ARCHETYPE: Real Meters — Loading Effects & Reading Errors

Here is a small scandal at the heart of every electrical measurement: the instruments lie. Not carelessly — systematically. Connect a voltmeter across a resistor and the voltmeter itself becomes a second path for current, quietly lowering the very voltage it was sent to report. Slip an ammeter into a branch and its own coil resistance adds to the branch, throttling the very current it displays. The meters are not faulty; they are physical. Every real instrument steals a little of the circuit it touches, and the theft has a direction, a size, and — this is the exam-worthy part — a formula. The trouble traces back to 1882, when Jacques-Arsène d'Arsonval perfected the moving-coil movement: a coil of fine wire twisting in a magnetic field in proportion to its current. Fine wire means resistance — so the instrument that measures electricity is itself a resistor, and the moment you wire it in, you are analysing a different circuit from the one you drew. Examiners at ICSE, JEE, NEET and A-Level mine this relentlessly: which of the two wiring arrangements for measuring a resistance errs high, which errs low, which suits a small resistance, and by what percentage. This archetype drills all of it — the two-arrangement analysis, the error formulas, the crossover rule, and the voltmeter that drags a potential divider's output down by a third. The ideal-meter conventions come from Circuit Elements, Symbols & Diagrams; here the meters grow up and get resistances.

Key takeaways

A real ammeter is a small resistance $r_A$ in series; a real voltmeter is a large resistance $R_V$ in parallel. Ideal limits: $r_A \to 0$ , $R_V \to \infty$ .
Measuring $R$ as $\dfrac{\text{voltmeter reading}}{\text{ammeter reading}}$ gives the resistance of what the meters actually enclose, never exactly $R$ .
Ammeter-near arrangement (voltmeter wrapped around both): reads $R + r_A$ — errs high, fractional error $\dfrac{r_A}{R}$ — best for large $R$ .
Voltmeter-near arrangement (voltmeter hugging $R$ alone): reads $\dfrac{RR_V}{R + R_V}$ — errs low, fractional error $\approx \dfrac{R}{R_V}$ — best for small $R$ .
Crossover: the two errors tie at $R = \sqrt{r_A R_V}$ — below it hug with the voltmeter, above it accept the ammeter's tax.
A voltmeter across a divider loads the tap: reading a 6 V tap as 4 V is normal physics, not a broken meter.
KEY IDEA: replace every meter by its resistance, redraw, and solve — the "measured value" is just the honest analysis of the circuit the meters created.

The core insight

Real meters are resistors — $r_A$ small and in series, $R_V$ large and in parallel — so every measurement analyses a modified circuit: the ratio $V/I$ from the dials returns $R + r_A$ in one wiring and $RR_V/(R+R_V)$ in the other, and choosing the wiring means choosing which systematic error you can afford.

Where this lives today

Every digital multimeter's datasheet leads with the two numbers this archetype is built on: input resistance on the voltage ranges (typically $10\ \mathrm{M\Omega}$ ) and burden voltage on the current ranges — the modern name for the ammeter's $Ir_A$ drop. Oscilloscope probes exist precisely so the scope does not become the divider ambush on the circuit under test; high-impedance sensors (pH electrodes at $10^{11}\ \Omega$ ) need buffer amplifiers because no passive voltmeter is high-resistance enough; and power electronics chooses milliohm current-sense shunts so the measurement does not eat the efficiency being measured. The meters got better; the theft never went away — it just moved down the decimal places.

The setup

One measurement, two honest ways to wire it — and they disagree. You want $R$ from simultaneous dial readings, $R_{\text{meas}} = V_{\text{read}}/I_{\text{read}}$ . The only choice is what the voltmeter bridges, and that choice decides which meter contaminates the answer.

Figure

The two wiring arrangements for measuring a resistance — each plants its systematic error in a different meter

The real ammeter is a low-resistance instrument placed in series: the current threads it, so its coil resistance $r_A$ (fractions of an ohm to a few ohms) simply adds to the branch. Its crime: it reduces the current it reports and adds its own voltage drop to whatever a wrapped-around voltmeter sees.

The real voltmeter is a high-resistance instrument placed in parallel: it reads the potential difference across its own terminals by letting a tiny sample current through itself. Its crime: that sample current is stolen from the element under test, so the parallel combination has less resistance — and less voltage — than the element alone. The ideal-limit conventions ( $r_A = 0$ , $R_V = \infty$ ) that let you ignore all this at first pass are set up in Circuit Elements, Symbols & Diagrams; this archetype is what happens when the training wheels come off.

The master pattern

Every meters problem is the same discipline:

Step 1. Redraw the circuit with each meter replaced by its resistance — ammeter as a small series box $r_A$ , voltmeter as a large parallel box $R_V$ .

Step 2. Solve the redrawn circuit with the toolkit from Resistors in Series & Parallel.

Step 3. Identify what each dial actually displays: the ammeter shows the current through itself; the voltmeter shows the voltage across itself.

Step 4. Form $R_{\text{meas}} = V_{\text{read}}/I_{\text{read}}$ and compare with the true $R$ : the difference is the systematic error, with a predictable sign.

Derivation AWhat each arrangement actually measures and its percentage error

We want exact expressions for the measured resistance $R_{\text{meas}} = V_{\text{read}}/I_{\text{read}}$ in both wiring arrangements, their signed errors, and the crossover resistance at which the two arrangements are equally wrong. The strategy is the master pattern itself: replace meters by resistances, track what each dial displays, and divide. Keep the signs of the errors distinct — one arrangement can only err high, the other only low, and examiners test exactly that asymmetry.

Step 1. Ammeter-near arrangement (voltmeter wrapped around ammeter + resistor in series). The same current $I$ threads the ammeter and $R$ , so the ammeter's dial is honest about $R$ 's current. But the voltmeter spans both elements:

V_{\text{read}} = I(R + r_A) \quad\Rightarrow\quad R_{\text{meas}} = \frac{V_{\text{read}}}{I} = R + r_A

The error is additive and positive: $\Delta R = +r_A$ , fractional error $\dfrac{r_A}{R}$ . Why it favours large R: a fixed 0.5 Ω theft is 5% of 10 Ω but only 0.001% of 50 kΩ — the ammeter's tax becomes invisible next to a big resistance.

Step 2. Voltmeter-near arrangement (voltmeter hugging $R$ alone). Now the voltmeter's dial is honest about $R$ 's voltage, but the ammeter upstream carries two currents — $R$ 's and the voltmeter's sample:

I_{\text{read}} = \frac{V}{R} + \frac{V}{R_V} \quad\Rightarrow\quad R_{\text{meas}} = \frac{V}{I_{\text{read}}} = \frac{RR_V}{R + R_V}

This is just the parallel combination of $R$ with $R_V$ — necessarily below $R$ . The error is negative: $\Delta R = -\dfrac{R^2}{R+R_V}$ , and for $R \ll R_V$ the fractional error is approximately $-\dfrac{R}{R_V}$ . Why it favours small R: the voltmeter steals current in proportion to R/R_V, so hugging a 10 Ω resistor with a 10 kΩ voltmeter costs only 0.1%.

Step 3. The crossover. Set the two fractional errors equal: $\dfrac{r_A}{R} = \dfrac{R}{R_V}$ , giving

R_{\text{cross}} = \sqrt{r_A R_V}

Below $\sqrt{r_AR_V}$ , the voltmeter-near wiring is more accurate; above it, ammeter-near. The geometric mean appears for the same structural reason as in the Wheatstone-flavoured problems: two competing fractional errors, one scaling as $1/R$ and one as $R$ .

Step 4. Anchor the numbers: $r_A = 0.5\ \Omega$ , $R_V = 10\ \mathrm{k\Omega}$ , so $R_{\text{cross}} = \sqrt{0.5 \times 10^4} \approx 71\ \Omega$ . Measuring a true $10\ \Omega$ : ammeter-near reads $10.5\ \Omega$ (+5%); voltmeter-near reads $10 \times 10^4/10010 = 9.99\ \Omega$ (−0.1%) — hug the small resistor with the voltmeter. Measuring a true $50\ \mathrm{k\Omega}$ : voltmeter-near reads $50\text{k} \parallel 10\text{k} = 8.33\ \mathrm{k\Omega}$ (−83%, catastrophic); ammeter-near reads $50000.5\ \Omega$ (+0.001%) — the ammeter's tax is nothing to a big resistance. Verified: all four error figures recomputed numerically — +5.0%, −0.0999%, −83.3%, +0.001%; crossover 70.7 Ω.

Loading a potential divider — the classic ambush

The same theft, staged one circuit earlier. A potential divider (built in the Potential Dividers module) taps a fraction of a supply: two $10\ \mathrm{k\Omega}$ resistors across $12\ \mathrm{V}$ put $6\ \mathrm{V}$ at the tap — on paper. Now measure that tap with a voltmeter of $R_V = 10\ \mathrm{k\Omega}$ . The voltmeter lands in parallel with the lower arm:

R_{\text{lower}}' = \frac{10 \times 10}{10 + 10} = 5\ \mathrm{k\Omega} \quad\Rightarrow\quad V_{\text{read}} = 12 \times \frac{5}{10 + 5} = 4.0\ \mathrm{V}

Figure

The divider ambush — a 10 kilo-ohm voltmeter drags a 6 V tap down to 4 V before reading it

The dial says $4.0$ V where theory promised $6.0$ V — a 33% shortfall from a perfectly healthy meter. Nothing failed: the circuit with the voltmeter attached genuinely has $4.0$ V at the tap. The rule of thumb the A-Level examiners want stated: a voltmeter only reads a divider faithfully if $R_V$ is much larger than the arm it bridges — ten times larger keeps the droop under ~5%, a hundred times under ~0.5%.

The variants examiners actually pose

variant	governing result	sign of error
ammeter-near, measure $R$	$R_{\text{meas}} = R + r_A$	high, $+\dfrac{r_A}{R}$
voltmeter-near, measure $R$	$R_{\text{meas}} = \dfrac{RR_V}{R+R_V}$	low, $\approx -\dfrac{R}{R_V}$
which wiring for small $R$	$R < \sqrt{r_AR_V}$	voltmeter-near
which wiring for large $R$	$R > \sqrt{r_AR_V}$	ammeter-near
correct a voltmeter-near reading	$I_R = I_{\text{read}} - \dfrac{V_{\text{read}}}{R_V}$ , then $R = \dfrac{V_{\text{read}}}{I_R}$	recover true $R$
voltmeter on a divider tap	tap arm becomes $R \parallel R_V$	reading droops
ammeter added to a branch	branch resistance grows by $r_A$	current drops
ideal-limit check	$r_A \to 0$ , $R_V \to \infty$	both errors vanish

Worked examples

Board level 1 — the voltmeter that dragged the reading (ICSE/A-Level pattern)

Two $5\ \mathrm{k\Omega}$ resistors in series are connected across a $10\ \mathrm{V}$ supply. A voltmeter of resistance $20\ \mathrm{k\Omega}$ is connected across the lower resistor. Find the theoretical tap voltage, the voltmeter's actual reading, and the percentage error.

Step 1. Unloaded theory: the divider splits evenly, $V_{\text{tap}} = 10 \times \dfrac{5}{5+5} = 5.0\ \mathrm{V}$ .

Step 2. Load the tap: the voltmeter sits in parallel with the lower arm, $5 \parallel 20 = \dfrac{5 \times 20}{25} = 4.0\ \mathrm{k\Omega}$ .

Step 3. Re-solve the divider: $V_{\text{read}} = 10 \times \dfrac{4}{5 + 4} = \boxed{4.44\ \mathrm{V}}$ .

Step 4. Error: $\dfrac{4.44 - 5.0}{5.0} = \boxed{-11\%}$ . The meter reads low — always low, for a parallel thief — and a $R_V$ only four times the arm resistance was never going to be gentle.

Board level 2 — same resistor, two honest arrangements (ISC pattern)

A resistor of true resistance $10.0\ \Omega$ is measured with an ammeter of resistance $0.5\ \Omega$ and a voltmeter of resistance $10\ \mathrm{k\Omega}$ , using each of the two standard arrangements. What value does each arrangement deliver, and which should be preferred?

Step 1. Ammeter-near: the voltmeter wraps resistor and ammeter, so $R_{\text{meas}} = R + r_A = 10.0 + 0.5 = \boxed{10.5\ \Omega}$ — 5% high.

Step 2. Voltmeter-near: the ammeter also carries the voltmeter's sample, so $R_{\text{meas}} = \dfrac{RR_V}{R + R_V} = \dfrac{10 \times 10000}{10010} = \boxed{9.99\ \Omega}$ — 0.1% low.

Step 3. Verdict: prefer voltmeter-near — check with the crossover, $R_{\text{cross}} = \sqrt{0.5 \times 10^4} \approx 71\ \Omega$ , and $10\ \Omega$ sits far below it.

Step 4. Ideal-limit sanity: set $r_A = 0$ in Step 1 and $R_V \to \infty$ in Step 2 — both collapse to exactly $10.0\ \Omega$ ✓.

Competitive level 1 — choose the wiring, then defend it (JEE-Main pattern)

With an ammeter of $r_A = 0.5\ \Omega$ and a voltmeter of $R_V = 10\ \mathrm{k\Omega}$ , resistances of roughly $5\ \Omega$ and $50\ \mathrm{k\Omega}$ must each be measured to better than 1%. Assign the correct arrangement to each and verify the errors.

Step 1. Crossover first: $R_{\text{cross}} = \sqrt{r_AR_V} = \sqrt{5000} \approx 71\ \Omega$ . Small resistor below it → voltmeter-near; big resistor above it → ammeter-near.

Step 2. The $5\ \Omega$ by voltmeter-near: error $\approx -\dfrac{R}{R_V} = -\dfrac{5}{10^4} = \boxed{-0.05\%}$ ✓ (the wrong choice, ammeter-near, would err $+r_A/R = +10\%$ — twenty times outside spec).

Step 3. The $50\ \mathrm{k\Omega}$ by ammeter-near: error $= +\dfrac{0.5}{5 \times 10^4} = \boxed{+0.001\%}$ ✓ (the wrong choice would read $50\text{k} \parallel 10\text{k} = 8.33\ \mathrm{k\Omega}$ — an 83% catastrophe).

Step 4. State the rule as the examiner wants it: ammeter-near for resistances large compared with $\sqrt{r_AR_V}$ ; voltmeter-near for small — each wiring hides its meter's theft where it does least harm. Timing note: compute the crossover once (~10 s) and both assignments follow instantly; spend the remaining minute on the two error evaluations. Full marks inside 2 minutes.

Competitive level 2 — un-stealing the reading (JEE-Advanced pattern)

In the voltmeter-near arrangement, the ammeter reads $2.0\ \mathrm{A}$ and the voltmeter ( $R_V = 200\ \Omega$ ) reads $20\ \mathrm{V}$ . A student reports $R = 20/2.0 = 10\ \Omega$ . Find the true resistance, and the percentage error of the naive report.

Step 1. Diagnose the theft: the ammeter's 2.0 A includes the voltmeter's sample current, $I_V = \dfrac{V_{\text{read}}}{R_V} = \dfrac{20}{200} = 0.10\ \mathrm{A}$ .

Step 2. The resistor's actual current: $I_R = 2.0 - 0.10 = 1.90\ \mathrm{A}$ .

Step 3. True resistance: $R = \dfrac{V_{\text{read}}}{I_R} = \dfrac{20}{1.90} = \boxed{10.53\ \Omega}$ .

Step 4. The naive $10\ \Omega$ errs by $\dfrac{10 - 10.53}{10.53} = \boxed{-5\%}$ — exactly the parallel-combination prediction, since $10 = \dfrac{10.53 \times 200}{10.53 + 200}$ ✓. Correcting a loading error is one subtraction — when the question supplies $R_V$ , the examiner expects the correction, not the excuse. Timing note: the entire correction is "subtract the voltmeter current, re-divide" — 60 seconds. The trap is skipping the diagnosis and certifying 10 Ω; the physics-aware answer is worth the extra minute.

Exam traps

Blaming the meter. Loading error is not inaccuracy; the meter reports its own circuit exactly. Answers that say "the voltmeter is faulty" where the syllabus wants "the voltmeter loads the circuit" lose the reasoning mark.

Confusing which arrangement errs which way. Ammeter-near can only read high ( $+r_A$ , additive); voltmeter-near can only read low (parallel theft). If your computed error has the wrong sign for the named arrangement, you have swapped the analyses.

"High resistance" reflexes. The voltmeter must have high resistance relative to what it bridges — 10 kΩ is superb across 10 Ω and useless across 50 kΩ. Never certify a voltmeter without comparing $R_V$ to the bridged element; the divider ambush lives entirely in that comparison.

Forgetting whose current the ammeter shows. In voltmeter-near wiring the ammeter reads resistor current plus voltmeter current. Given $R_V$ , subtract before dividing — reporting $V/I$ raw is the planted mistake in every "a student reports…" stem.

The crossover as geometric mean. $R_{\text{cross}} = \sqrt{r_AR_V}$ , not the average and not $R_V/r_A$ . With $r_A = 0.5\ \Omega$ and $R_V = 10\ \mathrm{k\Omega}$ it is ~71 Ω — remarkably low, which is why "use ammeter-near" is the right call for most laboratory resistors above ~100 Ω.

Series ammeter in a live branch. Adding an ammeter of $r_A$ raises the branch resistance and lowers the very current being measured: measured current ≤ original current, always.

Board conventions. ICSE and A-Level (CIE 9702) pose the divider-loading and percentage-error forms and expect the "loads the circuit" language; JEE-Main/Advanced and NEET favour the arrangement choice and the correction arithmetic. Ideal meters still rule unless the stem grants a meter resistance.

Origins · 1 minJacques-Arsène d'Arsonval and how this was discovered

Jacques-Arsène d'Arsonval

1851–1940 · Collège de France, Paris

Jacques-Arsène d'Arsonval trained as a physician, ran the Collège de France's biophysics laboratory, and needed something medicine could not buy in 1882: an instrument sensitive enough to read the microampere currents of nerves and muscle. His answer — a light coil suspended in the field of a permanent magnet, twisting against a hairspring in exact proportion to the current — became the moving-coil galvanometer, ancestor of every needle meter of the following century. Wired with a shunt it became the ammeter; with a series multiplier, the voltmeter — and because the working element is a coil of fine wire, every descendant inherited a resistance it cannot shed. The loading errors drilled in this archetype are d'Arsonval's coil asserting its physical existence. He never patented the movement, spending his later decades on medical high-frequency currents while his meter quietly instrumented the electrical age.

History

1820

— Schweigger's multiplier coil turns Oersted's twitching needle into the first crude galvanometer

1882

— d'Arsonval perfects the moving-coil movement: linear scale, microampere sensitivity, and an unavoidable coil resistance

1890

s — shunts and multipliers standardise the conversion of one movement into ammeters and voltmeters of any range

1920

s — vacuum-tube voltmeters push R_V to megohms, shrinking loading errors a thousandfold

1970

s — digital multimeters make 10 MΩ input resistance the default, and the divider ambush retreats to examination papers and high-impedance sensors

Going further

Texts. NCERT Physics Class 12, Ch. 3 · HC Verma Vol. 2, Ch. 32 (objective bank on meter errors) · CIE 9702 practical papers (loading, systematic error).

Provenance. Standard exam canon; every number recomputed: two-arrangement pair on a true 10 Ω ( $10.5\ \Omega$ / $9.99\ \Omega$ , +5% / −0.1%); catastrophe pair on 50 kΩ ( $8.33\ \mathrm{k\Omega}$ , −83% / +0.001%); crossover $\sqrt{0.5 \times 10^4} = 70.7\ \Omega$ ; divider ambush $12\ \mathrm{V} \to 4.0\ \mathrm{V}$ (10k/10k/10k) and $10\ \mathrm{V} \to 4.44\ \mathrm{V}$ (5k/5k/20k, −11%); correction example $I_V = 0.1$ A, $R = 20/1.9 = 10.53\ \Omega$ (−5% naive).

Prerequisites. Circuit Elements, Symbols & Diagrams · Resistors in Series & Parallel. Peers. Potential Dividers (the circuit this archetype ambushes) · Wheatstone & Meter-Bridge Numericals (the null-method escape from meter resistance) · Potentiometer Numericals (the zero-current voltmeter, perfected). Coming up. Galvanometer → Ammeter/Voltmeter Conversion in the magnetism chapter — how the shunts and multipliers are actually sized — and the "Redraw the Circuit" technique module, whose node-collapsing craft makes Step 1 of this archetype automatic.

If both arrangements are wrong, how does anyone ever measure a resistance correctly?

Three honest escapes. First, choose the lesser evil: pick the arrangement whose error is negligible for your $R$ (that is the whole crossover game). Second, correct arithmetically: if you know $r_A$ or $R_V$ , the theft is exactly computable and reversible — subtract $r_A$ , or un-parallel $R_V$ , as in the worked examples. Third, change the method entirely: null techniques like the Wheatstone bridge draw no current through the detector at balance, sidestepping meter resistance altogether — that story is told in Wheatstone & Meter-Bridge Numericals.

Why is an ammeter built with low resistance and a voltmeter with high resistance in the first place?

Each is designed to disturb its host minimally. The ammeter joins a branch in series, so any resistance it carries adds to the branch and lowers the current — hence: as low as manufacturable. The voltmeter joins in parallel, opening a new path that steals current — the higher its resistance, the smaller the theft. The design goals are opposite because the wiring topologies are opposite; the residual $r_A$ and finite $R_V$ are just how far real coils and multipliers can chase the ideals.

What actually happens if I wire them the wrong way round?

A voltmeter in series (huge resistance in the line) throttles the circuit to near-zero current — meaningless reading, no damage. An ammeter in parallel (tiny resistance as a bridge) is a short circuit: a huge current floods the meter, and the fuse — or the meter — dies. Hence the lab-manual prayer "ammeter in series, voltmeter in parallel," and the fuse behind every multimeter's current jacks.

My voltmeter reads 4 V where my calculation says 6 V. Which one is lying?

Neither. Your calculation described the circuit without the voltmeter; the voltmeter honestly reports the circuit with itself included. Across a 10 kΩ divider arm, a 10 kΩ voltmeter halves the arm's resistance and re-splits the divider — 4.0 V is the genuinely correct voltage of that loaded circuit. Want the unloaded 6 V? Use a meter with $R_V$ at least ten times the arm (droop under ~5%), or compute the correction, or measure with a null method that draws nothing at balance.

Is the percentage error formula R/R_V exact or an approximation?

An approximation, good when $R \ll R_V$ . The exact fractional error of the voltmeter-near arrangement is $-\dfrac{R}{R + R_V}$ ; for $10\ \Omega$ against $10\ \mathrm{k\Omega}$ that is $-0.0999\%$ versus the shortcut's $-0.1\%$ . The shortcut fails exactly where the arrangement itself fails — $R$ comparable to $R_V$ — which is your cue to have switched wiring anyway.

Do digital multimeters still have loading errors, or is this a needle-meter museum piece?

Alive and well, just smaller. A standard DMM presents about 10 MΩ on its voltage ranges — harmless across a 10 kΩ divider (0.05% droop) but ruinous across a 10 MΩ sensor (50% droop; pH probes and photodiode circuits live there). On current ranges the DMM inserts a shunt whose "burden voltage" is the modern $Ir_A$ — datasheets specify it precisely because milliohm shunts still bend low-voltage circuits. The decimal point moved; the archetype didn't.

Where does the galvanometer fit into all this?

The galvanometer is the raw d'Arsonval coil — the sensitive movement inside classic meters. A tiny parallel shunt turns it into an ammeter; a large series multiplier turns it into a voltmeter. The sizing formulas are their own examination industry, drilled in the Galvanometer → Ammeter/Voltmeter Conversion archetype coming in the magnetism chapter.