Transformer core lamination types: EI, UI, step-lap, and wound cores compared

This article is about the part everyone skips: how the lamination stack you choose quietly decides losses, noise, and manufacturing pain.

We’ll stay practical and a bit blunt: EI, UI, step-lap, and wound cores as they show up on real purchase orders.

1. First, align on terms

Very short recap, just to sync vocabulary:

• EI laminations – E and I stampings stacked to form shell-type cores. The generic workhorse from EI-26 up to EI-240+ in most catalogues.
• UI laminations – U and I stampings for core-type constructions, often used where coil insertion needs to be easy and window access matters.
• Step-lap cores – not a new shape, but a joint strategy: the yoke joints are cut and overlapped in several small “steps” instead of one big straight butt or simple lap.
• Wound cores – a strip of electrical steel slit and wound into a closed ring, then sometimes cut/opened for assembly. Includes 3-phase 3D wound constructions and variants with amorphous or nanocrystalline steel.

Everything else in this article assumes you’re comfortable with flux density, magnetostriction, and loss separation. So we’ll go straight to trade-offs.

2. EI lamination stacks – the default that keeps winning tenders

Most designers start from EI not because it’s “best”, but because the ecosystem around EI is mature:

• Standard size series from about EI-26 up to EI-240 or more are available worldwide, with thicknesses 0.23–0.35 mm (CRGO) and 0.35–0.50 mm (CRNGO).
• Tooling is cheap, stamping lines are everywhere, coil suppliers know the windows by heart.
• Repair shops know how to take them apart and put them back together without thinking.

What EI stacks usually give you

• Cost leverage – simple stamping, simple stacking. Several manufacturers explicitly highlight EI cores as cost-advantaged versus more complex constructions.
• Flexibility – one lamination size can serve multiple ratings by tuning stack height and window fill.
• Reasonable performance – with CRGO, burr control, and a decent stacking factor, EI cores will meet most distribution and control transformer specs without drama.

Where EI starts to look tired

• Joint region is often butt-lap or simple lap, so local flux crowding, higher no-load losses, and more audible hum compared with a good multi-step-lap or wound core.
• For higher ratings, the rectangular cross-section and joint gaps can make it harder to squeeze out every last watt of loss.

What to watch when buying EI lamination stacks

If you’re sourcing from multiple lamination factories:

• Specify burr height and measure it; high burrs destroy stacking factor and can increase local loss.
• Lock down coating type and resistance; mixing T2/T4 or different insulation systems in one core can change inter-laminar behaviour.
• Don’t rely only on “M3/M4/M5” labels; ask for guaranteed W/kg at your test B and frequency, not just catalogue numbers.

EI is still the default choice when your KPI spreadsheet is led by purchase price, reasonable efficiency, and easy local sourcing.

3. UI lamination stacks – when mechanics, power density, or assembly drive the design

UI cores usually show up in projects where the winding shop says:

“We want to wind the coils separately and then just drop them on.”

That’s the UI story in one sentence.

Why people move from EI to UI

• Easier coil insertion – core-type layout, big central window. Good for compact power supplies, UPS, welders, and some special transformers.
• Compact footprint – for the same rating, a UI assembly can give better power density and more straightforward mechanical support.
• Less complex clamping – the yoke and limbs are simpler to clamp and strap.

But you pay for this in other currencies

• For shell-type designs with tight leakage control, EI often still behaves more predictably.
• Joint locations and flux paths differ; if you swap EI to UI without re-optimising window and limb sections, you can get surprise hot spots.

From a lamination-stack point of view, UI is just another stamping set, but your whole mechanical layout changes. Procurement needs to think about matching UI series (UI-30…UI-100 etc.) to planned winding tooling.

If most of your production is still EI, flipping one product family to UI can increase complexity: extra tooling, extra stock keeping units, separate QA jigs. Sometimes worth it; sometimes not.

4. Step-lap cores – when a joint method starts to matter more than shape

Step-lap isn’t a shape; it’s a stacking method for the joints.

Instead of one abrupt transition where the limbs meet the yokes, you have several short overlaps arranged like stairs. Each lamination is shifted a little; flux sees a smoother path.

Studies and supplier data are consistent on a few points:

• Step-lap joints reduce no-load loss compared with simple mitered or butt-lap joints in the same steel, because the local peak flux in the joint is lower.
• They also cut magnetostriction-driven vibration and audible hum, which helps with noise limits in urban installations.
• Multi-step patterns (3–5 steps) usually track better than two-step “non step-lap” versions that some older designs used.

There’s a twist though: at equal core flux, no-load current and its harmonic spectrum can behave differently. One comparative test showed lower RMS no-load current for butt-lap than step-lap in a particular case, while the harmonic profile was actually worse for step-lap.

So step-lap is not magic. It shifts where and how you pay.

Cost and process impact

• Cutting is more complex: tight length tolerances and careful notch positions are required.
• Stacking workers need training or jigs; wrong sequence destroys the expected flux smoothing.
• Scrap optimisation is trickier for steel coils.

From a lamination-supplier quote, you’ll usually see step-lap yokes as a clear line item with higher price per kg than straight-cut yokes.

Where it makes sense:

• Medium and large distribution transformers, especially where energy-efficiency rules give a monetary value to every watt of core loss saved.
• Projects with strict noise specifications.

Below roughly tens of kVA and with short annual operating hours, the step-lap premium often doesn’t pay back; for 24/7 utility gear it usually does, and quickly.

5. Wound cores – continuous path, different economics

A wound core is built by spirally winding an electrical-steel strip (CRGO, amorphous, or nanocrystalline) into a closed loop, then cutting, annealing, and sometimes re-joining. Geometry can be rectangular, oval, or 3D triangular for 3-phase units.

Why manufacturers invest in wound-core equipment:

• Continuous magnetic path – no butt joints, no stacked gaps. That means lower local flux peaks and often lower core losses compared with equivalent stacked cores.
• For a given rating, you can reach smaller volume and lower weight, especially with high-grade CRGO or amorphous steel.
• Good noise behaviour; the absence of conventional joints reduces vibration sources.

The technology has been pushed further with 3D wound cores for 3-phase transformers, giving more balanced magnetic circuits and still lower no-load losses and inrush currents.

What holds some factories back:

• You need specialised winding and annealing lines, which means capex and dedicated know-how.
• Repair and rewinding are not as straightforward as pulling apart EI stacks.
• Window shapes are less forgiving; designs must respect what the winding machines can physically do.

In practice, wound cores dominate in:

• Many oil-immersed distribution transformers where utilities buy on total owning cost.
• Some current transformers and metering transformers needing low magnetising current and high accuracy.

If you already run a lamination stamping business, moving into wound cores is almost a different industrial game.

6. EI vs UI vs step-lap vs wound: quick comparison

Here’s a compact view for B2B decision-making. Treat the ranges as typical, not absolute.

7. Choosing lamination stacks for a real project

Most tenders or internal specs boil down to a handful of design drivers:

• Rated power and duty cycle
• Loss penalties (often monetised)
• Noise limits
• Available manufacturing equipment
• Repair philosophy

Let’s walk through a few common patterns.

7.1 Low-to-mid power, cost-sensitive, high mix

Small control transformers, machine tool supplies, small isolation transformers.

• Core choice – EI stacks almost every time.
• Reason: stampers offer standard EI series (EI-26…EI-240+), with low unit cost and easy sourcing, and designers can tune window fill quickly.
• Extra effort goes into: choosing lamination thickness and grade that meets your internal loss target without over-specifying.

7.2 High volume SKU, moderate power, semi-standard just-in-time production

Think: the same 3 or 4 ratings, produced all year.

Here you start to see:

• UI cores if coil insertion speed and assembly ergonomics dominate.
• EI cores with better steel and decent joint practice if you want to re-use existing tooling.

The important thing is to standardise lamination stacks early and lock them into drawings, so procurement doesn’t start mixing similar-looking stacks from different suppliers with small dimensional or coating differences.

7.3 Utility and distribution transformers with loss penalties

Once a utility or regulator attaches a cost to each watt of no-load loss, core construction moves closer to the top of the spec.

In this zone you normally see one of:

• EI or UI-style geometry with multi-step-lap joints.
• Wound cores built from high-grade CRGO or amorphous steel.

The right answer depends on your factory:

• If you already own stamping presses but no wound-core line, step-lap joints are often the lowest-pain path to better loss and noise numbers.
• If your volumes justify capex and you compete on efficiency, investing in wound cores can pay off in both loss figures and material optimisation.

7.4 Projects with aggressive noise requirements

Hospitals, dense urban substations, some commercial buildings.

• First move is usually step-lap stacking, careful clamping, and consistent core earthing (single-point earthing, no surprises).
• Wound cores with suitable steel grades also help but can be overkill if noise limits aren’t extremely strict.

8. Buying lamination stacks: small details that quietly matter

Engineers usually specify Bmax and steel grade; lamination specialists live in smaller details that make or break the result.

Some points worth writing into your specs or inspection plans:

1. Lamination thickness vs frequency At 50–60 Hz, 0.23–0.35 mm CRGO is standard for high-efficiency designs; thicker CRNGO may be used when cost dominates. Staying consistent across product lines helps with stock and predictable performance.
2. Stacking factor targets Specify the expected net iron cross-section vs gross. Burrs, coating thickness, and stacking discipline all affect it. If you ignore this, all your core-loss calculations drift.
3. Joint strategy spelled out Don’t just say “CRGO EI core”. Say “CRGO EI core with 5-step lap joints on top yoke, 3-step on bottom” (as an example) and attach a sketch. This saves arguments later.
4. Annealing and stress relief Cold working during stamping or winding introduces stress that degrades magnetic properties. Make sure your lamination supplier defines annealing cycles, especially for wound cores and high-grade CRGO.
5. Mixing batches and suppliers For critical units, avoid mixing stacks from different batches or suppliers in the same core; subtle differences in steel, coating, or burrs can show up as funny hotspot patterns or noise.
6. Measurement, not just certificates If losses matter, consider having representative cores from each batch measured for no-load loss and magnetising current at nominal flux density, not just trusting steel mill guarantees. Studies show that joint methods and stacking can shift behaviour even when the material certificate is identical.

9. FAQ: transformer lamination stacks in everyday decisions