NETA ETT Level 2 common mistakes
NETA ETT Level 2 · Understand the traps and how to avoid them.
Verifying absence of voltage by checking only phase-to-ground on one phase (or on the phase that 'looks' like the feed) and then applying locks and tags.
Candidates carry over the residential habit of a single non-contact tester check at a receptacle and never build the habit of testing all conductors, all combinations, on the load side of the isolation point.
What to do instead
Use the test–verify–test sequence every time: test the meter on a known live source, test every phase-to-phase and phase-to-ground combination on the load side of the disconnect, then re-verify the meter on the known source. Write the sequence out before the exam and rehearse it until it is automatic.
Selecting PPE from a memorized 'category' or from the equipment's nominal voltage instead of reading the arc flash label on the equipment being worked.
Training material often presents the PPE table as the primary method, so candidates treat 480 V gear as inherently low hazard and skip the label/incident-energy value that actually governs the boundary and PPE.
What to do instead
Drill the rule that the label governs when present; the NFPA 70E tables are permitted only within their stated equipment and fault-current limits. Practice reading a sample label and extracting incident energy, boundary, and required PPE before answering.
Reading a normally closed auxiliary contact on a schematic as 'open' because the equipment is energized, then tracing the circuit down the wrong path.
Candidates conflate the electrical state of the circuit with the mechanical state drawn on the print; schematics are drawn in the de-energized, unactuated condition.
What to do instead
Before tracing anything, restate that every contact is drawn in its normal (de-energized, unactuated) position. Redraw the relevant rung in the actual energized state, then trace power flow from the source through the redrawn rung.
Applying single-phase relationships to three-phase circuits - dropping the √3 in power, current, or voltage calculations, or mixing line-to-line and line-to-neutral values in the same formula.
Single-phase formulas are learned first and are reinforced by everyday work; under time pressure the √3 term is the first thing dropped, and phase subscripts are read casually.
What to do instead
Write the three-phase relationships explicitly at the top of scratch paper (S = √3·V_LL·I_L, and V_LN = V_LL/√3) and label every quantity as line or phase before substituting. Sanity-check the answer against a known magnitude - a 480 V, 100 A three-phase load is about 83 kVA, not 48 kVA.
Using whatever DMM and leads are on the truck to measure at a service entrance or outdoor distribution equipment, ignoring the meter's CAT rating.
Candidates assume any meter rated for 600 V can be used anywhere 600 V appears, not recognizing that the CAT rating describes transient environment, not steady-state voltage.
What to do instead
Match the meter and leads to the location: CAT IV at the service entrance and outdoor feeders, CAT III at distribution panels and fixed equipment, CAT II only for cord-and-plug loads. Check the CAT marking on both the meter and the leads before answering, and treat an untested or mixed-rating set as unacceptable.
Recording an insulation resistance reading as-tested, at whatever winding and ambient temperature it happened to be, and comparing it directly to a 20 °C reference value or to a previous test.
The reading is a number on the display and the temperature correction step feels optional; candidates do not appreciate that IR roughly halves for every 10 °C rise.
What to do instead
Record test voltage, duration, winding temperature, and ambient with every reading, and apply the temperature correction to a 20 °C base before comparing to limits or to prior trending data. If temperature was not recorded, treat the comparison as invalid rather than guessing.
Measuring breaker contact resistance, bus joints, or switchgear connections with a two-terminal ohmmeter and comparing the result to micro-ohm acceptance limits.
The meter is available, the resistance is small, and the candidate does not account for lead and contact resistance adding tens of milliohms to a reading that should be tens of micro-ohms.
What to do instead
Use a four-wire (Kelvin) low-resistance ohmmeter with the correct test current for the device, and null or account for lead resistance. Recognize that any value in the milliohm range on a bus joint is dominated by test-lead error, not by the joint.
Passing a transformer on a turns-ratio test when the ratio is within tolerance, without checking phase-angle deviation, excitation current, or phase-to-phase comparison across all taps.
The ratio number is the visible output, so the test is treated as pass/fail on one number; vector group and excitation current are less familiar and are skipped under time pressure.
What to do instead
Evaluate the complete TTR result set: ratio on every tap against nameplate, phase-angle deviation against the nameplate vector group, excitation current compared phase-to-phase and to prior data, and any phase that is an outlier investigated even when its ratio is technically in tolerance.
Timing a circuit breaker with the test set at whatever control voltage is convenient, or reporting only total clearing time without separating contact time, travel time, and bounce.
Candidates treat the timing set as a stopwatch and do not recognize that the coil is specified to operate at a rated control voltage; low control voltage stretches the time and looks like a slow breaker.
What to do instead
Test at the nameplate-rated control voltage (or the specified percentage of it), record travel, contact, and bounce separately, and compare each to the manufacturer's curve. Treat any timing run at off-rated control voltage as invalid.
Performing a fall-of-potential ground resistance test with the ground grid still bonded to the utility neutral, or with current and potential probes spaced too close together.
Disconnecting the grid is extra work and the test still produces a plausible number, so the isolation step gets skipped; probe spacing is often eyeballed rather than measured.
What to do instead
Isolate the electrode under test from parallel paths, then space the current probe at least 5× the diagonal of the grid (or per the test method used) and place the potential probe at the specified percentage of that distance. Take readings at several potential-probe spacings and confirm the plateau before reporting a value.