Ferrite Chokes and Cores
Ferrite is the most important material in the ham radio operator's RFI toolkit. It is a ceramic compound that looks unremarkable — a dull gray or black ring — but has magnetic properties that make it uniquely useful for suppressing interference at radio frequencies. Wind a few turns of coaxial cable through a ferrite toroid and you have a common-mode choke that can reduce feedline interference by 20 to 40 dB. Snap a ferrite clamp onto a USB cable and you can reduce computer noise getting into your receiver by 10 to 20 dB.
The challenge is that ferrite is not a single material. There are dozens of ferrite formulations, each optimized for a different frequency range, and using the wrong mix for your application can produce no useful suppression at all. A ferrite core optimized for 1 MHz will be nearly useless at 30 MHz. A core optimized for VHF will provide only weak suppression on 80 meters. Getting the right mix for the right frequency is what separates effective ferrite use from expensive frustration.
What Is Ferrite and Why Does It Work?
Ferrite is a ceramic magnetic material made by combining iron oxide (Fe₂O₃) with one or more other metal oxides — typically zinc, manganese, nickel, or cobalt — and sintering the mixture at high temperature into a hard, dense ceramic. The result is a material with a very high relative magnetic permeability (µr), which is the measure of how much better it concentrates magnetic flux compared to empty air. Depending on the formulation, ferrite permeability ranges from about 40 (for high-frequency nickel-zinc mixes) to over 15,000 (for some low-frequency manganese-zinc mixes).
When a wire passes through a ferrite core, the core adds inductance to that wire. The inductance is proportional to µr multiplied by N² (the square of the number of turns) and divided by the path length of the magnetic circuit. More turns, or a higher permeability core, gives more inductance. At RF frequencies, this inductance appears as impedance — specifically, a complex impedance with both an inductive (reactive) component and a resistive (lossy) component. The resistive component is the key to understanding why ferrite chokes are effective.
The lossy behavior of ferrite at RF frequencies is called magnetic loss, and it is intentional. Ferrite is not a lossless magnetic core like a high-quality inductor core at its operating frequency. Instead, for common-mode choke applications, we want the ferrite to be lossy — to convert some of the common-mode RF energy into a small amount of heat rather than reflecting it back as a reactive impedance. A purely reactive choke would present high impedance but would also create a resonance with parasitic capacitances that could actually make things worse at certain frequencies. A lossy ferrite choke presents broad-band resistive impedance that simply absorbs common-mode energy over a wide frequency range without creating resonances.
The crucial advantage of a ferrite common-mode choke is its selectivity. When you wind coaxial cable through a ferrite core, the differential-mode signal inside the coax creates equal and opposite magnetic fluxes in the core — they cancel exactly, and the core has no effect on the transmission line mode inside the cable. Only common-mode current, flowing in the same direction on both conductors simultaneously, creates net flux in the core. The ferrite therefore impedes only the common-mode current, leaving the wanted signal completely unaffected. This is not a theoretical idealization — it is a real measurable property of the device, and it is why you can add a choke balun to a feedline and see no change in SWR or signal strength while the common-mode noise drops by 20 dB.
Common ferrite core types for ham radio use. The mix number determines the frequency range; the physical size determines how much cable can be wound through and how much power the core can handle.
View LargerFerrite Mix Numbers and Frequency Ranges
Ferrite materials are identified by "mix numbers" — standardized designations assigned by manufacturers Fair-Rite Products and Amidon Associates, the two main suppliers in the amateur radio community. The mix number tells you the ferrite formulation and, by implication, the frequency range where it is most effective as a common-mode choke. Using the wrong mix is a common mistake that leads to expensive failures.
The underlying reason for different frequency performance is the magnetic loss mechanism. At low frequencies, ferrite cores can magnetize and demagnetize with each RF cycle without dissipating significant energy. As frequency increases, the magnetization cannot keep up and the material starts to lag behind — this lag shows up as magnetic loss (the resistive component of impedance). Each ferrite formulation has a specific frequency region where this loss is maximized and where it therefore provides the most effective suppression. Below that range the material is too efficient (too reactive, not lossy enough); above that range the permeability collapses and the material provides little impedance at all.
| Mix Number | Main Components | Relative Permeability (µr) | Best Frequency Range | Primary Ham Radio Use |
|---|---|---|---|---|
| Mix 31 | Manganese-zinc (MnZn) | ~1500 | 1–300 MHz (peak loss 25–250 MHz) | HF common-mode chokes; broadest HF coverage; best all-around choice for 3–30 MHz feedline chokes |
| Mix 43 | Manganese-zinc (MnZn) | ~850 | 10 MHz–1 GHz (good 25–300 MHz) | All-around HF/VHF suppression; very common in snap-on clamps; good for 14–148 MHz range |
| Mix 61 | Nickel-zinc (NiZn) | ~125 | 200 MHz–2 GHz | VHF and UHF suppression; lower impedance than Mix 31 at HF but very effective at 144 MHz and above |
| Mix 77 | Manganese-zinc (MnZn) | ~2000 | 0.1–10 MHz (peak loss 0.5–5 MHz) | 160-meter and 80-meter feedline chokes; AM broadcast band interference; very low frequency suppression |
| Mix 52 | Nickel-zinc (NiZn) | ~250 | 10 MHz–500 MHz | Moderate HF suppression; sometimes used where Mix 43 is unavailable |
The practical rule for most HF operators is: use Mix 31 or Mix 43 for 3–30 MHz feedline chokes, with Mix 31 being the better choice when you want maximum suppression over the entire HF spectrum. Mix 43 is widely available in snap-on clamp form and is adequate for applications above 14 MHz. For 160 meters (1.8 MHz) or for suppressing AM broadcast interference entering through an HF feedline, Mix 77 provides significantly better low-frequency performance. For 2-meter and 70-centimeter work, Mix 61 is the right choice.
Mix 31 has become the dominant choice for HF common-mode chokes in current amateur practice, largely due to detailed published data from W1JB (Joe Reisert) and K9YC (Jim Brown). Their measurements showed that Mix 31 provides higher common-mode impedance over the 2–30 MHz range than any other readily available ferrite material. A well-wound FT-240-31 choke can provide 2,000 to 5,000 ohms of common-mode impedance across the HF bands, compared to 500 to 2,000 ohms from an equivalent FT-240-43 choke.
Core Shapes and Sizes
Beyond the mix number, you need to choose the right physical form for your application. Ferrite is available in several shapes, and each has specific advantages.
Toroid ring cores are the most efficient shape for wound chokes. The toroid has no air gap — the magnetic path is completely through ferrite material — which gives it the highest inductance per turn and the most effective use of the ferrite material. The round shape also minimizes stray radiation from the core. For a feedline choke intended to carry full transmit power, a toroid wound with coaxial cable is the correct choice. Toroids also have the advantage that the winding direction does not matter for a choke application.
Split snap-on clamp cores are convenient because they can be added to existing cables without cutting or soldering. They consist of two ferrite halves in a plastic hinged housing that clips around the cable. The trade-off is that the split joint introduces an air gap that reduces effective permeability and therefore reduces impedance compared to an equivalent toroid. Snap-on clamps are useful for adding suppression to computer cables, USB cables, and audio cables where you do not want to disassemble connectors. For feedline common-mode chokes carrying transmit power, a wound toroid is significantly more effective.
Binocular cores (also called two-hole cores or "balun cores") have two parallel holes side by side. Wire passes through the holes in a figure-8 pattern. They are often used for wideband RF transformers and some choke applications where the wire path through the core can be made shorter. For common-mode chokes, binocular cores are less common than toroids.
Rod and sleeve cores are cylindrical. They are used for suppression of interference on cables where the cable is threaded inside a ferrite tube or sleeve rather than wound around a toroid. The tube places the ferrite around the cable rather than through it, providing some common-mode suppression but less than an equivalent wound toroid.
Core size notation. The naming convention used by Amidon Associates is widely used in amateur radio: FT-240-43 means "Ferrite Toroid, 2.40 inch outer diameter, Mix 43." The number between the FT prefix and the mix number is the outer diameter in hundredths of an inch. Common sizes are:
| Amidon Designation | Outer Diameter | Inner Diameter | Typical Use |
|---|---|---|---|
| FT-240 | 2.40 in (61 mm) | 1.40 in (36 mm) | HF feedline chokes with RG-8X or larger coax; high-power applications |
| FT-140 | 1.40 in (36 mm) | 0.90 in (23 mm) | HF feedline chokes with RG-58 or RG-8X; medium power |
| FT-82 | 0.82 in (21 mm) | 0.52 in (13 mm) | Small chokes on thin cable; low power or receive-only |
| FT-50 | 0.50 in (13 mm) | 0.31 in (8 mm) | Audio cables, small accessory cables; receive only; do not use for transmit feedlines |
For a transmit feedline choke on an HF station running 100 watts or more, the FT-240 is the standard choice. It accommodates heavier coax and has the thermal mass to handle the small amount of heat generated by common-mode current losses at legal power levels. The FT-140 is suitable for lower power or smaller coax. Never use FT-50 or FT-82 cores for transmit feedline chokes — they are too small for the power and will overheat.
How Many Turns? Impedance vs Frequency
The impedance of a ferrite choke increases with the square of the number of turns (N²) because each turn contributes inductance proportional to N, and inductance combines with the core's impedance characteristics to give total choke impedance. However, adding too many turns introduces parasitic capacitance between adjacent turns, which can short-circuit the choke at high frequencies, creating a self-resonance above which the choke loses effectiveness.
The impedance target for a common-mode choke depends on the application. For a feedline choke at an antenna feedpoint, you want at least 500 ohms of common-mode impedance at your operating frequency, with 1,000 ohms or more being significantly better. The higher the impedance of the choke relative to the source of common-mode current, the more effectively it blocks that current. With 1,000 ohms of choke impedance and a typical antenna impedance of 50 ohms, the common-mode current is reduced by a factor of about 20 (26 dB). With 5,000 ohms, the reduction is about 100:1 (40 dB).
An FT-240-43 toroid with 9 turns of RG-8X coax: according to Fair-Rite and Amidon published data, a single turn on this core gives approximately 14 µH inductance. With 9 turns, inductance is 14 × 9² = 14 × 81 = 1,134 µH — but in practice the published impedance charts are more reliable than first-principles calculation because they account for core losses directly.
Published data for this combination shows approximately:
- 1.8 MHz (160m): ~200 ohms — marginal for 160m use
- 3.5 MHz (80m): ~600 ohms — adequate
- 7 MHz (40m): ~1,200 ohms — good
- 14 MHz (20m): ~2,000 ohms — excellent
- 28 MHz (10m): ~1,500 ohms — good
For a station that operates primarily on 40m through 10m, this choke is excellent. For 160m and 80m primary operations, add Mix 77 cores or increase to 12 turns.
Practical rules of thumb for common-mode choke winding:
- FT-240-31, 9–12 turns of RG-8X: 1,000–5,000 ohms from 2–30 MHz — best all-HF choice
- FT-240-43, 8–10 turns of RG-8X: 600–2,000 ohms from 7–30 MHz — good for 40m and above
- FT-240-77, 10–14 turns: best for 160m and 80m; less effective above 10 MHz
- Stacking two FT-240-31 cores: approximately doubles the impedance — use when maximum suppression is needed
Stacking cores — placing two or three toroids in a row and winding through all of them as a group — multiplies the impedance approximately proportionally to the number of cores. Two FT-240-31 cores wound together with 9 turns of RG-8X will give approximately twice the impedance of a single core. This technique is particularly useful when a very high choke impedance is needed, such as for an off-center fed dipole or end-fed antenna where the common-mode drive is inherently strong.
Winding Techniques for Coax Feedlines
Winding a coax choke through a toroid is straightforward once you understand the conventions. The entire coaxial cable — jacket, braid, dielectric, and center conductor — passes through the center hole of the toroid. Each pass through the hole counts as one turn. The coax is not stripped or modified in any way; the ferrite acts on the outside of the assembled coax cable.
Winding a 9-turn coax choke on an FT-240-43 toroid. Count each pass of the cable through the center hole as one turn. The winding direction does not matter; spacing turns evenly reduces parasitic capacitance.
View LargerStep-by-step procedure for winding a 9-turn FT-240-43 choke on RG-8X coax:
- Cut a length of RG-8X about 2.5 meters (8 feet) — this is enough for 9 turns on an FT-240 with lead lengths remaining to reach connectors.
- Find the center of the coax and mark it lightly with tape. Start winding from the center so the two free ends are approximately equal in length when you finish.
- Pass one end of the coax up through the center hole of the toroid. This is turn 1.
- Bring the coax around the outside of the toroid and back up through the center hole from the same direction. This completes turn 2. Continue in the same direction.
- Count turns as you go: each pass through the center hole is one turn. Wind 9 turns total, spacing the turns evenly around the toroid. You do not need to cover the full circumference — 9 turns on a FT-240 will occupy about two-thirds of the core.
- Both ends of the coax should exit on the same side of the toroid when the winding is complete. This is not a rule — either side is fine — but having them exit on the same face makes connector installation easier.
- Attach PL-259 connectors to both ends using standard coax prep and soldering technique.
A few important points about winding technique:
- Winding direction does not matter for a choke. Unlike a transformer, a choke works identically whether wound clockwise or counterclockwise. Do not worry about it.
- Spacing matters for parasitic capacitance. Turns that are packed tightly together have more mutual capacitance between them, which creates a self-resonance at a lower frequency than necessary. Space the turns as evenly as possible around the core for best high-frequency performance.
- For larger coax (RG-213, LMR-400): Use the FT-240 size. Heavy coax may only fit 7–8 turns through an FT-240 before the hole is full. This is still adequate for most HF applications. For large-diameter hardline, it may be necessary to use a choke coil (coax wound in a coil of 6–8 turns, 6 inches diameter) rather than a ferrite toroid.
- For RG-58 or thin coax: The FT-140 core is fine, and you can wind more turns. An FT-140-31 with 12 turns provides excellent HF suppression and is more economical than the larger FT-240 for receive-only or QRP applications.
Snap-On Clamps and Their Use
Snap-on ferrite clamps are the quick-fix tools of the RFI world. They require no soldering, no disassembly of cables, and can be added in minutes. They are not as effective per unit as a properly wound toroid, but they are very useful for computer cables, USB cables, audio leads, and power supply cables where you want to add some suppression without modifying the cable.
A single snap-on clamp on a USB cable contributes approximately one to two effective turns of choke inductance, depending on the cable diameter and fit. The air gap at the split joint reduces effective permeability by about 30 to 50 percent compared to an equivalent wound toroid. For moderate suppression this is acceptable; for heavy common-mode suppression it is not enough.
You can increase effectiveness by making multiple loops of the cable through the same clamp. Instead of routing the cable straight through once, loop it around and pass it through the clamp a second or third time before the clamp closes. Two loops give approximately four times the impedance of one pass (due to the N² relationship). Some snap-on clamps have a larger bore specifically to accommodate multiple loops.
The best approach for computer noise in the shack is to use multiple clamps at strategic points: one at the computer end of each cable (stop the noise from getting on the cable) and one at the radio end (stop any noise that did get on the cable from entering the radio). Use Mix 31 or Mix 43 clamps for HF noise suppression.
When buying snap-on clamps, look for products marked Mix 31 or Mix 43 from Fair-Rite Products, Würth Elektronik, or TDK. Generic unbranded clamps from bargain suppliers may use Mix 43 or lower-grade ferrite — they may still provide some suppression but their performance is less predictable.
Power Handling
For receive-only applications and low-power QRP use, ferrite core size is not a concern. For transmit feedline chokes at 100 watts or above, you need to verify that your ferrite choke can handle the power without excessive heating or saturation.
The heating in a feedline choke comes from the small percentage of transmit power that is absorbed by the lossy ferrite as common-mode current flows through it. In a well-matched antenna system with low common-mode current, the power dissipation in the choke may be a fraction of a watt even at 1,500 watts transmit power. If the antenna system has a significant imbalance, the common-mode current could be tens of milliamps, and the ferrite will dissipate correspondingly more power.
As a practical guideline: an FT-240 core with 9–11 turns of RG-8X handles 1,500 watts PEP on HF with a good safety margin, provided the antenna system is reasonably balanced. During key-down operation at high power (CW or digital modes), monitor the temperature of the choke. It should feel warm to the touch — perhaps 40 to 60°C — but not hot enough to be uncomfortable to hold. If it gets too hot to touch, either the common-mode current is higher than expected (suggesting an imbalance problem to investigate) or the core is too small.
Never use FT-50 or FT-82 cores on transmit feedlines. Their small cross-sectional area limits both power handling and flux density before saturation. Even at low power levels, small cores near saturation produce non-linear behavior and can generate harmonics on the transmit signal.
Measuring Ferrite Impedance
The most reliable way to confirm that your choke provides adequate impedance at your operating frequency is to measure it directly. With a NanoVNA or other antenna analyzer, you can sweep the choke impedance from 1 to 30 MHz and verify that it meets the 500-ohm minimum (or 1,000-ohm preferred) target at all operating frequencies.
Measurement procedure: connect one end of the choke (with PL-259 connectors already installed) to Port 1 of the NanoVNA. Short the center conductor to the braid at the other end of the choke — this shorts the differential mode path and forces the analyzer to measure only the common-mode impedance through the ferrite. Set the NanoVNA to measure |Z| (impedance magnitude) and sweep from 1 to 30 MHz. The resulting curve shows common-mode impedance at every frequency.
Published ferrite data from manufacturers Fair-Rite and Amidon are available in their catalogs and on their websites. W1JB's ferrite comparison articles and K9YC's "A Ham's Guide to RFI, Ferrites, Baluns, and Audio Interfacing" document contain measured impedance data for the most common core types and turn counts. These are the definitive references for amateur ferrite choke design.
⚖ Experiment: Winding and Testing a Choke Balun
Wind a 9-turn coax choke on an FT-240-43 toroid and measure its common-mode impedance with a NanoVNA or antenna analyzer. This experiment lets you verify the core performs as published data predicts and understand the effect of turn count on impedance.
- FT-240-43 toroid core (Fair-Rite or Amidon)
- 2.5 m (8 ft) of RG-8X 50-ohm coaxial cable
- Two PL-259 connectors and a suitable crimp or solder tool
- NanoVNA or antenna analyzer capable of measuring impedance 1–30 MHz
- A short piece of wire or solder bridge to short center conductor to braid at one end
- Wind the RG-8X through the FT-240-43 toroid: 9 turns, each pass through the center hole counting as one turn. Space turns evenly around the core. The total coax used will be approximately 1.8 m for the winding, leaving lead lengths for the connectors.
- Install a PL-259 connector on each end of the coax.
- Connect one end to Port 1 of the NanoVNA. At the other end, short the center pin to the outer shield of the PL-259 connector with a jumper wire. This creates the shorted termination needed to measure common-mode impedance.
- Calibrate the NanoVNA for the 1–30 MHz range (open, short, load calibration at Port 1).
- Set the NanoVNA to display |Z| (impedance magnitude). Sweep 1–30 MHz and record the impedance at 1.8 MHz (160m), 3.5 MHz (80m), 7 MHz (40m), 14 MHz (20m), and 28 MHz (10m).
- Now unwind two turns to make a 7-turn choke. Repeat the measurement. Observe how the impedance changes — lower at all frequencies, and the peak moves to a higher frequency.
- Add back 4 turns to make an 11-turn choke. Measure again. Observe the higher impedance but note whether a self-resonance dip appears at a lower frequency due to increased parasitic capacitance.
With 9 turns on FT-240-43, you should see impedance rising from approximately 200 ohms at 1.8 MHz to a peak of 1,500–2,000 ohms somewhere in the 10–20 MHz range, then declining slightly at 28 MHz. The 7-turn version will show lower impedance throughout. The 11-turn version will show higher impedance at low frequencies but may show a self-resonance dip somewhere in the 20–28 MHz range where parasitic capacitance resonates with the choke inductance. This experiment directly confirms the published data and teaches you how turn count affects choke performance.
Frequently Asked Questions
Can I use any ferrite from a junk box?
Maybe — but the mix number matters greatly, and identifying unmarked cores is difficult. Ferrite from computer power supplies, old CRTs, and surplus electronics is often Mix 43 or Mix 77, but without markings you cannot be certain. Gray or charcoal-colored cores from computer equipment are commonly Mix 43; darker, heavier cores may be Mix 77. However, a core that looks similar to an FT-240-43 might be a completely different formulation that provides negligible impedance at your operating frequency. For reliable results, buy new cores from Fair-Rite or Amidon with known mix numbers. The cost of an FT-240-31 is modest compared to the time spent troubleshooting a choke made from the wrong material.
How do I know if my choke balun has enough impedance?
The target is at least 500 ohms of common-mode impedance at your operating frequency, with 1,000 ohms or more strongly preferred. To measure: connect one end of the completed choke to your antenna analyzer or NanoVNA. Short the center conductor to the braid at the other end. Sweep impedance from 1 to 30 MHz and check the value at your operating frequencies. If you see less than 500 ohms at your main operating frequency, try adding more turns or switching to Mix 31 which provides higher impedance over the HF range. Published data from W1JB and K9YC gives reliable impedance values for standard core and turn combinations so you can design to specification before building.
Will a choke balun affect the SWR on my feedline?
No — a properly constructed choke balun has no effect on the differential-mode (transmission line) impedance inside the coax, and therefore no effect on SWR, signal levels, or power transfer to the antenna. The ferrite only impedes current that flows on the outside surface of the braid as a common-mode current. Inside the coax, the transmission line mode is completely unaffected. If you see a change in SWR after adding a choke, it usually means the choke was actually changing the antenna's resonant frequency because the feedline was previously acting as part of the antenna — which confirms the feedline was radiating before the choke was installed.
Test Your Knowledge
Answer the questions below to check your understanding. Every answer can be found in the lesson above.