Why I Won’t Trust a Ferrite Stack to Casual Glue or Tape
Why I Won’t Trust a Ferrite Stack to Casual Glue or Tape
A ferrite stack that rattles inside its box is not finished RF engineering. It must stay aligned through winding, transport and temperature changes without rubbing the insulation or chipping the rings. That is why I prefer a controlled retention process over a quick drop of whatever glue is nearby or another turn of tape.
RF.Guru working definition: Common-mode current is the non-cancelling phasor-sum current in a specified set of conductors, evaluated at a defined cross-section and using a declared current-direction convention. In the intended differential transmission-line mode, the outgoing and return currents are equal and opposite, so their phasor sum is zero. When they do not cancel, the remaining current must close through another reference or return path—such as the outside of a coax shield, a mast, equipment chassis, station wiring, nearby structures, earth, the operator, or distributed coupling through the environment.
This broader working definition is especially useful in practical antenna systems. On transmit, non-cancelling current on the outside of the coax can make the feedline and connected structures part of the radiating antenna system unless that path is intentional, clearly defined and properly controlled—for example by providing the required return path and placing a suitable common-mode choke at the correct boundary.
The shortcut is familiar: add a drop of instant adhesive, wrap the rings with whatever tape is on the bench, or tighten a clamp until nothing moves. My objection is not to the existence of glue or tape. It is to treating temporary fixation as permanent mechanical design. A stack can pass an electrical check today and still move against its winding after transport or repeated heating. Keeping the geometry stable is part of keeping the RF assembly dependable.
The engineering position: for a permanent bonded stack, I favour a controlled structural joint, with a suitably specified two-part epoxy as a candidate—not an unidentified instant glue or general-purpose tape wrap. Its purpose is stable alignment and protection from movement, not better ferrite chemistry. A qualified clamp or banding system can solve the same mechanical problem; the cable still needs independent strain relief.
Why the Mechanical Choice Deserves Its Own Design
With complete toroids held in a planned position, winding placement does not have to follow a moving stack. A designed joint or holder can restrain relative movement before it becomes chipped ferrite or worn insulation. That is the useful advantage over casual retention: a defined load path and repeatable geometry, rather than hoping tack and wrapping tension remain unchanged.
For a bonded assembly, I want the contact area, alignment and bond line chosen deliberately. A suitable structural epoxy gives the designer options for working time, joint toughness and cure control; those options are the reason to consider it. They are not a guarantee that any two-part resin is gentle on ferrite. Make the mechanical choice first, then use measurements to check that the assembled component still does its electrical job.
The aim is not the strongest possible blob of glue. It is enough restraint to stop harmful movement without transferring unnecessary expansion, clamp load or cable pull into a brittle core. An assembly that needs routine separation may be better served by a purpose-designed removable holder.
Ferrite Is Strong in Compression and Vulnerable to Mishandling
Ferrite is a sintered ceramic. It is hard and wear-resistant, but brittle. TDK warns that mechanical shock, rapid temperature change, tensile loading and high static or cyclic stress can crack ferrite. Fair-Rite similarly treats chips and cracks as handling and acceptance issues whose electrical significance depends on size and orientation.
Stress is not only a mechanical concern. Magnetostriction and stress sensitivity mean that mounting forces can change permeability and therefore inductance or complex impedance. TDK's application guidance notes that higher core stress can reduce initial permeability. A core can remain in one piece yet move far enough electrically to matter in a narrow or high-impedance design.
That leads to two practical rules:
- Do not make a ferrite ring, bonded stack or split-core seam carry cable weight, connector torque or enclosure impact.
- Do not assume that a visually intact assembly has preserved its original AL, R+jX response, loss or thermal margin.
Closed Toroids and Mated Core Halves Are Different Problems
When complete toroids are stacked face to face, the main magnetic path remains inside each closed ring. Flux in one ring does not need to cross the adhesive layer into the next ring. The interface is therefore primarily a mechanical and thermal boundary, although stress, winding placement and proximity still influence the finished device.
An E-core, U-core, pot core or split ferrite is different. Two parts complete one magnetic circuit across their mating faces. Dust, coatings, tape, adhesive or an uneven clamp at that joint changes magnetic reluctance. Even a small unintended gap can alter AL, magnetising inductance, stored energy, leakage and flux distribution.
Do not transfer one rule between geometries: a thin film between two complete rings is not the same electromagnetic feature as material inserted between the mating faces of a split magnetic circuit.
Fast Fixture Time Is Not Long-Term Qualification
Ordinary hobby “super glue” is attractive because it fixtures quickly and wicks into close joints. That speed does not qualify it for a ferrite assembly. Surface coating, porosity, joint gap, humidity, cure depth, temperature, impact, peel load and differential thermal expansion all affect the result.
It is also inaccurate to say that every cyanoacrylate is inevitably brittle or unsuitable. Henkel publishes toughened cyanoacrylates with specified gap, impact, peel and operating-temperature properties. Those values belong to the exact adhesive and test substrates; they do not certify a ferrite choke or transformer. A product can perform well in one close-fitting ceramic joint and fail another assembly because the stress state, coating, cure or environment differs.
A qualified cyanoacrylate process therefore needs:
- confirmation that the adhesive bonds the actual bare or coated ferrite;
- a permitted bond-line range and surface-preparation method;
- cure conditions that reach the full joint rather than only the exposed edge;
- mechanical data relevant to shear, peel, impact and thermal cycling;
- moisture and temperature limits for the intended environment; and
- post-cure magnetic, insulation and mechanical acceptance checks.
Without those records, an instant adhesive remains a workshop experiment rather than a controlled production method.
A Controlled Epoxy Joint Is More Than Mixing Two Tubes
Two-part epoxies offer many combinations of toughness, modulus, cure time, shrinkage, glass-transition temperature, thermal conductivity and electrical properties. That makes epoxy useful—but it also prevents “use epoxy” from being a complete specification.
A very stiff bond can transfer differential thermal expansion into the ferrite. An overly thick bond can change a mated core gap, move a winding, add dielectric material near a high-field node or increase a thermal interface. A fast cure can bring different shrinkage, exotherm, working-time and void behaviour from a slower formulation. Ferroxcube's study concerns thin bonds on the outer legs of mated core halves, not arbitrary stacks of complete toroids. It discusses thermally induced stresses, cure and moisture ageing; several tested joints were strong enough that the specimen fractured in the ferrite. The result supports deliberate adhesive/process selection, not a universal glue winner for every core geometry.
Select the exact resin and hardener from the manufacturer's technical and safety data. Control mix ratio, dispense, bond thickness, cure temperature, cure time and fixture pressure. A trade name or “five-minute” label cannot replace those process limits.
Tape Can Insulate, Band or Merely Hold Temporarily
Electrical tape, polyester tape, polyimide film and glass-cloth tape are not interchangeable. Backing, adhesive, stretch, tensile strength, thermal index, dielectric performance, flame behaviour, chemical resistance and ageing all differ.
Some tapes are genuinely intended for coil covering, anchoring, banding or core and layer insulation. 3M, for example, specifies its Glass Cloth Electrical Tape 27 for dry electrical applications that include coil covering and banding. IEC 60454-3-8 defines performance categories for woven-fabric electrical tapes, while also making clear that application selection still depends on the user's actual requirements.
That is very different from wrapping a warm ferrite stack with general-purpose vinyl tape and assuming permanent clamp load. A pressure-sensitive adhesive can creep, flag or change adhesion with temperature and contamination; an elastic backing can relax; application tension can place unknown compression on the core; and multiple layers can change heat flow and winding clearance.
Use tape only when its datasheet and the assembly design support the intended function. If it is insulation, specify the dielectric, temperature, overlap and abrasion requirements. If it is banding, specify retained tension, creep, ageing and reject criteria. If it is temporary fixturing, ensure it is removed or formally accepted before service.
Clamps and Holders Need Controlled Pressure
A removable clip, spring, holder or compliant pad can offer excellent serviceability. Core manufacturers routinely provide clips and bobbins for particular shapes. The advantage is not that clamps are inherently gentle; it is that a designed system can control where force enters the ceramic and how it changes over temperature.
A hard point clamp or uneven screw can crack an edge. Excess pressure can alter the magnetic result, while too little allows fretting and impact. The holder material can soften, shrink or creep. Metal hardware near the core may also add eddy-current loss or change winding capacitance and electric-field clearance.
Follow the core manufacturer's approved clip, mating-surface and pressure guidance. Where a custom holder is necessary, spread load over suitable faces, include compliant support where justified, keep the winding and cable independently strain-relieved, and measure the assembled component rather than inferring success from clamp torque.
Mechanical Retention Is Also a Thermal Design
Ferrite loss, winding loss and nearby conductor loss create heat. Adhesive, tape, pads, potting and holders change the thermal path from each core to air, the enclosure and neighbouring parts. A bond can help share heat or form an insulating interface depending on its thickness and thermal conductivity. A wrap can protect insulation or reduce surface heat transfer. Neither effect is safely assumed.
Adhesive service temperature must not be confused with a ferrite temperature or RF power rating. Check glass-transition behaviour, modulus, creep, dielectric properties and chemical ageing below any decomposition limit. The complete assembly must stay within the limits of the core, adhesive or tape, coax, wire insulation, connectors, holder and enclosure.
RF and thermal safety: inhibit transmission before touching a choke or transformer. Ferrite and windings can remain hot, and a high common-mode impedance can support substantial RF voltage. Increase power only under controlled conditions with representative frequency, mismatch, waveform, duty cycle and cooling.
Insulation and Retention Are Separate Specifications
A material that holds cores together is not automatically adequate winding insulation. Creepage, clearance, dielectric strength, puncture, abrasion, flame behaviour, tracking, moisture and temperature must suit the voltage and environment. Adhesive squeeze-out can reduce a clearance; a sharp ferrite edge can cut film; a metal clamp can create a new capacitive or conductive path.
Conversely, an excellent insulating tape may not retain clamp force for the life of the product. Give each layer a declared job. If one tape or adhesive serves two functions, it must satisfy both sets of requirements in the finished geometry.
Qualify the Completed Assembly
A useful qualification sequence is straightforward:
- Define the environment. Record transport shock, vibration, temperature and humidity range, outdoor exposure, service interval and cable loads.
- Identify every material. Record ferrite manufacturer, part, coating and lot where available, plus adhesive or tape product, holder and insulation.
- Measure before retention. Capture AL or inductance where relevant and complex R+jX over the operating frequencies using a repeatable fixture.
- Control assembly. Record surface preparation, bond line, mix and cure, tape overlap and tension, or clamp position and pressure.
- Inspect after assembly. Look for chips, cracks, displaced windings, sharp-edge contact, voids, squeeze-out and lost clearances.
- Repeat the magnetic measurement. Compare the same reference planes and fixture after cure or clamping; investigate material changes rather than hiding them in tolerance.
- Apply relevant conditioning. Temperature change, humidity, shock and vibration severities must represent the intended product. IEC 60068 methods can provide test procedures, but the product specification still sets the required severity and acceptance criteria.
- Repeat inspection and R+jX. Mechanical survival alone is insufficient if inductance, impedance, loss or resonance has shifted outside limits.
- Run powered thermal tests. Use representative loads, mismatch, frequency, waveform, duty cycle, enclosure and cooling. Observe individual cores, bond or tape, winding and connectors until stable.
- Document serviceability. Define inspection, replacement and reject criteria, including what happens after an enclosure is opened or a bonded part is dropped.
IEC 60068-2-6, IEC 60068-2-27 and IEC 60068-2-14 provide standardized vibration, shock and temperature-change methods. Citing one of them is not a durability claim by itself: mounting, severity, duration, sample count, operating state and acceptance criteria must all be declared.
Primary manufacturer and standards sources
- TDK Ferrites and Accessories — application notes and mechanical-stress warnings
- TDK Ferrites and Accessories — processing, clamping and gluing notes
- Ferroxcube — Gluing of Ferrite Cores
- Fair-Rite — ferrite product life, cracks, chips and operating limitations
- Henkel — a toughened cyanoacrylate example with product-specific gap, impact and temperature data
- 3M — Glass Cloth Electrical Tape 27 applications and technical data
- IEC 60454-3-8 — woven-fabric pressure-sensitive electrical tapes
- IEC 60068-2-6 — sinusoidal vibration testing
- IEC 60068-2-27 — shock testing
- IEC 60068-2-14 — change-of-temperature testing
My bottom line: do not ask a casual drop of glue or a stretchy tape wrap to be the mechanical design of a permanent RF component. Choose a controlled bond or holder that keeps the ferrite and winding in their intended positions, with cable loads taken elsewhere. For a bonded stack, a suitable structural epoxy is a useful starting choice; its exact joint and cure still matter. The benefit is deliberate, repeatable retention. The tests confirm that benefit—they are not a substitute for choosing it.
Mini-FAQ
- Why prefer controlled bonding over a quick tape wrap? A designed bond controls contact, alignment and relative movement instead of relying on unspecified wrapping tension or tack. A suitable epoxy is one option, not a universal winner; qualified banding or a removable holder can also work.
- Is cyanoacrylate always unsuitable for ferrite? No. Suitability depends on the exact adhesive, ferrite surface, joint gap, cure, impact, peel, temperature, moisture and ageing requirements.
- Is epoxy always safer than instant adhesive? No. Epoxy modulus, shrinkage, bond thickness, glass-transition behaviour and thermal expansion can transfer damaging stress or alter a mated magnetic joint.
- Can electrical tape retain a ferrite stack? A specified tape may serve as insulation, anchoring or banding, but general-purpose tape cannot be assumed to retain clamp load through heat, vibration and ageing.
- Does adhesive between stacked toroids create an air gap? Not in the magnetic path of complete closed rings. Adhesive between the mating faces of core halves is different and can change magnetic reluctance.
- How tightly should ferrite cores be clamped? Use the core and holder manufacturer's approved mounting method. There is no universal clamp torque or pressure for every ferrite shape and material.
- Does the retention method determine the RF power rating? No. Power capability belongs to the completed core, winding, insulation, retention, connector, enclosure and cooling assembly under declared RF and thermal conditions.