Voltage Drop Troubleshooting

Voltage Drop Problems Quick Answer

Use this when the lights dim, a motor stalls, or your DMM reading refuses to match the spreadsheet.

1. Measured Voltage Drop Doesn’t Match the Calculator

What it looks like: Spreadsheet or the advanced calculator says ~2–3% on the design gauge. Under load your meter shows 6–10% at the receptacle, well pump, or welder outlet. Idle voltage looks almost normal.

Why it happens (extra R the formula never saw): Textbook drop is I × R_conductors. Real drop is I × (R_copper/Al + R_{every lug, splice, breaker, receptacle, and oxidised connection}). Any series contact resistance that was ignored in the model shows up as “mystery” volts once current flows.

Likely causes checklist:

• Loose set-screw or under-torqued lug — especially Al terminations and service/neutral bars. Heat cycles loosen them; resistance climbs without changing AWG.
• Corroded aluminium or outdoor/marine terminations — white oxide and moisture raise contact R. Copper–Al mixes without listed connectors and inhibitor compound.
• Wrong gauge actually installed — scheduled as 6 AWG, pulled as 8; or mixed colours left unmarked in long conduits. Strip and measure OD / read printing.
• Undersized or shared neutral not in the model — calculator assumed dedicated return; field circuit returns on a skinny multiwire neutral.
• Breaker / fuse holder / pin-and-sleeve contact resistance — source-side voltage healthy; load sees drop across a warm enclosure.

Fixes:

• De-energise safely and retorque to manufacturer values; look for discoloration, fretting, or melted plastic behind devices.
• On Al, clean, use listed connectors, and apply anti-oxidant where the listing requires it (Noalox-class inhibitor is common in those kits).
• Verify conductor size against the schedule and the print on the jacket — then re-run the calculator with actual AWG.
• Measure millivolt drop across suspect joints under load with a quality True-RMS meter when safe — high mV drop on a tight length of conductor is a contact story, not an ampacity chart story.

2. Circuit / Equipment Fails Even Though the Calc “Passes”

What it looks like: NEC-flavoured 3% / 5% recommendation is green on paper. Compressors hard-start, VFD trips on undervoltage, LED drivers flicker on tip-up cycles, welders bog only when other loads share the feeder.

What the calculator is not doing for you:

• Inrush / locked-rotor amps — steady-state FLC may pass; motors draw multiple × FLC at start. Voltage collapses for a few cycles and the starter “chatters” or the drive protects.
• Ambient / conductor temperature — hot attics and rooftops raise copper resistance (~0.4%/°C). Design calc often uses 75°C table R without field heat.
• Harmonics / non-linear current — True-RMS current is higher than a cheap meter’s average sensing; electronic loads heat neutrals on MWBCs.
• Shared neutrals and return paths — single-phase model ignores how return current shifts line-to-neutral volts on the weak leg.
• Source stiffness — generator, soft utility, or heavily loaded transformer adds source impedance the branch calc never includes.

Fixes:

• Re-run voltage drop at a realistic peak or start amp, not just nameplate continuous. Soft-starts / VFDs change the amp profile — still verify DC bus / undervoltage trip thresholds.
• Separate ampacity derating from voltage drop; both must pass. Use Voltage Drop vs Ampacity when the conflict is heat vs resistance.
• Measure source voltage under the same load condition as the load. A soft primary makes every “passed” branch calc look wrong at the tool.

3. Long-Run Feeds Still Sag (Generators, Wells, Detached Shops)

What it looks like: Detached garage welder, barn pump, or generator transfer still brown-outs even after “we upsized.” Spreadsheet used 200 ft when the trench is 200 ft one-way — or Al feeders were treated like copper.

Common field errors unique to long work:

• Entering round-trip length when the formula already multiplies by 2 (or vice versa).
• Ignoring Al resistivity (~1.6× copper) on USE-2 / URD / multiplex.
• Sizing only the final branch while the underground feeder eats most of the budget — stack feeder + branch in the advanced calculator’s combined mode.
• Generator AVR struggling: voltage at the panel looks OK unloaded, collapses under large step loads.

Fixes:

• Model feeder and branch separately; keep total within your design target (informational NEC notes cite ~3% branch / ~5% feeder+branch — verify with AHJ).
• For wells / irrigation, include motor start amps and elevation if control boards are voltage-sensitive.
• After upsizing, re-check conduit fill and ampacity derating — larger wire solves drop and can re-break bundling math. Details: long-run sizing guide.

4. Multi-Wire Branch Circuit Voltage Drop Mistakes

What it looks like: One countertop leg runs heaters fine; the other flats out or trips GFCI/AFCI. Neutral feels hot. Line-to-neutral volts unequal under load though both “circuits” share a handle-tied breaker.

Why MWBC math bites voltage drop: On a proper 240/120 MWBC, return currents partially cancel in the shared neutral. If phases land on the same leg, or a multiwire pair is broken into two unrelated circuits, the neutral carries more than the model assumed. Calculators that treat each 120 V circuit as an independent round-trip overcount or undercount depending on how you entered the neutral.

Likely causes:

• Both ungrounds on the same phase (no 240 V potential, doubled neutral current).
• Open or high-resistance shared neutral — line-to-neutral balloons on one leg (dangerous for appliances).
• Treating MWBC as two full L–N–L loops in software without reflecting cancellation.

Fixes:

• Verify opposite phases on a two-pole breaker; clamp neutral under simultaneous load.
• Never land mixed MWBC neutrals on a single-pole discontinuity without a bonded handle and proper identification.
• When diagnosing, measure each L–N at the load and L–L at the source — open-neutral signatures look like “voltage drop” until you see >120 V on one leg unloaded.

5. Where to Measure — and How Meters Lie

What it looks like: “We checked voltage — it’s fine.” Probe was at the panel with tools plugged in far away, or measured unloaded receptacles. Calc and field numbers will never reconcile.

Correct differential method:

• Energise the real load (rated tool, heater, motor).
• Measure V_source at the supplying breaker or feeder lugs (same phase).
• Measure V_load at the equipment terminals or receptacle under that same load.
• Drop ≈ V_source − V_load. Percentage = drop / nominal (or / V_source, be consistent).

Common measurement mistakes:

• Probing hot-to-ground instead of hot-to-neutral on 120 V circuits (misses neutral drop).
• Using an averaging meter on distorted waveforms — True-RMS matters for VFDs / welders.
• Long meter leads inductive kick; hold probes steady; don’t compare different meter brands without a sanity check.
• Recording panel voltage after the load staged off — soft sources recover and hide the sag.

Related how-to: formula and percent method live in How to Calculate Voltage Drop; code notes in NEC Voltage Drop Requirements.