Three Ethernet cable constructions compared: U/UTP (unshielded), F/UTP (overall foil), and S/FTP (braid + individual pair foil) — the shielding layers directly determine EMI immunity and alien crosstalk performance

The Construction Difference: What's Inside UTP vs STP

UTP (Unshielded Twisted Pair)

UTP cable contains four pairs of copper conductors, each pair twisted at a different rate to minimize electromagnetic coupling between pairs. The conductors are insulated with polyethylene or FEP (fluorinated ethylene propylene), and the entire assembly is wrapped in an outer jacket — typically PVC, LSZH, or CMR-rated compound. There is no metallic shielding layer anywhere in the construction.

The physics of UTP noise rejection depends entirely on differential signaling: each pair carries equal and opposite signals. When external electromagnetic fields strike both conductors of a twisted pair, they induce nearly identical common-mode voltages. The Ethernet PHY receiver uses a differential amplifier with a common-mode rejection ratio (CMRR) of 40–60 dB to subtract out this common-mode noise, leaving only the differential signal. This works well in electrically quiet environments — but the CMRR ceiling is what ultimately limits UTP performance in high-EMI settings.

STP (Shielded Twisted Pair) — A Family, Not a Single Design

"STP" is the most misused term in structured cabling. It is not a single construction but a family of shielded cable designs defined by which layers carry shielding and what type of shielding they use. At minimum, an STP-class cable adds one or more conductive layers — aluminum/polyester foil, tinned copper braid, or both — positioned either around individual pairs, around the entire four-pair bundle, or at both levels simultaneously.

The physical result is a cable that is approximately 15–30% heavier, 1–3 mm larger in outside diameter, noticeably stiffer (minimum bend radius increases from 4× to 8× cable OD), and absolutely dependent on a continuous ground path from end to end for the shield to function as anything other than an accidental antenna.

ISO/IEC 11801 Shielding Nomenclature Decoded

The XX/YY TP Format

The international standard ISO/IEC 11801 defines a precise two-part naming scheme that eliminates the ambiguity of marketing labels like "STP" or "FTP":

Format: Overall_Shield / Pair_Shield TP

• Before the slash — describes the shield wrapping all four pairs together: U = none, F = foil (aluminum/polyester, 100% optical coverage), S = braid (tinned copper, typically 60–85% optical coverage), SF = braid layered over foil (maximum coverage)
• After the slash — describes the shield on each individual pair: U = unshielded pairs, F = foil per pair, S = braid per pair (rare in practice)

Complete Shielding Type Reference Table

For a complete breakdown of every ISO/IEC 11801 shielding code — how to decode any cable label containing U, F, S, or SF combinations, the precise differences between F/UTP and U/FTP, and why SF/UTP and S/FTP are not the same product — see our dedicated STP vs UTP Cable: Complete Terminology Guide.

How Shielding Works: Faraday Cage Physics

The Three Mechanisms of Shield Performance

When an electromagnetic wave encounters a conductive shield, three physical processes occur simultaneously:

1. Reflection: A portion of the wave energy bounces off the shield surface. This is the dominant mechanism at lower frequencies and accounts for most of the shielding effectiveness below 100 MHz. Reflection efficiency depends on the impedance mismatch between air (377 ohms) and the shield material.
2. Absorption: Energy that penetrates the shield surface is converted to heat as it travels through the conductive material. Absorption becomes the dominant mechanism above 100 MHz and increases with material thickness and conductivity. A 0.025 mm aluminum foil layer provides roughly 40 dB of absorption loss at 500 MHz.
3. Multiple reflections: At shield interfaces (foil-to-braid, braid-to-air), residual energy bounces between boundaries, creating additional attenuation. This effect is significant in SF/UTP and SF/FTP constructions where the braid-foil gap creates a secondary reflective cavity.

Foil vs. Braid: When Each Matters

This is why premium Cat6A and Cat8 cables use S/FTP construction: the braid handles low-frequency magnetic coupling and provides a low-impedance ground path, while the individual pair foils block high-frequency pair-to-pair capacitive crosstalk. Each layer does what it does best.

Grounding: The Make-or-Break Factor

The Ground Path Must Be Continuous

For a shielded cabling system to function, the shield must form an unbroken electrical path from the equipment ground at one end to the equipment ground at the other. Every component in the channel must participate:

• Shielded cable — the foil/braid runs the full length with an integral drain wire
• Shielded RJ45 connectors — the metal shell captures the drain wire and makes 360-degree contact with the cable shield
• Shielded keystone jacks — metal housing with internal grounding contacts that bond to the patch panel
• Grounded patch panel — metal panel bonded to the telecommunications grounding busbar (TGB) per ANSI/TIA-607-D
• Grounded equipment — the switch or server chassis provides the far-end ground reference

Break any link in this chain and the shield becomes a floating antenna — actively degrading performance rather than protecting it.

Single-End vs. Both-Ends Grounding

There are two valid grounding topologies, and choosing the wrong one for your environment creates ground loops:

• Both-ends grounding (standard): The shield is bonded to ground at both the patch panel and the equipment end. This provides the lowest shield impedance and the best high-frequency performance. Required for most commercial and data center installations. Risk: if the two ends connect to different ground potentials (e.g., separate building ground rods), the shield carries 60 Hz ground-loop current, which couples into the signal pairs as hum.
• Single-end grounding (special case): The shield is bonded at the patch panel only; the equipment end floats. This eliminates ground-loop current but reduces high-frequency shielding effectiveness by roughly 6 dB and is not recommended for frequencies above 100 MHz. Used primarily in audio/video installations where 60 Hz hum is the dominant concern.

Proper shielded connector termination: the drain wire must make continuous contact with the metal connector shell — this single step determines whether your STP cable is a shield or an antenna

Alien Crosstalk: Why Cat6A Changed the Game

What Alien Crosstalk Is (ANEXT / PSANEXT)

Alien crosstalk is electromagnetic coupling between adjacent cables in a bundle — distinct from internal crosstalk (NEXT/FEXT), which occurs between pairs within the same cable. It is measured as:

• ANEXT (Alien Near-End Crosstalk): Noise coupled from a disturbing pair in one cable to a disturbed pair in an adjacent cable, measured at the same end
• PSANEXT (Power Sum Alien Near-End Crosstalk): The sum of ANEXT contributions from all disturbing pairs in the six surrounding cables in a standard seven-cable test bundle

Why It Became a Certifier Parameter at Cat6A

At Cat5e (100 MHz) and Cat6 (250 MHz), alien crosstalk is measurable but rarely limiting — the twist-rate variation between pairs and cables provides enough random decoupling. At Cat6A (500 MHz), the wavelength becomes short enough (~60 cm) that bundle coupling becomes highly efficient. In a U/UTP Cat6A bundle of 48 cables, PSANEXT is the parameter most likely to fail certification — not insertion loss, not internal NEXT.

The standard TIA-568-C.2 requires PSANEXT testing of Cat6A links in bundles of seven cables — and this is a mandatory, not optional, part of field certification. A Cat6A U/UTP cable that passes all other parameters can still fail PSANEXT if the bundle density exceeds the cable's design assumptions.

How Shielding Eliminates ANEXT

The individual pair foil in U/FTP and S/FTP construction effectively creates a Faraday cage around each pair, blocking the electric field that drives capacitive crosstalk. Cable-to-cable ANEXT becomes limited to magnetic coupling through the braid — and at 500 MHz, magnetic coupling is far weaker than capacitive coupling in the near field.

High-density data centers using Cat6a 10G cabling face severe alien crosstalk with UTP, which STP cables effectively suppress. Learn testing specs and layout optimization in our AMPCOM’s article: Alien Crosstalk in High-Density Bundles: Testing & Mitigation.

Cost Analysis: Cable + Installation + Lifecycle

The Three Cost Dimensions Nobody Accounts For

The purchase price of the cable spool is the smallest part of the total cost equation. A complete cost analysis for UTP vs. STP must include:

The Rework Math That Changes the Equation

Installation comparison: UTP bundles (left) are lighter and faster to pull but face ANEXT challenges in high-density trays; S/FTP (right) adds weight and termination time but passes certification predictably

Installation Considerations by Shield Type

Physical Properties That Impact Labor

Shield Continuity Testing

UTP certification requires only wire map, length, insertion loss, NEXT, PSNEXT, ACR-F, PSACRF, and return loss. STP certification adds one critical test: shield continuity. A DC resistance measurement between the connector shells at both ends must show continuity (typically < 1 ohm for a 90-meter permanent link). Any open circuit indicates a broken shield path — and the link fails certification even if all signal parameters pass.

Fluke DSX-8000 and equivalent certifiers perform shield continuity as part of the autotest sequence when the shielded adapter is selected. If the tester reports "Shield Open," the most common causes are: drain wire not captured in the connector, foil torn during jacket stripping, or unshielded keystone jack installed in the channel.

Decision Framework: When UTP Works, When STP Is Mandatory

The Four-Question Decision Tree

Instead of starting with cable type, start with your deployment conditions and work backward:

Question 1 — What speed?
Gigabit (1000BASE-T) → UTP is fine for nearly all environments at standard distances. 10GBASE-T (Cat6A/500 MHz) → the ANEXT question becomes dominant. 25G/40GBASE-T (Cat8/2000 MHz) → S/FTP is mandatory; there is no UTP Cat8.

Question 2 — What is the EMI environment?
Run a spectrum analyzer sweep (or rent one for a day) at the cable pathway locations. If you detect field strengths above 3 V/m at frequencies from 30 MHz to 500 MHz, you are in the territory where UTP CMRR may be insufficient. Common high-EMI indicators: proximity to VFD motor drives, MRI rooms, arc welding stations, AM/FM broadcast antennas within 500 meters, or shared conduit with power cables.

Question 3 — What is the bundle density?
If your cable trays will carry more than 24 Cat6A cables in a shared pathway, the PSANEXT math favors shielding. The cost of S/FTP cable becomes cheaper than the cost of ANEXT remediation when bundle size exceeds roughly 30 cables.

Question 4 — What is the building's grounding infrastructure?
S/FTP only works if you have a proper TGB (Telecommunications Grounding Busbar) bonded to the building's electrical ground. In older buildings without a verified TGB, deploying STP is risky — you may spend more on cable and still fail certification because you cannot provide a valid ground reference. If the building lacks a TGB, either budget for the grounding retrofit or stay with UTP and accept the bundle density limits.

Quick-Reference Decision Matrix

Real-World Deployment Scenarios

What's Changing: Cable Shielding Trends Through 2030

Single-Pair Ethernet (SPE) and the Shielding Question

IEEE 802.3cg (10BASE-T1L) and 802.3bp (1000BASE-T1) define single-pair Ethernet for industrial IoT and building automation. These standards operate over a single twisted pair — no differential cancellation between pairs, because there is only one pair. Shielding becomes essential: 10BASE-T1L over 1 km requires overall foil shielding as a practical minimum, and industrial SPE deployments near motor drives almost universally use S/FTP-like construction on the single pair. The cable is simpler (one pair), but the shielding requirement is stricter per pair than traditional four-pair Ethernet.

Cat8 and the End of Unshielded Copper

Cat8 (ISO/IEC Class I, 2000 MHz) is the first Ethernet category where there is no unshielded option. The 2 GHz frequency ceiling means the wavelength inside the cable is approximately 10 cm — short enough that even a single pair twist cycle functions as an efficient antenna element. TIA-568-C.2-1 defines Cat8 exclusively as S/FTP or SF/FTP construction. For 25GBASE-T and 40GBASE-T at 30-meter channel lengths, the S/FTP construction is not a "better" choice — it is the only choice.

POE++ and Shield Heat Dissipation

IEEE 802.3bt (Type 4 PoE, 90W per port) introduces a thermal dimension to the shielding decision. When a cable bundle carries 90W on multiple pairs simultaneously, the DC resistance of the copper generates heat — approximately 1.5–2.0°C per watt in a 48-cable bundle. Shielded cables dissipate this heat more efficiently because the metal foil and braid act as thermal spreaders, conducting heat away from the conductor bundle to the cable jacket surface. In a TIA TSB-184-A compliant PoE bundle, S/FTP cables show 3–5°C lower internal temperature rise than U/UTP cables at the same power level — a difference that directly affects insertion loss (copper resistance increases ~0.4% per °C).

Cat8 S/FTP certification: at 2000 MHz, shield continuity testing is not optional — the Fluke DSX performs it as part of every autotest sequence

Key Questions & Answers

AMPCOM Technical Team

Industry experts with 17+ years in enterprise network infrastructure and structured cabling systems