Transformer Calculations Explained
From the nameplate to amps, ratios, and sizing — with the distinctions that keep the numbers honest
1. Read the nameplate first
A transformer's rating is stated in kVA — apparent power — not in kW, because the transformer's own limits are voltage (insulation, core flux) and current (winding heating), regardless of the load's power factor. The nameplate also carries the rated voltages of each winding, the frequency, and usually the percent impedance.
Two habits prevent most mistakes. Keep rated voltages separate from measured operating voltages — the nameplate says what the transformer was designed for, a meter says what the supply is doing today. And note the voltage basis on three-phase units: nameplate voltages are line-to-line unless marked otherwise.
2. From kVA to amps
Rated current follows from the rating and the voltage. Single-phase: I = S / V. Three-phase: I = S / (√3 × V), with S the total three-phase rating and V the line-to-line voltage.
A worked example: a 75 kVA three-phase transformer with a 208 V secondary carries 75,000 / (√3 × 208) = 208.2 A per line at full load. The same arithmetic runs in reverse — enter any two of rating, voltage, and current and the third follows.
3. Line quantities are not winding quantities
In a three-phase transformer the connection matters. A wye winding sees the line-to-line voltage divided by √3; a delta winding sees it directly. Two different ratios therefore describe the same transformer: the line-voltage ratio and the winding turns ratio.
The classic case is 480 V delta to 208 V wye. The line-voltage ratio is 480 / 208 = 2.31 : 1, but the winding turns ratio is 480 / (208 / √3) = 4.00 : 1. Reporting only "ratio = 2.31" for such a unit describes the system voltages, not the windings that a turns-ratio test would measure.
4. Rated current is not actual consumption
The rated full-load current is the ceiling the transformer was designed to carry continuously — not the current that flows. The connected loads decide the actual current, and most transformers spend their lives well below rated current.
The same distinction guards the other direction too: a rated-current figure is not a breaker size, not a safety judgment, and not proof that a specific load will work. Checking a real load means comparing its apparent-power demand against the rating — the calculator's load task does exactly that comparison, with the demand entered, never guessed.
5. Sizing with explicit margins
Choosing a rating involves three separate decisions that must not blur together: the calculated load, an allowance for future growth (a percentage of the load), and a target utilization (how much of the transformer's capacity you plan to use).
The margin arithmetic with a 60 kVA load: 20% growth alone requires 72 kVA; an 80% utilization target alone requires 75 kVA; applying both requires 90 kVA. Three different answers from the same load — which is why the calculator keeps the three inputs separate and defaults every margin to zero, so the base result stays visible.
6. What the advanced checks tell you — and what they cannot
The advanced checks each run a small, published model and say so next to the result. Voltage regulation converts the manufacturer's own figure between no-load and full-load voltage — and asks which voltage the percentage is divided by, because published sources use both conventions and the same voltages give different percentages.
The fault-current check divides the rated current by the per-unit impedance — an infinite-source screening bound, not a site study: no source or cable impedance, no motor contribution. The loss model (fixed losses plus load losses scaled by the load fraction squared) reproduces the standard textbook results, including peak efficiency where copper loss equals core loss. The V/Hz indicator flags over-fluxing risk when supply frequency drops; it approves nothing.
Run your own numbers
The calculator runs every task from this guide — rated current, ratios, load checks, sizing with explicit margins, and the advanced checks — with the assumptions stated next to each result.