Shop feeder overload, voltage sag and a high-resistance earth
A two-technician repair shop in Nashik runs from a single-phase 230 V domestic connection. The energy meter and main switch are 60 m from the shop, and the entire shop is fed by one 2.5 mm² aluminium twin-core cable in conduit, protected by a 16 A Type B MCB. The bench carries three 60 W soldering stations, one 700 W hot-air rework station, a 1 hp (746 W) air compressor and about 200 W of lighting and fans. No-load socket voltage is 228 V; with everything running except the compressor it reads 214 V. Whenever the compressor starts, the tube lights flicker, the rework station resets and the DVR reboots, and a clip-on ammeter shows a brief 30 A surge. An earth tester gives 18 Ω for the shop electrode. The owner now wants to add a 1.5 kW room air conditioner on the same circuit and says 'just fit a bigger MCB'.
1. Estimate the shop running current and check whether the existing 2.5 mm² aluminium feed meets a 3% voltage-drop limit. Show your working and state what you would recommend.
Read the worked answer
Load list (convert everything to input power)
- Soldering: 3 × 60 W = 180 W
- Hot air: 700 W
- Compressor: 746 W shaft output, allow about 1 kW input
- Lighting and fans: 200 W
- Total ≈ 2.1 kW, so I = 2100 / 230 ≈ 9 A (the motor runs at roughly 0.8 pf, so the current is a little higher than the wattage alone suggests).
Voltage drop on the existing feed (aluminium, ρ ≈ 0.028 Ω·mm²/m)
- R per conductor = 0.028 × 60 / 2.5 = 0.672 Ω
- Loop resistance (out and back) = 1.34 Ω
- Drop at 9 A = 9 × 1.34 = 12.1 V ≈ 5.3% of 230 V, well above the 3% (6.9 V) figure normally applied to a final circuit. This matches the measured 228 V no-load against 214 V loaded.
- At the 30 A starting surge: 30 × 1.34 = 40 V, so the terminal voltage collapses to about 190 V for a fraction of a second. That is exactly why the SMPS-fed rework station and the DVR reset while the tube lights dim; the equipment is not faulty.
Recommendation – replace the feeder with 6 mm² copper: R = 0.0175 × 60 / 6 = 0.175 Ω per conductor, loop 0.35 Ω, drop at 9 A = 3.2 V (1.4%) and at 30 A only 10.5 V. With the 1.5 kW AC added (about 6.5 A), design the feeder for 15–16 A and give the AC its own 20 A circuit. A bigger MCB is the wrong fix: it does not reduce the drop, and it would leave the 2.5 mm² aluminium (roughly 18–20 A capacity) without overload protection.
2. Explain the mechanism behind the compressor-start disturbance and give three remedies, ranked with your reasoning.
Read the worked answer
A 1 hp capacitor-start induction motor draws a locked-rotor current of roughly 5–8 times full-load current. Full-load current is about 4.5 A, so the 30 A surge measured is normal, not a fault. What turns it into a shop-wide problem is the shared, high-impedance feeder: 30 A through the 1.34 Ω loop drops 40 V, and every appliance on that feeder sees the dip. Equipment with a wide-input SMPS rides out small dips, but a 40 V dip lasting a few tens of milliseconds is enough to drop the internal rail below the reset threshold.
Remedies, ranked
- Reduce the source impedance – change the 2.5 mm² aluminium feeder to 6 mm² copper. This fixes the root cause for every load and also creates capacity for the AC.
- Separate the circuits – a dedicated circuit and Type C MCB for the compressor (Type C tolerates 5–10 × In magnetic trip, so normal starting does not trip it while a genuine short still does), and a separate power circuit for the bench electronics and DVR. Put sensitive loads on a small 1 kVA stabiliser or line-interactive UPS.
- Attack the motor itself or the operation – fit a soft starter or correct the start capacitor, and stop running the compressor while a rework job is in progress.
Order matters: the cable upgrade removes the cause, the circuit split stops the fault spreading, and the UPS only protects the last metre.
3. The earth electrode measures 18 Ω. Explain the risks, state a realistic target for this shop, and describe how you would improve and verify the earthing.
Read the worked answer
18 Ω is high for a shop with earthed metal enclosures and a water line. Under an insulation fault, the touch voltage on a bench enclosure is roughly I_fault × R_earth, and a high resistance also limits fault current so the 16 A MCB may never trip on magnetic or even thermal element quickly. For a large installation IS 3043 works towards about 1 Ω, but for a small single-phase shop a practical target is below 5 Ω, backed up by a 30 mA RCD, which is the real protection against electric shock.
Improvement – do not rely on the existing single pipe. Drive two or three 3 m galvanised iron pipes at least 3 m apart, connect them in parallel to the main earth bar, and backfill with bentonite or the traditional salt-and-charcoal layer so the pit holds moisture. Bond the bench frames, the SMPS enclosures and the DVR to the same bar with a single-point earth, so you do not create a ground loop. Re-test in dry weather, not after rain.
Verification – use a three-terminal fall-of-potential earth tester with the potential spike at 62% of the current-spike distance; a multimeter between line and earth proves nothing. Confirm with a loop-impedance test and an RCD trip-time test, and record the values in the maintenance log.