Permalloy Transformer Cores: a practical, engineer‑first guide

Permalloy cores can feel a bit like a “secret menu” item in magnetics: not right for every job, but magical for low‑level signals, ultra‑low noise, and demanding instrumentation when nothing else quite measures up. This guide distills what matters in the lab and on the line: where permalloy shines, how it compares to ferrites, silicon steels, and nanocrystalline alloys, and how to spec and source cores that actually hit the numbers after heat treat and assembly. 

• What you’ll get from this article
  • A plain‑English map of the permalloy family (45/49/80% Ni and “supermalloy”)
  • Side‑by‑side property comparisons you can use in first‑pass selection
  • Heat‑treatment and handling checkpoints that make or break permeability
  • Sourcing language that vendors understand (ASTM A753, A596 ring tests)
  • Design gotchas around DC bias, lamination thickness, and magnetostriction

Permalloy, at its core, is a nickel–iron soft‑magnetic alloy. The most famous grade for transformer work is the 80% Ni variant (often called HyMu 80, Moly Permalloy, or ASTM A753 Alloy 4), prized for extremely high initial and maximum permeability and very low coercivity—attributes that let you move flux with tiny magnetizing forces and minimal distortion. 

The permalloy family at a glance

• Alloy 1 (≈45% Ni Fe–Ni): higher Bsat than 80% Ni, moderate permeability
• Alloy 2 “High Perm 49” (≈48–49% Ni): highest Bsat (~1.5–1.6 T) among Ni–Fe, good permeability; favored where headroom against saturation matters
• Alloy 3 (≈75–78% Ni, Cu/Cr additions): very high permeability, used more for shielding/specialty parts
• Alloy 4 “HyMu 80/Moly Permalloy” (≈80% Ni, ~5% Mo): the go‑to for ultra‑high permeability laminations and tape‑wound cores
• Supermalloy (≈75% Ni, ~5% Mo): even higher permeability at the cost of mechanical robustness and Bsat margin 

Permalloy’s appeal is simple: at low magnetizing forces (think audio, sensors, instrumentation), nothing couples flux quite as gently. The trade‑off is lower saturation flux density than steels and a need for careful heat treatment and handling. In practice, designers pick 80% Ni when they need vanishingly small excitation current, ultra‑low distortion at small signals, and minimal magnetostriction “singing.” 

• Where permalloy cores earn their keep
  • Small‑signal audio transformers (mic and line‑level), magnetic pickups, tape heads
  • Precision instrument transformers and transducers at 50/60 Hz to a few kHz
  • High‑attenuation shields built into transformer structures
  • Any low‑level interface where core noise, hysteresis, and magnetizing current must be minimized 

Quick comparison: permalloy vs. the usual suspects

The numbers below are representative of widely used grades after proper hydrogen anneal. Always confirm with supplier datasheets and your specific lamination thickness and heat‑treat path.

Sources underpinning the table: HyMu 80 and Alloy 49 from Carpenter and MuShield; classic permeability/Bsat ranges from Lee’s Electronic Transformers; ferrite and nanocrystalline ranges from vendor datasheets and application notes. Always reconcile to your vendor’s datasheet for the specific grade and thickness you’ll use. 

• Translating those numbers into choices
  • Choose Permalloy 80 when magnetizing current and low‑level linearity dominate and your flux density stays well below ~0.2–0.3 T in service.
  • Choose High Perm 49 when you need “permally‑like” behavior but can’t accept the 0.6–0.8 T ceiling of 80% Ni.
  • Stay with silicon steel for bulk power at 50/60 Hz; it’s cost‑effective and robust.
  • Favor ferrite above ~50–100 kHz; resistivity wins, losses are low, parts are compact.
  • Consider nanocrystalline for chokes/filters or when you want high μ and ~1.2 T Bsat in the 50 Hz–100 kHz window. 

Heat treatment and handling: where μ is won or lost

Here’s the uncomfortable truth: you don’t “buy” high permeability—you create it with the right anneal and you can destroy it with careless handling. HyMu 80 and related alloys require a hydrogen anneal (dew point typically below about −40 °C) at roughly 1100–1180 °C for a few hours, followed by controlled cooling. This step relieves stress, grows grains, and unlocks the permeability vendors quote. After final anneal, bending, punching, or even a firm knock can degrade μ; many shops perform the “perfection anneal” as the very last step, and package parts to avoid stress and stray magnetization during shipping. 

• Heat‑treat checklist for HyMu 80 laminations/toroids
  • Final hydrogen anneal after all forming, stamping, or welding
  • Verify furnace dew point (≤ −40 °C) and soak 2–4 h at ~1100–1180 °C
  • Control cool through 700–300 °C at a few °C/min (supplier‑specific)
  • Avoid mechanical shocks post‑anneal; re‑anneal if parts were stressed
  • Test rings per ASTM A596 to confirm permeability/coercivity targets 

Permalloy’s mechanical and magnetoelastic behavior also helps keep transformers quiet: the magnetostriction around 80–82% Ni crosses near zero, which reduces strain‑induced noise and helps with ultra‑low hum designs. The exact magnetostriction depends on precise composition and even minor alloying; published work pegs the “zero λ” near ~81.5% Ni. 

• Practical implications of “almost zero” magnetostriction
  • Less audible hum from magnetostriction compared to steels
  • Lower stress sensitivity, but not immunity—cold‑work still hurts μ
  • Composition tweaks (e.g., Mo, Cu) can shift magnetostriction slightly; lock the grade in your spec 

Lamination thickness, eddy currents, and why 0.1–0.2 mm matters

Eddy losses scale with the square of lamination thickness and frequency. If you halve lamination thickness, you can quarter the eddy‑current loss component (all else equal). That’s why audio‑grade permalloy laminations often live around 0.1–0.2 mm, and why tape‑wound toroids perform so well at low flux densities. Use the simple proportional form Pe ∝ f^2·B^2·t^2 for first‑order trade studies, then validate with your vendor’s core‑loss data. 

• Core‑form choices that pay off
  • Tape‑wound toroids minimize leakage and headroom loss; great for small‑signal
  • EI/C laminations are easier to assemble and cost less; specify insulation and stacking factor
  • Avoid air gaps unless you’re deliberately biasing; 80% Ni’s low Bsat means gaps eat headroom fast 

How to spec and source permalloy cores (so vendors don’t guess)

You’ll get better parts, faster, if your PO reads like a test plan. Include the alloy, product form, heat treatment, and the numbers you will actually measure on receipt.

• Sourcing checklist
  • Alloy and standard: “ASTM A753 Alloy 4 (HyMu 80) laminations, thickness X mm” or “ASTM A753 Alloy 2 (High Perm 49)”
  • Heat treatment: “Final hydrogen anneal per supplier practice achieving μ and Hc targets; provide furnace dew point, time/temperature”
  • Magnetic targets: “Ring test per ASTM A596; min μ at B=40 G; Hc max at B=5–10 kG; Bsat (≥ X kG)”
  • Mechanical/finish: insulation coating class, stacking factor, burr limits, flatness
  • Handling/packaging: nonmagnetic packing, avoid residual magnetization, maintain ID traceability to heat and anneal batch 

When you need examples of “real” parts, look at small‑signal audio units built on 80–85% Ni laminations: their wideband linearity at millivolt levels shows what the material can do when the flux stays small and the anneal is right. 

• Typical acceptance tests to run in‑house
  • DC ring test (A596) for μ and Hc on coupons from your lot
  • Low‑level B–H loop at your actual frequency
  • Sweep‑tone distortion at expected flux density (for audio)
  • Temperature drift of magnetizing current at operating B

Design patterns that work (and a few that don’t)

In low‑level audio (say a 600 Ω to 15 kΩ step‑up), an 80% Ni core lam stack or tape‑wound toroid lets you run milligauss to low‑gauss flux swings with negligible hysteresis contribution, yielding clean low‑frequency extension at sane sizes. Keep the peak flux density conservative—on the order of a few hundred mT at most for headroom—and avoid DC bias unless you gap the core (which sacrifices μ). For power/instrument transformers where flux runs higher, High Perm 49 offers the headroom to keep distortion down before saturation. 

• Common mistakes to avoid
  • Assuming catalog μ without matching the lamination thickness and anneal
  • Letting parts get bumped post‑anneal (μ drops silently)
  • Using 80% Ni where DC bias is unavoidable and no gap is provided
  • Skipping a ring test on receipt—lot‑to‑lot variation is real 

A note on competing materials

Nanocrystalline cores are outstanding for common‑mode chokes and some power magnetics thanks to high μ and ~1.25 T Bsat, but above ~100 kHz their eddy losses rise versus ferrites. Ferrites dominate at high frequency for precisely this reason. None of that makes them better or worse than permalloy—it simply means you should pick the tool that matches frequency, flux swing, and the signal levels you care about. 

• Quick rules of thumb
  • LF, tiny signals, lowest excitation: Permalloy 80
  • LF with more volts and current: High Perm 49
  • HF power conversion: Ferrite
  • Wideband chokes/filters or LF‑HF bridges: Nanocrystalline (validate losses at your f)