AdvanceReview
…Checking

Physics notesMechanicsUnits & Experiments

Independent topic page

Units, Errors & Experiments

Classify as dimensions, significant figures, percentage error or a named experiment in 15 seconds; then one propagation rule or least-count protocol finishes the stem.

Portion MechanicsFocus Steady marks from dimensional checks, significant figures, error propagation and named syllabus experimentsTarget 60–90sPortion share 32.3%Trend Falling share
This unit is the paper’s measurement language. Advanced mixes dimensional consistency, significant-figure rounding, random/systematic error and apparatus protocols (Vernier, screw gauge, pendulum-g, Young’s modulus, capillary, calorimeter, u–v, resonance column, Ohm’s law, meter bridge/PO box). Treat every stem as: what is measured, what is computed, and how uncertainty travels. One-line protocol: measure → correct → compute → propagate. If you skip a step, the option that looks neat is usually wrong.

How Advanced asks this

  • Check dimensional consistency or find an unknown exponent in a proposed formula
  • Propagate absolute/percentage error through a derived quantity
  • Significant figures after arithmetic or instrument reading
  • Vernier/screw-gauge reading with zero error; least-count traps
  • g by simple pendulum: T²–L slope, or percentage error from δT and δL
  • Young’s modulus by Searle, capillary rise, calorimeter heat balance, u–v lens/mirror, resonance column, Ohm’s law, meter bridge or PO-box resistance

Traps that leak marks

  • Treating a dimensionally correct option as physically proven
  • Forgetting the factor 2 on δT/T in g∝1/T²
  • Using observed reading without zero-error correction
  • Confusing absolute error with percentage error in options
  • Reporting calculator digits beyond significant-figure rules
  • Ignoring end correction in resonance column
  • Using one (u,v) pair when the experiment expects several or a graph
  • Calling a dimensionally correct option physically proven
  • Dropping the factor 2 on δT/T when g∝1/T²
  • Using observed reading without zero-error correction

01. A dimensionally homogeneous equation is necessary but not sufficient; dimensionless constants and trig arguments still need physical justification.

02. Significant figures report measurement precision, not calculator display. Multiplication/division follows the weakest relative precision; addition/subtraction follows the weakest absolute precision (fewest decimal places).

03. For a product/quotient P = k Aᵃ Bᵇ …, percentage error adds as |a|δA/A + |b|δB/B + … when relative errors are small and independent.

04. Named experiments ask you to know what is measured directly (least count, zero error) versus what is computed (slope, intercept, formula). Graph method beats single-point substitution when the syllabus expects it.

05. Treat every stem as a measurement pipeline: what is read, what is corrected, what is computed. Dimensions kill wrong formulas; they do not prove right ones.

06. Error propagation is a log-diff ledger: write the computing formula first, then map δ through powers and products with absolute exponents. Significant figures report instrument honesty, not calculator vanity.

07. Named experiments score when you know least count, zero error and what the graph slope actually equals—single-point substitution is often the trap option.

Official · 2026

Syllabus for this topic

Official syllabus points for Units, Errors & Experiments only — not the full Mechanics portion map. Portion hub: Mechanics · Mechanics analysis.

Syllabus on this page is topic-scoped: Units, Errors & Experiments. The parent portion (Mechanics) also covers other topics — use the portion hub for the full map.

Must cover (this topic)

  • Units and dimensions; dimensional analysis; least count; significant figures
  • Error analysis for listed experiments: Vernier, screw gauge, g by pendulum, Young’s modulus, surface tension by capillary, calorimeter specific heat, u–v focal length, resonance column, Ohm’s law, meter bridge / post office box

Gaps / watch-outs

No topic-specific exclusions flagged — finish the unit points below for Units & Experiments.

Source: https://jeeadv.ac.in/documents/jee-advanced-2026-syllabus.pdf

Unit 01

Units, Errors & Experiments

medium

Exam focusDimensional analysis, significant figures, least count, and named experiment error setups.

Syllabus points

  • Units and dimensions; dimensional analysis; least count; significant figures
  • Error analysis for listed experiments: Vernier, screw gauge, g by pendulum, Young’s modulus, surface tension by capillary, calorimeter specific heat, u–v focal length, resonance column, Ohm’s law, meter bridge / post office box

Do / check

  • Write least count and % error formula before computing
  • Know what each named experiment measures

Full Physics syllabus →

Cross-topic stems

Mixed joins for this topic

Advanced rarely asks a chapter in isolation. These are tagged multi-idea Physics stems that touch Units, Errors & Experiments — practise the knowledge bridge, not only the single-chapter drill.

No tagged mixed stem matched this topic’s keywords yet. Still revise the joins listed under Trends, and scan the mixed joins board for Mechanics / EM / Optics bridges.

All mixed joins →

Why methods matter

Common methods that change your score

Dimensional checks and error propagation decide many General stems before any apparatus story begins.

01

Dimension sieve first

Most wrong formulas die on dimensions alone; saves the 90-second budget for genuine physics.

Do thisAssign M,L,T (and I,θ if needed) to every symbol; kill options that fail homogeneity before algebra.

Why it works

Most wrong formulas die on dimensions alone; saves the 90-second budget for genuine physics.

See it once

If period T∝ρᵃ gᵇ λᶜ, match [T] to powers of density, g and wavelength; solve three equations for a,b,c.

02

Log-diff error map

Turns every power and product into a linear percentage-error ledger without memorising cases.

Do thisTake ln of the computing formula, differentiate, replace d→δ and take absolute values.

Why it works

Turns every power and product into a linear percentage-error ledger without memorising cases.

See it once

g=4π²L/T² ⇒ δg/g = δL/L + 2 δT/T; timing error usually dominates.

03

Zero-error before formula

Systematic bias is not random; it shifts the measured value before propagation.

Do thisCorrect every instrument reading for zero error, then apply least count; never correct the final answer instead.

Why it works

Systematic bias is not random; it shifts the measured value before propagation.

See it once

Positive zero error means the instrument reads high with jaws closed—subtract that from the observed reading.

04

Graph over single point

Reduces random scatter and matches how Advanced phrases pendulum-g, Ohm’s law and meter-bridge stems.

Do thisWhen multiple (x,y) pairs exist, extract slope/intercept matching the linearised law.

Why it works

Reduces random scatter and matches how Advanced phrases pendulum-g, Ohm’s law and meter-bridge stems.

See it once

T² = (4π²/g)L: slope m=4π²/g ⇒ g=4π²/m, not a lonely T at one L.

Processed examples

Sample questions · full working

Read the impact line first. Then notice which family/model was chosen. The solution is written the way your rough sheet should look in the hall.

01

bankFamily · Dimensional consistency

Sample question

If F=αv + βx with F force, v speed and x length, find [β]/[α].

This stem rewards early recognition of the dimensional consistency model.

Model chosenEquate dimensions of each term to force.

Processed solution

  1. [α][v]=[F] ⇒ [α]=MT⁻¹.
  2. [β][x]=[F] ⇒ [β]=MT⁻².
  3. [β]/[α]=T⁻¹.

AnswerT⁻¹

Takeaway: Each additive term must match the left-hand dimensions.

02

standardFamily · Named experiment setup

Sample question

A Vernier has main-scale least count 1 mm and 10 Vernier divisions equal to 9 main-scale divisions. Find the Vernier least count.

This stem rewards early recognition of the named experiment setup model.

Model chosenVLC = 1 MSD − 1 VSD.

Processed solution

  1. 1 MSD=1 mm.
  2. 1 VSD=0.9 mm.
  3. VLC=0.1 mm.

Answer0.1 mm

Takeaway: Least count is the difference of one main and one Vernier division.

03

twistFamily · Error propagation

Sample question

In a simple pendulum experiment, T = 2.00 s ± 0.01 s and L = 1.000 m ± 0.001 m. Find the percentage error in g from T=2π√(L/g).

This Advanced-level variation couples hidden constraints with the error propagation model.

Model chosenPropagate relative errors with the power of T doubled.

Processed solution

  1. From g=4π²L/T², δg/g = δL/L + 2(δT/T).
  2. δL/L=0.001 and δT/T=0.01/2.00=0.005.
  3. δg/g=0.001+2(0.005)=0.011.
  4. Percentage error is 1.1%.

Answer1.1%

Takeaway: Powers in the formula multiply relative errors.

Reusable models

Common question families

Advanced stems cluster. Name the family in 10–15s, write the fast model, then run the same path. Do not invent a new plan each time.

01

Dimensional consistency

Looks likeCheck equation, find [X], or compare options by dimensions

Fast modelWrite base dimensions for every symbol before algebra.

Same solve path

  1. List quantities and their dimensions.
  2. Equate powers of M, L, T (and A if EM).
  3. Solve for unknowns or reject inconsistent options.
  4. Sanity-check with a known special case.

Rough skeleton

  • [ ] table
  • power equate
  • result

Why faster

Recognition fixes the model before arithmetic begins.

Refuse this

Do not start blind expansion or casework before identifying the governing model.

02

Error propagation

Looks likeLeast count, % error, or combined measurement uncertainty

Fast modelUse relative-error addition with powers as multipliers.

Same solve path

  1. Write the governing formula.
  2. Take logarithmic differentials.
  3. Replace dx/x by relative/percentage errors.
  4. Insert least-count or given uncertainties.

Rough skeleton

  • formula
  • δf/f
  • % error

Why faster

Recognition fixes the model before arithmetic begins.

Refuse this

Do not start blind expansion or casework before identifying the governing model.

03

Significant figures and rounding

Looks likeReport a computed result to correct precision

Fast modelLimit digits by the least precise measured input.

Same solve path

  1. Identify limiting significant figures.
  2. Compute with one guard digit.
  3. Round only at the end.
  4. Match the asked reporting style.

Rough skeleton

  • limiting digits
  • compute
  • round

Why faster

Recognition fixes the model before arithmetic begins.

Refuse this

Do not start blind expansion or casework before identifying the governing model.

04

Named experiment setup

Looks likeVernier, screw gauge, pendulum-g, Y, capillary, calorimeter, optics, resonance, Ohm/bridge

Fast modelName what is measured and the working formula before numbers.

Same solve path

  1. Identify the apparatus and measurand.
  2. Write least count and the formula.
  3. Propagate error if asked.
  4. State the physical meaning of the intercept/slope if graphed.

Rough skeleton

  • setup
  • formula
  • LC/error

Why faster

Recognition fixes the model before arithmetic begins.

Refuse this

Do not start blind expansion or casework before identifying the governing model.

Hall timing

60–90s process

  1. 0–15s
    Decide: dimensions, sig-figs, error propagation, or named experiment protocol.

    Underline given data, unknowns and constraints. Name the mechanical model before recalling any formula; discard relations whose assumptions are already broken.

  2. 15–35s
    List measured quantities with least counts/zero errors; write the computing formula or graph relation.

    Draw axes, free-body or energy/impulse diagram. Write governing equations symbolically with consistent sign and reference.

  3. 35–75s
    Propagate errors or extract slope/intercept; apply zero correction before using a reading.

    Solve algebraically as far as possible; substitute numbers only once. Keep vectors, components and units on the rough sheet.

  4. 75–90s
    Check units/dimensions, number of significant figures, and whether systematic bias (zero error, parallax, end correction) was handled.

    Check dimensions, limiting cases and physical sense. For multi-correct, test every option independently rather than forcing one narrative.

After the model is locked

Quick formula bank

Dimensional check

[LHS] = [RHS]; trig/exp arguments dimensionless

Proposed formulas and option sieves

Relative/percentage error

δx/|x|; % error = 100 δx/|x|

Single measured quantity

Product/power propagation

δP/|P| ≈ |a|δA/A + |b|δB/B + … for P∝AᵃBᵇ…

Derived quantity from independent measurements

Sum/difference

δ(A±B) = δA + δB (maximum-error convention in JEE)

Addition/subtraction of measured values

Vernier least count

LC = 1 MSD − 1 VSD = (1 MSD)/n

n Vernier divisions match (n−1) or n main-scale divisions as stated

Screw gauge

LC = pitch / number of circular divisions; reading = MSR + CSR×LC − zero error

Diameter/thickness with zero correction

Pendulum g

T = 2π√(L/g) ⇒ g = 4π²L/T²; use slope of T² vs L

g by simple pendulum

Young’s modulus (Searle)

Y = (MgL)/(πr²ℓ) with ℓ extension; often from load–extension slope

Wire under tension experiment

Mental bank

  • Homogeneous ≠ correct; only necessary
  • Max % error: add relative contributions with |exponents|
  • δ(A±B)=δA+δB (JEE maximum error)
  • Least count sets the smallest meaningful digit
  • Zero error is systematic—correct readings first
  • Pendulum: g from slope of T² vs L
  • Meter bridge: R = S(l/(100−l)) at balance (wire 100 cm)
  • Resonance column: end correction L_e ≈ 0.3d; λ/4 ≈ L+L_e
  • Homogeneous ≠ true; only necessary
  • δP/|P| ≈ Σ |aᵢ| δAᵢ/Aᵢ for P∝∏Aᵢ^{aᵢ}
  • Zero error before any reading enters a formula
  • Pendulum g from slope of T² vs L, not one (L,T) pair

Make it automatic

Practice · checklist

Daily

  • Two dimensional-consistency sieves and two log-diff error propagations.
  • One Vernier/screw reading with zero error and one pendulum or meter-bridge numerical.

Weekly

  • One mixed set covering Young’s modulus, capillary/calorimeter and Ohm/PO-box.
  • Rewrite every experiment you miss as: measured → corrected → computed → uncertainty.

100% checklist

  • Model classified (dimensions / sig-figs / error / experiment)
  • Formula written before error map
  • Zero error and least count applied correctly
  • Relative-error exponents absolute-valued
  • Graph slope used when multiple points exist
  • Named-experiment assumptions recalled (end correction, balance condition)
  • Final digits match significant-figure or % demand
  • Units/dimensions of answer checked
  • Model classed: dimensions / sig-figs / error / experiment
  • Correction then formula then propagation, in that order
  • Final digits match LC or % demand; units checked
Mechanics hubAll Physics notesSyllabusSubject analysisMixed joins