Induction motor testing, starting and speed control

In short

Testing tells you a squirrel-cage motor's equivalent-circuit parameters, losses, efficiency and temperature rise without having to load it fully; starting deals with the 5-7 times rated inrush current a motor draws when switched directly on line; speed control deals with changing the running speed of a machine that is naturally almost constant-speed. The three subjects are examined together because they all follow from the same equivalent circuit and torque-slip relation. In GATE, ESE and university exams this topic carries predictable numericals (no-load and blocked-rotor data → parameters, star-delta starter ratios, V/f calculations). On the job it is the daily work of maintenance, panel-design and drives engineers at ABB, Siemens, Schneider, Danfoss, L&T, Bharat Bijlee, Kirloskar, CG Power, NTPC, Tata Steel and every state DISCOM, where IS/IEC 60034, IS 12615 efficiency classes and BEE labelling apply.

Explained

Tests on a three-phase induction motor

• DC resistance test: measure stator resistance per phase with a DC source or Kelvin bridge; for a star winding R1 = (measured line resistance)/2, for delta R1 = 1.5 × measured line resistance. Correct to the reference temperature (75 °C or as per IS/IEC 60034-1).
• No-load (running-light) test: run the motor uncoupled at rated voltage and frequency. Slip is almost zero, so the rotor branch is nearly open-circuit; the input is mostly core loss + friction and windage + small stator copper loss. Gives the magnetising branch R0 (or Rc) and Xm, and power factor is poor (0.1-0.3).
• Blocked-rotor (locked-rotor) test: clamp the rotor, apply a reduced voltage (typically 20-30% of rated) to get about rated current. Slip = 1, the magnetising branch is neglected, so the input is almost all copper loss. Gives the series branch Rsc and Xsc; R2′ = Rsc − R1. Must be done quickly to avoid overheating; for deep-bar/double-cage rotors do it at reduced frequency (about 25% of rated, per IEC 60034-28 practice) because X is frequency-dependent while skin-effect-inflated R is not.
• Splitting X1 and X2′: no test separates them. Standard assumption X1 = X2′ = Xsc/2 for wound-rotor and NEMA Design A and D motors; for Design B and C the standards give other ratios (0.4:0.6 for B, 0.3:0.7 for C).
• Separation of no-load losses: repeat the no-load test at falling voltage; core loss ∝ V², friction and windage is constant. Plot no-load loss (minus stator copper loss) against V² and extrapolate to V = 0 — the intercept is friction + windage.
• Load test (brake test): load the motor with a rope brake, eddy-current or DC generator dynamometer; read V, I, W, speed and torque to plot efficiency, power factor, torque and current against output. Direct but needs full-load power and a big dynamometer.
• Slip measurement: stroboscope on a marked shaft; or connect a centre-zero DC milliammeter across one rotor phase of a slip-ring motor and count the slow needle swings — each swing cycle is one slip cycle, so s = (slip frequency)/(supply frequency).
• Other type tests: temperature-rise (heat-run) test and insulation-class check, high-voltage/dielectric test, insulation-resistance (megger, with polarisation index), vibration (IS 12075), noise (IS 12065), degree of protection IS/IEC 60034-5 (IP codes), and efficiency determination by summation of losses to IS 15999 / IEC 60034-2-1 which is the basis for the IE classes.
• Circle diagram: the locus of stator current with varying slip is a circle. Drawn from no-load and blocked-rotor data: plot the no-load current phasor and the (scaled) short-circuit current phasor, draw the circle through them with its centre on the perpendicular bisector lying on a horizontal through the no-load point. Reading off the diagram gives input, output, torque, slip, efficiency and maximum torque/output lines at any load. Output line joins the no-load point to the short-circuit point; torque line joins the no-load point to the point separating stator and rotor copper loss.

Why starting needs attention

• At standstill s = 1, the rotor is a short-circuited secondary, so the motor behaves like a short-circuited transformer: starting current = 5-8 times full-load current while starting torque is only 1.5-2.5 times full-load torque (poor, because the rotor power factor is bad at s = 1).
• Consequences: voltage dip on the feeder (lights flicker, contactors drop out), winding and cage heating, mechanical shock to couplings and belts. Hence DISCOM/CEA-linked supply conditions typically require a reduced-voltage starter above a threshold rating (commonly around 5 HP/3.7 kW, but the limit is set by each state's supply code — confirm the local code).
• Key fact used in every numerical: with a reduced-voltage starter, torque ∝ V² at a given slip, and motor current ∝ V. So you cannot cut inrush without cutting torque.

Starting methods

1. Direct-on-line (DOL): contactor + overload relay + fuse/MCCB. Full torque, full inrush. Used up to about 5-7.5 kW, and for anything needing full starting torque. Cheapest and most robust.
2. Stator-resistor or reactor starting: series resistance/reactance in the stator, cut out as the motor runs up. Current reduced to x·I_DOL, torque to x²·T_DOL. Resistors waste energy; reactors are used for large motors.
3. Star-delta starting: the motor must be delta-connected in normal running with all six leads out. Started in star, so phase voltage = V_L/√3 → line current = 1/3 of DOL and starting torque = 1/3 of DOL torque. Changeover by timer; open-transition causes a current spike, so closed-transition (Korndorfer-type) or 3-contactor star-delta with transition resistors is preferred. Only suitable for light-starting loads (fans, pumps, lathes).
4. Autotransformer starting: a tapped autotransformer (typically 50/65/80% taps) feeds the stator. At tapping x: motor current = x·I_DOL, supply line current = x²·I_DOL, torque = x²·T_DOL. Better torque per ampere of line current than star-delta; tapping can be chosen; costlier.
5. Rotor-resistance starting (slip-ring motors only): external resistance in the rotor circuit through brushes and slip rings. Adding R2 increases starting torque (maximum torque can be made available at start) and reduces starting current — the only method that improves both. Used for crane hoists, crushers, ball mills, mine winders. Resistance is cut out in steps as speed builds.
6. Soft starter (solid-state): anti-parallel thyristors per phase ramp the voltage (and hence torque) up over a set time, with current limit, kick-start and soft-stop features. Compact, no moving contacts, but it only reduces voltage — it cannot raise torque — and it injects harmonics during the ramp; usually bypassed by a contactor when up to speed.
7. VFD start: a variable-frequency drive starts at low frequency and low voltage, so full torque is available at little more than rated current (typically 1.1-1.5 × FLC), a small fraction of DOL inrush. The best start, and gives speed control as a bonus; costliest.
8. Protection that accompanies a starter: thermal overload relay (or electronic motor-protection relay) set near FLC, single-phasing/phase-unbalance protection, under-voltage release, earth-fault protection, and for large motors stator differential, locked-rotor and winding-RTD protection.

Speed control from the stator side

• Speed N = (120f/P)(1 − s), so only f, P, s (and voltage, which changes s) can be changed.
• Stator voltage control: reduce V by autotransformer or AC voltage controller. Torque ∝ V², so speed falls only on a load whose torque rises with speed (fans). Narrow range, poor efficiency, high slip losses and derating; used for small fans and pumps.
• V/f (constant-flux) control — the industry standard: keep V/f constant so that flux φ ∝ V/f stays at its design value; the torque-slip curve shifts bodily along the speed axis with nearly constant maximum torque. Below about 10-15 Hz, add a voltage boost to offset the I·R1 drop. Above base frequency, voltage is already at maximum, so flux falls — a constant-power (field-weakening) region with falling torque. Implemented by a PWM VSI drive (rectifier → DC link → IGBT inverter); scalar V/f for pumps and fans, vector/field-oriented control (FOC) or direct torque control for high-dynamic loads, with closed-loop or sensorless speed feedback.
• Constant-current / slip-frequency control is used in some drives; frequency control alone without raising voltage saturates nothing but reduces torque, while raising f without V weakens flux.
• Pole changing: switch the stator winding to a different pole number — two separate windings, or a consequent-pole / Dahlander connection giving 2:1 speed ratio, or pole-amplitude modulation (PAM) for ratios like 4:6. Gives only fixed, discrete speeds (e.g. 1,500/750 rpm) but is cheap and efficient — two-speed cooling-tower fans, lifts, washing machines.
• Supply-frequency note: Indian supply is 50 Hz, so a 4-pole motor's synchronous speed is 1,500 rpm; drives commonly go up to 2-3 × base frequency subject to bearing and mechanical limits.

Speed control from the rotor side (slip-ring motors)

• Rotor-resistance control: increasing R2 increases slip at a given torque; maximum torque is unchanged (it does not depend on R2) but occurs at higher slip. Simple, good torque, but slip power is burnt as heat, so efficiency falls roughly as (1 − s). Used for intermittent duty: cranes, hoists, stirrers.
• Slip-power recovery: instead of wasting rotor slip power, rectify it and return it to the supply.
  • Static Scherbius / Kramer (sub-synchronous converter cascade): rotor diode bridge → DC link → line-commutated inverter → supply transformer. Converter rating is only the slip-power rating (say 20-30% of motor rating), so it is economical for large fans, pumps and compressors with a limited speed range. Modern wound-rotor wind generators (DFIG) use the same idea with a bidirectional converter.
  • Kramer drive returns the power mechanically through a DC machine on the same shaft; largely historical.
• Cascade (concatenation) connection: two motors on one shaft, the slip power of the first feeding the second. Cumulative cascade speed N = 120f/(P1 + P2), differential cascade N = 120f/(P1 − P2). Obsolete in practice, still examined.
• Injected-EMF (slip-EMF) control: inject a voltage of slip frequency into the rotor in phase or antiphase to raise or lower speed above/below synchronous.

Indian standards, codes and practice

• IS/IEC 60034-1 — rating and performance: duty types S1 to S9, temperature-rise limits by insulation class, tolerances, voltage/frequency operating zones. IS/IEC 60034-5 (IP), -6 (cooling, IC codes), -7 (mounting, IM codes), -8 (terminal markings), -12 (starting performance), -30-1 (IE efficiency classes).
• IS 12615 — energy-efficient three-phase squirrel-cage induction motors, defining IE1 (standard) to IE4 (super-premium); IE2 has been the mandatory minimum under BEE/Energy Conservation Act, 2001 (as amended in 2022) notifications, with a move to IE3 for many ratings — the exact mandatory level and ratings covered change by notification, so verify the current BEE/Ministry of Power order before quoting it.
• IS 12065 (noise limits), IS 12075 (vibration severity), IS 15999 / IEC 60034-2-1 (loss-summation efficiency testing), IS 8789 (performance values for general-purpose motors), IS/IEC 61800 series for adjustable-speed drives (61800-3 EMC, 61800-5-1 safety, 61800-9 energy efficiency).
• Older composite specification IS 325 for three-phase induction motors has been largely overtaken by the IS/IEC 60034 series; check its current status before citing it in a tender.
• Installation: CEA (Measures relating to Safety and Electric Supply) Regulations and IS 732 / National Electrical Code (SP 30) govern motor circuit sizing, earthing and protection; harmonic limits for VFD installations follow IEEE 519 / the DISCOM's own code, usually handled with line reactors or 12-pulse/active front ends.

Key points

• No-load test - rated V, no load, s ≈ 0: gives magnetising branch (R0, Xm), core loss and friction-windage loss; power factor is very low.
• Blocked-rotor test - reduced V, s = 1: gives Rsc and Xsc, hence R2′ = Rsc − R1 and leakage reactance.
• X1 = X2′ = Xsc/2 - a standard assumption, not a measurement.
• DC test - R1 = half the measured line resistance for star, 1.5 times for delta.
• Circle diagram - current locus from no-load and short-circuit data; reads off torque, output, slip and efficiency graphically.
• Starting current - 5-8 × FLC on DOL, because at s = 1 the rotor is a short-circuited secondary.
• Torque ∝ V² at a given slip; motor current ∝ V — the key trade-off in all reduced-voltage starters.
• Star-delta - 1/3 of DOL line current and 1/3 of DOL torque; needs a delta-rated motor with six terminals.
• Autotransformer at tap x - line current x²·I_DOL, torque x²·T_DOL; better torque per line ampere than star-delta.
• Rotor-resistance starting - the only method that raises starting torque while cutting current; slip-ring motors only.
• Maximum torque is independent of R2; R2 only decides the slip at which it occurs (s_maxT = R2/X2).
• Soft starter - thyristor voltage ramp: smooth start, no torque gain, harmonics during ramp.
• V/f control - constant flux, constant maximum torque below base speed; constant-power field weakening above it; IR boost at low frequency.
• Pole changing - discrete speeds only (Dahlander 2:1, PAM other ratios), but efficient and cheap.
• Slip power = s × air-gap power; rotor-resistance control wastes it, slip-power recovery returns it to the line.
• Converter rating in a Scherbius drive is only the slip power, not the full motor rating — why it suits large limited-range fan/pump drives.
• IE classes come from IS 12615 / IS/IEC 60034-30-1, tested by loss summation under IS 15999.

Formulas and examples

Example 1 - Equivalent circuit from test data. A 400 V, 50 Hz, 4-pole, star-connected squirrel-cage motor: DC test R1 = 0.6 Ω/phase. No-load: 400 V, 6.0 A, 1,200 W. Blocked rotor: 100 V, 20 A, 1,500 W.

• No load: cos φ0 = 1200/(1.732 × 400 × 6) = 1200/4156.9 = 0.289, so sin φ0 = 0.957. V_ph = 400/1.732 = 231 V.
• Xm = 231/(6 × 0.957) = 231/5.744 = 40.2 Ω; R0 = 231/(6 × 0.289) = 231/1.732 = 133.4 Ω.
• Fixed loss = 1,200 − 3 × 6² × 0.6 = 1,200 − 64.8 = 1,135 W (core + friction and windage).
• Blocked rotor: Rsc = 1500/(3 × 20²) = 1.25 Ω; Zsc = (100/1.732)/20 = 57.74/20 = 2.887 Ω; Xsc = √(2.887² − 1.25²) = √(8.335 − 1.563) = 2.60 Ω.
• R2′ = 1.25 − 0.6 = 0.65 Ω; X1 = X2′ = 2.60/2 = 1.30 Ω.

Example 2 - Starter comparison for the same motor (full-load current 15 A).

• DOL current at 400 V = 20 × (400/100) = 80 A = 5.33 × FLC.
• Star-delta (if the motor were delta-rated for 400 V): line current = 80/3 = 26.7 A, torque = 33% of DOL torque.
• Autotransformer at the 65% tap: motor current = 0.65 × 80 = 52 A, line current = 0.65² × 80 = 33.8 A, torque = 0.4225 × DOL = 42% of DOL torque.

Example 3 - Rotor resistance for maximum starting torque. A slip-ring motor has R2 = 0.03 Ω/phase and standstill X2 = 0.30 Ω/phase.

• s_maxT = 0.03/0.30 = 0.10, i.e. maximum torque occurs at 10% slip.
• a = 0.10, so Tst/Tmax = 2(0.1)/(1 + 0.01) = 0.2/1.01 = 0.198 — only ~20% of maximum torque at start.
• For maximum torque at standstill, total rotor resistance must equal X2: R2 + Rext = 0.30 → Rext = 0.27 Ω/phase (referred to the rotor).

Example 4 - V/f control. A 415 V, 50 Hz, 4-pole motor runs at 1,440 rpm at rated torque. Required speed ≈ 840 rpm at the same torque.

• Rated slip speed = 1,500 − 1,440 = 60 rpm. At constant torque and constant flux, slip speed stays 60 rpm.
• Required Ns = 840 + 60 = 900 rpm → f = 900 × 4/120 = 30 Hz; V = 415 × 30/50 = 249 V.
• Slip = 60/900 = 0.0667 (6.7%). Shaft power = rated × 840/1,440 = 58.3% of rated — torque is held, power scales with speed.
• A centrifugal pump at this speed would need only (840/1440)³ ≈ 20% of rated power, which is why VFDs save so much on pumps and fans.

Example 5 - Power flow and efficiency. A 3-phase motor draws 50 kW at 415 V, 50 Hz, 4 poles; stator copper loss 2 kW, core loss 1 kW, slip 4%, friction and windage 0.8 kW.

• Pag = 50 − 2 − 1 = 47 kW; rotor copper loss = 0.04 × 47 = 1.88 kW; Pm = 47 − 1.88 = 45.12 kW.
• Shaft output = 45.12 − 0.8 = 44.32 kW; efficiency = 44.32/50 = 88.6%.
• ωs = 2π × 1500/60 = 157.08 rad/s → developed torque = 47,000/157.08 = 299 N·m; speed = 1,500 × 0.96 = 1,440 rpm, ω = 150.8 rad/s → shaft torque = 44,320/150.8 = 294 N·m.

Where it is used

• Panel design and switchgear: selecting DOL/star-delta/soft starters and overload relays for pumps, compressors and conveyors — core work at L&T Switchgear, Schneider, Siemens, C&S and thousands of panel builders.
• Drives and automation: commissioning VFDs (ABB ACS, Siemens SINAMICS, Danfoss FC, Delta, Fuji) on cooling-tower fans, ID/FD fans, boiler feed pumps and extruders; parameter setting, V/f versus vector mode, auto-tuning.
• Energy efficiency and audits: BEE-certified energy auditors replace oversized IE1 motors with IE3/IE4 units and fit VFDs on throttled pumps; measured savings justify the capex in rupees per year.
• Plant maintenance: no-load and blocked-rotor checks, megger/PI tests, current-signature and vibration analysis, temperature-rise checks after rewinding — at NTPC, Tata Steel, JSW, refineries, cement plants and municipal water works.
• Traction and heavy drives: Indian Railways' three-phase locomotives (WAG-9, WAP-7) use VVVF-controlled cage induction motors; metro rolling stock the same; cranes and mills still use slip-ring motors with rotor-resistance or slip-power-recovery drives.
• Manufacturing and test floors: routine and type testing of motors to IS/IEC 60034 and IS 12615 at Bharat Bijlee, CG Power, Kirloskar Electric, Havells, with NABL-accredited test labs.

Interview questions

• Q: Why does the no-load test give the magnetising branch? A: At no load the slip is nearly zero, so R2′/s is very large and the rotor branch carries almost no current; the input current is essentially magnetising plus loss current.
• Q: Why is the blocked-rotor test done at reduced voltage? A: At s = 1 the impedance is only the series leakage impedance, so rated voltage would force 5-8 times rated current and burn the windings; reduced voltage gives about rated current.
• Q: How do you separate core loss from friction and windage? A: Run the no-load test at several voltages, subtract stator copper loss, plot the remainder against V² and extrapolate to zero voltage — the intercept is friction and windage, which is speed- not voltage-dependent.
• Q: Why can star-delta starting not be used for a crusher? A: It gives only one-third of DOL starting torque, which is far less than a high-inertia, high-friction load needs; the motor would stall in star.
• Q: At the same supply line current, which gives more torque — star-delta or autotransformer? A: The autotransformer, because the motor sees x·V while the line sees only x² of DOL current; at 33% line current the autotransformer tap is 0.577 and gives the same 33% torque, but any higher tap gives more torque per line ampere with a selectable trade-off.
• Q: Does adding rotor resistance change the maximum torque? A: No. Tmax depends only on V1, R1 and leakage reactance; rotor resistance only moves the slip at which it occurs.
• Q: Why keep V/f constant in a variable-frequency drive? A: Flux is proportional to V/f; holding it constant keeps the flux at design value, avoiding saturation at low frequency and loss of torque at high frequency.
• Q: What happens above base frequency in a VFD? A: The voltage is already at maximum, so flux falls as 1/f — the motor enters a constant-power, reducing-torque field-weakening region.
• Q: Why is a soft starter not a substitute for a VFD? A: It only ramps voltage during the start and then bypasses; it cannot change the running speed and gives no energy saving on a throttled pump.
• Q: What is slip power and why recover it? A: Slip power = s × air-gap power, the portion that appears in the rotor circuit; in resistance control it is wasted as heat, while a Scherbius drive feeds it back to the supply, keeping efficiency high over the control range.
• Q: Why is the converter in a slip-power-recovery drive much smaller than the motor? A: It handles only the slip power over the required speed range, typically 20-30% of motor rating.
• Q: What tests would you do on a motor that tripped on overload after a rewind? A: Insulation resistance and polarisation index, DC resistance balance between phases, no-load current and loss check, blocked-rotor current for shorted turns, and a heat-run with temperature rise compared against the insulation class limits in IS/IEC 60034-1.

Common mistakes

• Mixing line and phase quantities in test calculations — Zsc must use the phase voltage, and Rsc = W/(3I²) uses phase current.
• Forgetting that the blocked-rotor resistance includes R1, so writing R2′ = Rsc instead of Rsc − R1.
• Treating no-load input power as pure core loss and ignoring friction, windage and the small stator copper loss.
• Claiming star-delta starting reduces current to 1/√3 — it is 1/3 of the DOL line current (and 1/3 of the torque).
• Saying rotor resistance increases maximum torque; it only shifts the slip at which maximum torque occurs.
• Changing frequency in a drive without scaling voltage — lowering f alone saturates the core, raising f alone weakens the flux and torque.
• Quoting a withdrawn or superseded IS number in a report; check the current IS/IEC 60034 or IS 12615 edition and the latest BEE efficiency notification.

Related topics

Induction motors · DC machines · Synchronous machines · Power electronics: inverters and converters · Electric drives and traction · Electrical machine design basics