuit
uses direct current (DC) — a steady voltage, a steady current. Radio signals
aren't like that. To broadcast a radio wave, the current has to oscillate back
and forth, rising and falling like a sine wave. This is alternating current
(AC).
An AC wave has:
•
a wavelength — the distance from one peak to the next,
•
a frequency — how many peaks pass a fixed point per
second,
•
and a speed, which equals wavelength × frequency.
Here's the key detail: for
ordinary household AC power (50–60 Hz), the wavelength is thousands of
kilometers long — far longer than any wire in your house, so reflections barely
matter. But Smith was working with radio frequencies in the megahertz range,
where the wavelength shrinks to tens of meters. His transmission line was over
2 km long — many, many wavelengths — so any reflection became a serious
problem.
Why does a mismatch cause
reflection at all? Think of shaking a slinky that's tied to a second slinky
with different weight per unit length. When a wave hits the junction between
the two slinkies, part of it continues through and part bounces back — simply
because the wave “feels” a different environment on each side. The size of the
bounced-back wave relative to the original is called the reflection
coefficient. If the two slinkies are identical, there's no bounce, and the
reflection coefficient is zero.
Electrically, the property that
plays the role of “mass per unit length” is called characteristic impedance.
It's worth pausing on that name, because it trips people up: this is not the DC
resistance you'd measure across the cable with a multimeter (that's close to
zero for a good copper conductor). Characteristic impedance is a property of
how voltage and current waves travel along the cable — set by the cable's
geometry (conductor size, spacing) and the insulating material between the
conductors, not by how resistive the copper is.
If the transmission line's
impedance doesn't match the antenna's impedance, some of the signal reflects —
exactly like the slinkies. In Smith's real setup, his transmission line was 50
ohms and his antenna array was 12.5 ohms — a serious mismatch, and a serious
reflection.
When the forward wave (heading
toward the antenna) and the reflected wave (heading back toward the
transmitter) overlap on the same cable, they combine into a standing wave — a
pattern of high and low voltage points that stays fixed in place along the line
rather than traveling. This is exactly what engineers measure as SWR (standing
wave ratio), and it's the most common real-world reason anyone reaches for a
Smith chart today: a bad SWR reading on a piece of test equipment is the
practical symptom of the mismatch problem this whole article is about.
If the reflections are bad
enough, the peak voltage on the line can reach twice the input voltage, which
is enough to physically burn out a cable.
3. Why a resistor alone doesn't fix it
The obvious fix might be to add
a resistor to bring the two impedances closer together. It doesn't work well,
for two reasons:
1.
Resistors waste power. They dissipate energy as heat —
which is exactly the loss you're trying to eliminate.
2.
Impedance isn't just about size — it's also about
timing.
That second point needs
unpacking, and it's where the real physics (and the complex numbers) come in.
4. Capacitors and inductors shift timing, not just size
Picture a capacitor: two
conductive plates with a gap between them. Apply a voltage, and charge builds
up on the plates — but the charge (current) doesn't rise and fall in step with
the voltage. Current into a capacitor is proportional to how fast the voltage
is changing, not to the voltage itself. So at the instant the voltage reaches
its peak, it has (just for that instant) stopped changing — and since current
depends on the rate of change, the current at that exact instant is zero. The
current reaches its own peak a quarter-cycle before the voltage does. Engineers
say current leads voltage by 90° in a capacitor (equivalently, voltage lags current).
An inductor (a coil of wire)
does the opposite. A changing current creates a magnetic field, and that field
pushes back against changes in current. The result: voltage leads current by
90° in an inductor.
A plain resistor doesn't shift
timing at all — voltage and current rise and fall together, in lockstep.
So a real transmission line
isn't just resisting the signal; it's also shifting its timing depending on how
much capacitance and inductance are present. To properly “match” two systems,
you need to match both the relative size of voltage and current and the timing
(phase) between them. A single number (resistance) can't capture that. You need
something that carries two pieces of information at once — and that's exactly
what complex numbers are for.
5. Complex numbers, explained simply
Forget “imaginary numbers are
numbers you square to get −1” for a moment — that's the textbook definition,
but it's not the most useful way to think about them here.
Instead, picture a flat map —
the complex plane. It has a normal horizontal axis (real numbers, like the
number line you already know) and a vertical axis (the “imaginary” axis). Any
point on this map is a complex number: part real, part imaginary.
The genius of this plane is
what happens when you multiply:
•
Multiplying by an ordinary (real) number scales things.
Multiply by 3, and everything stretches to three times its size, in the same
direction.
•
Multiplying by the imaginary unit (electrical engineers
call it j, not i, because i is already used for current) rotates things by 90°.
Start at the point 1 on the real axis. Multiply by j, and you rotate 90° to
land on the imaginary axis. Multiply by j again, and you rotate another 90° —
landing on −1. Do it twice more and you're back where you started, having gone
all the way around.
So multiplying by a complex
number does two things simultaneously: it scales (changes size) and rotates
(changes angle/timing). That's precisely the two things a real AC circuit does
to a signal — resistors change size, capacitors and inductors shift timing
(phase). Complex numbers are a natural, compact way to describe both effects
with a single quantity.
Engineers bundle this into a
value called impedance, written Z:
Z = R + jX
•
R is resistance — the “real,” everyday size-changing part.
•
X is reactance — the phase-shifting part contributed by
capacitors (negative X) and inductors (positive X).
Impedance is still just Ohm's
Law generalized for AC: Z = V / I — but now both V and I are complex numbers,
so Z captures the ratio of their sizes and the phase shift between them, all in
one number. On the complex plane, resistance is the horizontal position,
inductive reactance points “up,” capacitive reactance points “down.”
6. The infinity problem
To eliminate reflections,
engineers try to cancel the reactance first (add an inductor to cancel a
capacitor's effect, or vice versa — they point in opposite directions on the
complex plane, so they subtract out), leaving only resistance. Then, if that resistance
still doesn't match the line, there's a genuinely clever trick: rather than
adding a lossy resistor, you exploit the fact that impedance changes as you
move along the transmission line (because the forward and reflected waves
interfere differently at different points). Somewhere along the line, there's a
point where the resistance naturally matches. Find that point, cancel the
leftover reactance there with a lossless capacitor or inductor, and you've
eliminated the reflection — with no power wasted as heat.
The catch: doing this by hand
with the raw equations (first worked out by Oliver Heaviside decades earlier)
meant slogging through long, awkward calculations by hand or slide rule. Smith
wanted a graphical shortcut. But there was a geometric obstacle in his way.
A real impedance can range from
zero (a short circuit — all current, no voltage) to infinite (an open circuit —
all voltage, no current). Every value in between is possible. How do you draw a
chart that includes infinity, on a finite piece of paper?
7. Folding infinity into a circle
Smith brought in two
mathematician colleagues, Frell and McCrae, and together they used a property
of complex-number functions. Consider the transformation:
new value = 1 / Z
Apply this to the entire
complex plane, and something remarkable happens: straight lines curve, the
whole plane warps — and everything that used to run off toward infinity gets
pulled inward and packed into a finite region. Zoom in on any small patch of this
warped plane, and it still looks locally like a plain square grid — angles and
local shapes are preserved even though the picture as a whole is bent.
Mathematicians call this a conformal map. It's the mathematical equivalent of
one of those elastic map projections that folds an infinite plane onto the
surface of a sphere: nothing gets torn, but everything gets curved.
That's the trick that makes the
impossible possible: a transformation that takes an infinite range of impedance
values and folds them into a finite circle, without losing any information.
1/Z is a useful demonstration
of that folding trick, but it isn't quite what ended up on the chart. It's fair
to ask: if 1/Z already solves the infinity problem, why go any further?
8. From impedance to reflection coefficient
The answer is that Smith used a
transformation closely related to 1/Z, but even better suited to the physics:
the reflection coefficient, usually written Î (the Greek letter gamma), defined
as the reflected wave divided by the forward wave. In terms of impedance, it
works out to:
Î = (Z − Z₀) / (Z + Z₀)
where Z is the impedance being
plotted and Z₀ is the transmission line's characteristic impedance. You don't
need to derive this to use the chart — the important part is what it buys you.
It carries exactly the same information as impedance (each impedance value
corresponds to one unique point on this new plane), but with a crucial
advantage over plain 1/Z: since a reflected wave can never be bigger than the
wave that created it, the reflection coefficient's size can never exceed 1. The
infinity problem doesn't just get folded away — it disappears entirely, by
definition, and the boundary of the Smith chart (the outer circle) has a direct
physical meaning: it's the edge case of total reflection.
Applying the transformation:
•
Every vertical line of constant resistance on the
impedance plane becomes a circle on the reflection coefficient plane. Zero
resistance gives the biggest possible circle (running from −1 to 1). Resistance
equal to the line's own characteristic impedance (a “perfect match”) gives a
circle that passes right through the center. As resistance climbs toward
infinity, the circles shrink and cluster near the point (1, 0).
•
Every horizontal line of constant reactance also
becomes a circle — above the center for inductance, below for capacitance,
shrinking as the reactance increases, and squashing flat into the horizontal
axis itself when reactance is zero.
Overlay both families of
circles, and you get the finished Smith chart: a dense grid of resistance
circles and reactance circles, all packed inside one circle of radius 1.
9. Reading and using the chart
To use it, an engineer first
normalizes the measured impedance by dividing by the line's characteristic
impedance (commonly 50 ohms). The point of this step: it makes any transmission
line look like a generic “1-ohm system” from the chart's point of view. A
single printed chart, with its center marked “1,” can then be reused for a
50-ohm system, a 75-ohm cable-TV system, or anything else — you just divide by
whatever Z₀ applies before you plot, and multiply back by Z₀ when you read a
result off. For example, a measured impedance of 36 + 74j ohms on a 50-ohm line
normalizes to:
Z = 0.7 + 1.5j
Plot that point by finding
where the “0.7” resistance circle crosses the “1.5” reactance circle. The
distance from that point to the center of the chart is the magnitude of the
reflection coefficient (in this example, about 0.68 — meaning the reflected
wave is 68% the size of the forward wave, a substantial mismatch).
The goal is always to walk that
point to the center of the chart — the point where resistance equals 1 (a
perfect match) and reactance is zero (no leftover phase shift), meaning zero
reflection.
Two important facts about the
chart make this practical:
3.
Moving along the transmission line traces a circle on
the chart — because the reflection coefficient's magnitude stays constant as
you move (only its phase angle rotates), and a full 360° rotation corresponds
to moving exactly half a wavelength along the physical line.
4.
A dangling, unconnected stub of extra cable, cut to a
specific length, can supply any pure reactance you need — cancelling out
whatever's left over — without wasting any power as heat, since nothing is
being resistively dissipated.
So the whole matching process
becomes: (1) plot your impedance, (2) rotate along the constant-resistance
circle (by physically moving along the line, or adding a length of line) until
the resistance equals 1, then (3) add a stub of the right length to cancel the
remaining reactance, landing you exactly at the center. In practice, this is
exactly what engineers do in the field — measuring a mismatched antenna system,
calculating a stub length using the chart, cutting a piece of coaxial cable to
that length, and watching a wasted signal loss disappear into a clean, fully
matched transmission line.
10. Why it stuck around
Smith's chart was slow to catch
on — it took about two years and several rejections before a technical magazine
agreed to publish it in 1937. Around the same time, engineers in Japan (Tosaku
Mizuhashi) and the Soviet Union (Amiel Vulpert) independently arrived at
essentially the same graphical solution — three groups, working in isolation,
converging on the same elegant geometry.
It was World War II that
cemented its place in engineering. At the MIT Radiation Laboratory, scientists
building microwave radar to detect submarines needed fast, reliable ways to
keep their systems free of reflections, and Smith's chart became an everyday
working tool. When the war ended, those engineers carried it into industry,
universities, and textbooks — and the name stuck to Smith's version
specifically, even though it had been discovered independently elsewhere.
Today, computers can calculate
an optimal impedance match instantly, no chart required. But the Smith chart is
still taught in classrooms and still built into professional RF measurement
equipment, because a computer can hand you the answer without teaching you
anything — while the chart shows why a fix works and gives an engineer a
visual, physical intuition for which direction to move. It's less like a
formula and more like a hand-drawn map — “go down the street, turn left, go
further” — guiding you intuitively from where you are to where you want to be.
11. Why it's still on every screen in the lab
This is also why the chart
never actually disappeared, even after computers made hand calculation
obsolete. Turn on a modern Vector Network Analyzer (VNA), open an antenna
analyzer app, or fire up RF simulation software like ADS, HFSS, CST, or the RF
tools in LTspice, and a Smith chart is almost always sitting right there on
screen — often as the default view. The reason is human, not mathematical: a
table of complex impedance numbers changing as you sweep frequency is hard to
absorb at a glance, but a trace moving across a Smith chart is something a
trained eye can read instantly — “that's inductive, that's a decent match,
that's drifting toward trouble.” The chart survives not because the math still
requires it, but because human pattern recognition is faster than reading
numbers off a screen.
The short version
•
Radio signals sent down a mismatched cable partially
bounce back, wasting power and potentially damaging equipment.
•
Matching requires controlling both the size and timing
of a signal — which is why engineers use complex numbers (impedance =
resistance + j·reactance) instead of a single number.
•
Real impedances range from zero to infinity, which
can't fit on an ordinary chart.
•
A mathematical trick (a conformal map, using the
reflection coefficient instead of impedance directly) folds that entire
infinite range into one finite circle without losing any information.
•
The result — the Smith chart — turns a tedious algebra
problem into a visual, geometric one: walk your point to the center of the
circle, and your reflections vanish.
-
more details : https://www.youtube.com/watch?v=GK2pZ_oVU1o
