Cable Shield Quality: Transfer Impedance & Effective...

Why Cable Shielding Matters More Than You Think

You've routed your sensitive analog signal through a shielded cable, connected the shield at both ends, and yet your EMC pre-scan still shows a nasty spike at 150 MHz. Sound familiar? The problem often isn't whether you have a shield — it's how effective that shield actually is at the frequencies that matter.

Here's the thing most engineers miss: cable shielding effectiveness isn't some magic number you look up once in a datasheet and forget about. It's frequency-dependent, construction-dependent, and length-dependent. A cable that gives you 80 dB of shielding at 10 MHz might only manage 45 dB at 200 MHz, and that difference is often the gap between passing and failing your radiated emissions test.

The shield construction matters enormously. Braid? Foil? Spiral wrap? Each behaves differently as frequency climbs. Then there's the DC resistance of the shield material itself, which dominates at low frequencies but becomes less important as skin effect kicks in. And of course, cable length — longer runs mean more opportunity for coupling, which degrades your shielding effectiveness proportionally.

Understanding how these parameters interact isn't just academic curiosity. It's the difference between confidently shipping a product and scrambling to explain to management why you need another three weeks and a complete cable redesign. The Cable Shield Effectiveness calculator lets you model this quickly without building spreadsheets or digging through IEEE papers every time you need an estimate.

Transfer Impedance: The Key Metric

Transfer impedance, $Z_T$, is what separates hand-waving about "good shielding" from actual quantitative analysis. It tells you exactly how much voltage appears on the inner conductor per unit length when current flows on the outside of the shield. The formal definition is:

where $V_{inner}$ is the induced voltage on the inner conductor, $I_{shield}$ is the current flowing on the shield, and $L$ is the cable length.

Think of it this way: if you have 1 amp of shield current over 1 meter of cable and that induces 50 millivolts on your inner conductor, your transfer impedance is 50 mΩ/m. Lower numbers are better — they mean less coupling from the shield to your signal.

At low frequencies, below a few MHz, transfer impedance is straightforward. It's basically just the DC resistance per unit length of the shield material. Copper braid? Aluminum foil? Whatever the resistivity is, that's what dominates. Simple.

But as frequency increases, things get interesting. Two competing physical effects start fighting each other:

Skin effect pushes current toward the outer surface of the shield. This is actually helpful — the current concentrates on the outside, meaning less magnetic field penetrates to the inner conductor. Transfer impedance drops as frequency rises, sometimes dramatically. Braid leakage and porpoising work against you. In braided shields, the weave pattern creates small gaps between the wires. At low frequencies these don't matter much because the magnetic field just flows around them. But at higher frequencies, these apertures start acting like little antennas, letting field through. The braid effectively becomes more transparent. This effect increases transfer impedance with frequency.

For a solid tubular shield — think coax with a continuous copper tube — transfer impedance decreases monotonically with frequency due to skin effect:

where $t$ is the shield wall thickness and $\delta$ is the skin depth at frequency $f$:

The skin depth tells you how deeply the current penetrates into the conductor. At 100 MHz in copper, it's only about 6.6 micrometers. At 1 GHz it's barely 2 micrometers. The current is essentially riding on the surface.

For braided shields, which is what most of us use in practice, the behavior is more complex. Transfer impedance typically drops with frequency initially, reaches a minimum somewhere between 1 MHz and 30 MHz depending on the braid geometry, then starts climbing again as porpoising takes over. This is why a cable that performs beautifully at 10 MHz can be surprisingly leaky at 200 MHz. The physics changes.

Shielding Effectiveness From Transfer Impedance

Once you know $Z_T$, you can calculate shielding effectiveness (SE) in decibels. This is where we compare the transfer impedance to a reference impedance — usually the circuit impedance or the 50 Ω test system impedance. A common simplified expression is:

where $Z_0$ is your reference impedance (typically 50 Ω) and $L$ is the cable length in meters.

Higher SE values mean better shielding. As a rough guide: 60 dB is decent for most commercial applications, 80 dB is good and will handle most industrial environments, and 100+ dB is excellent — military or medical grade. But these are just guidelines. Your actual requirements depend on your circuit impedances, signal levels, and what interferers you're dealing with.

Notice that SE degrades linearly with cable length in dB terms. Double the cable length and you lose 6 dB of shielding effectiveness. This is why keeping cable runs short is such a universal rule in EMC design — it's not superstition, it's physics.

Worked Example: Evaluating a 2-Meter Braided Shield Cable at 100 MHz

Let's work through a realistic scenario. You're using a 2-meter cable with a tinned copper braid shield. The manufacturer's datasheet (if you're lucky enough to have one that's actually complete) specifies a shield DC resistance of 15 mΩ/m. You need to know if this will give you adequate shielding at 100 MHz, where you've got some switch-mode power supply harmonics causing trouble.

Inputs:

• Shield DC resistance: $R_{DC} = 15 \text{ mΩ/m}$
• Cable length: $L = 2 \text{ m}$
• Frequency: $f = 100 \text{ MHz}$

First, let's calculate the skin depth for copper at 100 MHz. Copper's resistivity is $\rho = 1.68 \times 10^{-8}$ Ω·m:

Just 6.6 micrometers. That's thinner than a human hair. For a typical braid with an effective thickness around 0.1 mm (100 μm), the ratio $t/\delta \approx 15$. This means skin effect is very significant — the current is strongly concentrated on the outer surface.

Now here's where it gets tricky. If this were a solid tube, we could calculate transfer impedance directly from the skin depth. But it's a braid, not a solid shield. The weave pattern adds a mutual inductance term that increases transfer impedance at high frequencies. The porpoising effect I mentioned earlier.

Typical braided cables at 100 MHz, assuming decent quality with around 85% optical coverage, exhibit transfer impedances in the range of 10–100 mΩ/m. The exact value depends on the braid angle, the number of carriers, and how tightly it's woven. Let's assume the calculator determines $Z_T \approx 50 \text{ mΩ/m}$ at 100 MHz — this is realistic for an 85% coverage braid, maybe slightly optimistic depending on construction.

The total transfer impedance over the 2-meter cable length is:

Now we can calculate shielding effectiveness referenced to 50 Ω:

That's... not great. It's marginal for many EMC requirements. If your specification calls for 60 dB of shielding — which is pretty common for commercial products — you're 6 dB short. That might not sound like much, but remember, decibels are logarithmic. You're off by a factor of 2 in voltage terms.

What are your options? You could shorten the cable run to 1 meter, which would gain you 6 dB and get you to 60 dB. Or you could switch to a higher-coverage braid — 95% optical coverage instead of 85%. That can reduce transfer impedance by a factor of 3 to 5 at high frequencies, potentially getting you down to 15 mΩ/m or better. That would give you about 70 dB of shielding effectiveness, which provides some margin.

The best option? Move to a braid-plus-foil construction. A good braid-over-foil cable can achieve transfer impedances below 5 mΩ/m at 100 MHz. With the same 2-meter length, that would give you:

Now you've got 14 dB of margin over your 60 dB requirement. Much better.

You can verify all of this by plugging these exact values into the Cable Shield Effectiveness calculator. It'll show you the results instantly, and you can sweep frequency to see exactly where your shielding holds up and where it starts degrading. This kind of quick analysis can save you from discovering problems during formal compliance testing, when fixes are expensive and schedule-impacting.

Practical Tips for Improving Shield Effectiveness

Increase braid coverage. This is often the easiest win. Going from 85% to 95% optical coverage can reduce transfer impedance by a factor of 3 to 5 at high frequencies. The difference between 85% and 95% sounds small, but in terms of the apertures in the weave, you're reducing the leakage area significantly. Yes, 95% coverage cable costs more. It's still cheaper than failing EMC testing. Use combination shields. A braid-over-foil construction gives you the best of both worlds — the low-frequency performance and mechanical durability of the braid, plus the high-frequency sealing of the foil. The foil provides a continuous conductive barrier with no apertures, while the braid handles the mechanical stress and provides a low-impedance termination point. For really demanding applications, double-shielded cables with foil-braid-foil-braid construction are available, though they're stiff and expensive. Minimize cable length. This is the most obvious recommendation, but it's worth repeating because people ignore it constantly. Since shielding effectiveness degrades linearly with length in dB terms, shorter cables always win. If you can cut that 2-meter run down to 1 meter, you've just gained 6 dB. Sometimes the best EMC fix is just better mechanical packaging that allows shorter interconnects. Terminate the shield properly. Most engineers know this in theory but mess it up in practice. A pigtail ground connection — where you strip back the shield, twist it into a wire, and connect it to a ground pin — can add 10 to 20 mΩ of impedance at the connector. At high frequencies, that might be more impedance than the entire cable shield. Use 360-degree backshell terminations wherever possible. The shield should connect to the connector body with a continuous circumferential contact, not a single-point pigtail. Yes, good backshells are expensive. So is re-spinning your board after failing radiated emissions. Watch out for resonances. This one catches people by surprise. Cable lengths that are odd multiples of λ/4 at your problem frequency can create standing waves on the shield. At these resonant lengths, the shield current distribution changes and shielding effectiveness can drop dramatically — sometimes by 20 or 30 dB at specific frequencies. If you see sharp dips in your EMC scan at particular frequencies, check if your cable length corresponds to a quarter-wave resonance. The fix is usually just changing the cable length by 10% or so to shift the resonance away from your problem frequency.

When to Worry (and When Not To)

For low-frequency applications — audio, slow serial buses below 1 MHz, DC power distribution — even a modest braid with 15 mΩ/m DC resistance provides excellent shielding. At these frequencies, transfer impedance is essentially just the DC resistance, and the total transfer impedance over any reasonable cable length is tiny compared to typical circuit impedances. You'd have to work hard to have shielding problems below 1 MHz with any decent cable.

The real challenges emerge above 30 MHz. This is where braid leakage starts dominating and transfer impedance can rise rapidly with frequency. If you're dealing with high-speed digital signals (USB 3.0, HDMI, Gigabit Ethernet), switch-mode power supply harmonics (which can extend well into the hundreds of MHz), or any application where radiated emissions in the 100 MHz to 1 GHz range matter, you need to take shield quality very seriously.

I've seen designs where the engineer spec'd a perfectly adequate cable for the fundamental signal frequency, not realizing that the harmonics or clock frequencies were much higher and would leak right through the shield. Then they're stuck explaining to the test house why they need to come back in three weeks with a different cable. Don't be that engineer.

Try It

Grab your cable's DC resistance spec and the length of your run, then open the Cable Shield Effectiveness calculator. Sweep across your frequencies of concern and see exactly where your shielding holds up and where it doesn't. It takes maybe 30 seconds to get a realistic estimate.

Is it perfect? No — there are second-order effects the simplified model doesn't capture, like the exact braid geometry or the impact of connector transitions. But it's accurate enough to tell you if you're in the right ballpark or if you need to reconsider your cable choice before you commit to a design. That's often all you need. Better to find out now with a quick calculation than during formal compliance testing when the clock is ticking and every day of delay costs money.