For a small motor drawing only a few amps of full-load current, the standard 100:5 or 50:5 CT ratio delivers too little secondary current for a motor protection relay to reach an accurate pickup band, so the field fix is to pass the motor's supply conductor through the CT window multiple times to multiply the apparent primary current [S2].
Because the CT secondary current follows the ampere-turns balance (Is × Ns = Ip × Np), wrapping the cable through the window three times turns a 5 A motor current into a 15 A effective primary seen by the relay, which lands cleanly inside the 1–12 A typical sensing window of modern low-voltage motor relays [S2][S3].
Why the small-motor CT sizing problem exists
Compact IEC low-voltage motor control center drawers physically cannot accept the C400 or C800 relay-class CTs sized for medium-voltage switchgear; SEL's published case study for a 5,000-drawer Central Asia oil and gas facility targeted CTs small enough to fit inside removable drawers carrying loads from a few amps up to 2,500 A and short-circuit levels up to 20 kA [S2]. The most common instrument transformer secondary ratings are 120 V for voltage transformers (VTs) and 5 A for CTs, so relay sensing electronics are built around that 5 A secondary full-scale figure [S5].
When the motor's running current sits at the bottom of that 5 A envelope, ratio error, phase displacement, and CT saturation margin all collapse, because the CT is operating far below its designed knee-point voltage. ENC's published motor-protection CT, with primary input 0–400 A, secondary 0–200 mA or 0–10 V, ±0.2% ratio error, and 10×Iin overload rating, illustrates the typical parameter set engineers must reconcile when the actual motor FLA is only a few amps [S3].
Primary turns: the math, the wrap, the limits
For a single-conductor pass through a window CT, Np = 1, so secondary current equals primary current divided by the CT ratio. Increasing Np to 2, 3, or 4 by looping the cable through the window Np times raises the magnetizing ampere-turns by the same factor, which lets the same physical core sit at a higher point on its excitation curve and reduces ratio error at low primary current [S2][S4].
CT ratio error is specified under ±0.2% and phase displacement under ±15 minutes on commercial motor-protection CTs, with insulation resistance above 1,000 MΩ and insulation voltage above 3 kV [S3]. The practical wrap count ceiling is set by physical window diameter, lead dress, and the fact that the supply conductor must still be able to carry full-load current and short-circuit current without nuisance heating; SEL's selection technique keeps the effective secondary voltage during a fault below the CT's calculated knee-point voltage to avoid saturation that would cripple the instantaneous and short-time elements [S2].
Comparison: 3 ways to handle the small-FLA relay sensing problem
Option 1, multi-turn primary loop: cheapest, no extra hardware, only works above a minimum FLA of roughly 1–2 A, and burdens the cable with extra bend radius inside the drawer [S2]. Option 2, dedicated low-ratio CT (e.g. 10:5 or 25:5) wound specifically for the motor FLA: cleanest secondary waveform and best phase accuracy, but requires a CT stocking program and a different CT part number for every motor rating band on the same MCC [S2][S3]. Option 3, relay with a 1 A secondary input or with software-selectable CT ratio compensation: eliminates the wrap, but restricts the engineer to a specific relay family and is not always available in legacy protection relay retrofits [S2].
On cost per drawer, SEL's paper concludes the multi-turn primary technique on a stock 50:5 or 100:5 CT typically wins for the sub-50 A FLA range, because the same CT part can be re-used across the full drawer population with only the wrap count changing at the termination step [S2]. On accuracy, a dedicated low-ratio CT still wins because ratio error stays inside the published ±0.2% band even at 5% of nominal primary, whereas the multi-turn approach trades some accuracy for hardware simplicity [S3].
Field commissioning: prove it, do not assume it
After the wrap is dressed, primary-injection through the CT is the verification step that confirms the motor protection relay trips at the expected setting across the instantaneous, short-time, and earth-fault elements, because hand calculations on ampere-turns balance ignore the actual burden, lead resistance, and any residual CT saturation at fault current [S1][S2]. SEL's real-time simulation validation on a representative low-voltage motor relay is the same principle applied at the bench before the drawer ships to site [S2].
A secondary shorting jumper or shorting switch must remain across any unused CT secondary tap during the wrap-and-test process, because an open-circuited CT secondary with current flowing in the primary will generate the voltage needed to push the rated secondary current through whatever path exists, including the relay's input, the test set, or an arc to ground; field reports describe CTs that burned up in switchgear after a secondary was left open [S4]. The fix documented in one multi-tap CT ground-fault event was to relocate the ground wires and remove the short across the unused taps, after which W1 and W2 winding currents matched in the differential meter [S4].
Standards and definitions that govern the wrap
CTs for motor protection in electrical power systems are installed around one of the motor's supply conductors and provide an isolated, low-level current to the motor protection relay for monitoring and trip decisions on overcurrent, earth fault, and locked-rotor conditions [S3]. A residual connection of three phase CTs and a ground-fault relay is the standard wiring for low-resistance grounded systems on medium-voltage motors, with the same ampere-turns physics applying whether the motor draws 50 A or 5 A [S5].
Where the motor protection scheme requires differential coverage, device 87 on motors 750 kW and above on ungrounded systems, and on grounded systems where the applied ground-fault protection is not considered sufficient, the multi-turn primary loop on a window CT serves as the self-balancing differential current sensor, with the number of turns that must be short-circuited before detection occurs depending on the motor winding arrangement, the relay pickup, and the CT ratio [S5]. That same self-balancing window CT is what makes the multi-turn wrap technique viable for sub-750 kW LV motors, where the CT is acting as a phase monitor rather than a true differential element [S2][S5].
Limits and failure modes to plan for
Three failure modes recur in published field data. First, CT saturation at the calculated symmetrical fault current collapses the secondary waveform and delays or blocks instantaneous tripping, which is the reason SEL's selection procedure is built around the calculated knee-point voltage at the relay burden [S2]. Second, open-circuited CT secondaries, including the common practice of shorting unused taps on a multi-tap CT, can divert secondary current away from the intended relay input and either cause a trip on a healthy system or prevent a trip on a faulted one [S4]. Third, treating the wrap count as a free parameter without re-checking the CT's continuous thermal rating (typically 1× to 2× FLA for metering CTs, higher for protection CTs) leads to insulation degradation at the cable entry point over years of thermal cycling [S2][S3].
For maintenance teams building the inspection plan around condition-based monitoring of motor protection assets, the practical signal is whether the relay's metered secondary current on each phase matches the calculated value Ip × Np ÷ Ns across the full load range; deviations above 1–2% in steady state are the early indicator of a wrap that has loosened, a CT that is approaching saturation, or a secondary wiring resistance drift that the wrap technique is more sensitive to than a dedicated low-ratio CT would be [S2][S3].
For related coverage, see Prefabricated Power Modules for Data Centers: 2026 Spec Reality.