Toroid and Ferrite Core Selection Guide
Every balun, unun, and common-mode choke in this series depends on choosing the right core — get the mix or size wrong and you get a choke that runs hot, saturates at power, or does nothing at all. This guide compares the ferrite mixes and iron powder cores actually used in ham radio construction, explains the AL value math you need for turn counts, and covers how to size a core so it survives your power level.
What a core is doing electrically
A toroid core raises the inductance of a winding far beyond what the same wire would produce in air, by concentrating magnetic flux inside a high-permeability ceramic (ferrite) or iron powder ring. In a common-mode choke, that inductance is placed in series with the outside of the coax shield, presenting high impedance to common-mode current while leaving the differential (wanted) signal inside the coax untouched. In a transformer-style balun or unun, the core couples two or more windings so tightly that the turns ratio sets the impedance transformation ratio.
The core mix determines two things that matter more than anything else for a ham application: how much impedance you get per turn at your operating frequency, and how much heat the core generates when it is doing real work (absorbing common-mode current or handling reactive power). Pick a mix suited to 3–30 MHz and use it at 440 MHz and it will do almost nothing. Pick a mix suited for VHF and use it at 1.8 MHz and you will need an impractical number of turns to get useful impedance.
Ferrite vs. powdered iron
Ferrite cores (the "#" mix numbers: #31, #43, #52, #61, #77, #75) are ceramic, have high initial permeability (mu-i from about 40 to over 5000), and are the correct choice for common-mode chokes and broadband transformers on HF and VHF. Powdered iron cores (the "T-" and "-2/-6/-10/-12" mix numbers, and the yellow/red/black color-coded toroids) have much lower permeability (mu-i around 1–35), store energy without saturating as easily, and are the correct choice for resonant or narrowband tuned circuits — such as antenna tuner inductors and low-pass filters — not for wideband chokes.
Reading the mix numbers
Manufacturers (Fair-Rite, Amidon, Micrometals) assign a mix number to each ferrite formulation that sets its permeability and useful frequency range. The physical size is a separate code (e.g. FT-240 = 2.4-inch OD ferrite toroid, FT-140 = 1.4-inch OD). A full part number like FT-240-43 tells you both: 2.4-inch OD, mix #43.
- #31: Highest permeability of the common mixes (mu-i ≈ 1500). Best low-frequency choke performance from 1.8–30 MHz; loses effectiveness above about 30 MHz.
- #43: The general-purpose ham radio mix (mu-i ≈ 850). Good broadband choke and balun performance from about 2–200 MHz — the default choice if you are only buying one mix.
- #61: Lower permeability (mu-i ≈ 125), lower loss at VHF/UHF. Better choice above 50 MHz where #43 core loss starts to matter more.
- #77: Similar range to #31 (mu-i ≈ 2000) with different loss characteristics; sometimes substituted for #31 in choke designs.
Sizing for power: saturation and heating
A core doesn't fail by "not working" under power — it fails by overheating or saturating, which shows up as SWR that changes with transmit power, a core that cracks, or wire insulation that melts at the surface. Flux density in the core rises with RF voltage across the winding and falls with core cross-sectional area, frequency, and turns squared. Undersized cores on high-SWR or high-power feed lines (a common-mode choke seeing a poor match, or a 4:1/9:1 unun on a mismatched antenna) are the most frequent failure case in this series.
| Mix | Type | Initial Permeability | Best Frequency Range | Typical Ham Use |
|---|---|---|---|---|
| #31 | Ferrite | ~1500 | 1.8–30 MHz | 1:1 and low-ratio HF current baluns, shack-entry chokes |
| #43 | Ferrite | ~850 | 2–200 MHz | General-purpose HF/VHF chokes, EFHW ununs, multiband baluns |
| #52 | Ferrite | ~250 | 25–200 MHz | VHF transformers, less common in ham baluns |
| #61 | Ferrite | ~125 | 20–300+ MHz | VHF/UHF chokes where #43 core loss is too high |
| #77 | Ferrite | ~2000 | 1–20 MHz | Alternative to #31 for low-band chokes and baluns |
| Iron powder -2 (red) | Powdered iron | ~10 | 1–30 MHz | Tuner inductors, low-pass filter coils, NOT chokes |
| Iron powder -6 (yellow) | Powdered iron | ~8 | 3–40 MHz | Tuner inductors, higher-Q than -2 mix at HF |
Turns-for-Inductance Calculator
Materials for stocking and identifying cores correctly
Assortment of ferrite and powdered iron toroid cores for comparison
Selecting the Right Core for Your Build
Work through this sequence before winding anything — most "my choke doesn't work" problems trace back to skipping steps 1 and 3.
Identify the application, not just the frequency
Decide whether you need a wideband common-mode choke (ferrite), a transformer winding for a balun/unun (ferrite), or a tuned inductor for a tuner/filter (powdered iron). This decision comes before picking a mix number — using the wrong core type is the single most common mistake, and no amount of turn-count math fixes it.
Pick the mix for your frequency range
Use the mix comparison table above. For a general HF choke or unun covering 80–10m, #43 is the safe default. For a low-band-only (160/80/40m) build where maximum impedance per turn matters more than VHF coverage, #31 or #77 wins. For anything you expect to also see 6m or 2m energy on (e.g. a shack-entry choke that should also block VHF common-mode current), #43 still performs reasonably; pure VHF work favors #61.
Size the core to your power level and expected SWR
Match core size (OD) to power: FT-82 size (0.8-inch OD) is adequate up to roughly 100W with a reasonable match; FT-140 (1.4-inch) handles up to about 500W–1kW; FT-240 (2.4-inch) is the standard choice for full legal-limit HF stations or any application (like an EFHW unun) that regularly sees elevated SWR before the antenna is fully tuned. Undersizing here is what leads to cracked or overheated cores under real operating conditions, not just at the rated CW power into a perfect 50 ohm load.
Calculate turns from the AL value
Once mix and size are chosen, look up the AL value (inductance per turn squared, in nH/turn²) from the manufacturer datasheet for that exact part number — AL varies by mix and core size, not just mix alone. Use it to solve for the turns needed to hit your target inductance. Use the calculator above for this, or work it by hand:
Verify with a NanoVNA before relying on the core
Wind the calculated turns, then measure the actual choke impedance (or transformer inductance) with a NanoVNA rather than trusting the datasheet AL value alone — real-world AL varies core-to-core by 10–20%, and a cracked or mislabeled core will measure far outside the expected range. For a common-mode choke, look for at least 500 ohms of common-mode impedance across your operating bands; for a transformer winding, confirm the inductance is high enough that reactance stays well above the load impedance at your lowest operating frequency.
| Symptom | Most likely cause | Diagnosis | Fix |
|---|---|---|---|
| Core cracks or feels hot within seconds of transmitting | Core undersized for power/SWR, or wrong mix chosen for the load it's seeing | Check SWR at the choke/transformer location, not just at the transmitter; a matched-looking SWR at the shack can still mean high internal voltage if the line has standing waves | Step up to a larger core size (e.g. FT-140 to FT-240) or add a second core in series/stacked to share the flux |
| Choke has little or no effect on common-mode current (RF in the shack persists) | Powdered iron core used where ferrite was needed, or mix chosen for the wrong frequency range | Measure common-mode impedance with a NanoVNA; powdered iron mixes will show only tens of ohms where a ferrite choke should show 500+ | Rewind on the correct ferrite mix (#31/#43/#61 per the comparison table) for the actual operating frequency |
| Choke works fine at low power but SWR shifts as power increases | Core approaching saturation at higher RF voltage | Reduce power and see if the shift disappears; recalculate flux density at full power using the formula above | Increase core size, add turns (raises L, reduces flux density for a given voltage), or split the winding across two stacked cores |
| Measured inductance is far off from the AL-value calculation | Wrong or counterfeit core, or turns counted incorrectly (through-hole passes vs. physical turns) | Recount turns using the standard convention (each full pass through the center counts as one turn), and re-verify core dimensions against the datasheet with calipers | Re-wind with correct turn count, or replace the core if dimensions don't match any known catalog part |
| Wire insulation is discolored or melted right where it exits the core | Localized overheating from core loss at that operating frequency, usually a mix mismatch (e.g. #43 pushed hard well above 200 MHz, or #31 used at 6m/2m) | Confirm operating frequency against the mix's rated range in the comparison table | Switch to a mix rated for the actual frequency in use (#61 for higher VHF/UHF, #43 for general HF/low-VHF) |
Can I substitute a #43 core for a #31 core?
Yes, for general multiband HF use — #43 is the more versatile mix and is the standard recommendation when only stocking one mix. You will get somewhat lower common-mode impedance on 160/80m with #43 than a purpose-built #31 choke of the same turn count, so on strictly low-band chokes, #31 or #77 remains the better choice if you already know the application is low-band-only.
What AL value do I actually need?
There's no single target — it depends on the application. For an HF common-mode choke, aim for enough turns on a #43 or #31 core to reach roughly 10-15 uH, which typically yields 500+ ohms of common-mode impedance across most of the HF range. For a balun/unun transformer winding, the required inductance is set by keeping the winding's reactance several times higher than the impedance it's transforming at your lowest operating frequency — check the specific balun or unun guide for the exact target.
Are iron powder cores ever acceptable for a balun or choke?
Not for a wideband common-mode choke or a 1:1/4:1/9:1 current balun — their low permeability means you would need an impractical number of turns to get useful impedance, and even then, core loss characteristics aren't optimized for that job. Powdered iron belongs in tuned inductors (antenna tuner coils, low-pass filters) where you want energy storage without saturation, not broadband impedance.
When should I stack two cores instead of using one larger core?
Stacking two identical cores (placing one directly on top of the other before winding) roughly doubles the effective cross-sectional area, which lowers flux density for the same power and turns — useful when you're already using the largest commonly available size (FT-240) but still seeing thermal issues at very high power or poor SWR. It's a practical alternative to hard-to-find oversized cores, at the cost of a bulkier finished choke or transformer.
Does the color of a powdered iron core tell me the mix?
For Amidon/Micrometals-style toroids, yes — it's a widely used convention: red is mix -2, yellow is mix -6, black is mix -10, green/white is mix -12, blue is mix -17, and yellow/white is mix -26. Always confirm against the seller's listing or a datasheet if the core came without documentation, since color coding conventions can vary between manufacturers and counterfeit or relabeled cores do turn up in the secondary market.
How do I identify an unmarked or unknown core?
Measure OD, ID, and height with calipers and compare against catalog dimensions to narrow down the physical size code. Then wind a known number of test turns (10 is convenient), measure inductance with a NanoVNA or LCR meter, and back-calculate AL = (L in nH) / N². Compare that AL value and the core's color/appearance against the datasheet table for your best-guess size to identify the likely mix.