Week 4 · Lesson 4

UTP/STP Cabling & Network Interface Cards

The humble RJ-45 patch lead is the most common network cable on Earth, and the Network Interface Card decides how fast it runs. This lesson shows how twisted pair defeats interference, how the two wiring standards differ, which of the three cable types to use where, and what speed, duplex, and autonegotiation actually mean.

CCNA 200-301 Domain 1.0 Network Fundamentals Verified 2026-10-11
Part 1

UTP construction

Unshielded twisted pair (UTP) is the cable that fills offices, homes, and server rooms. It carries an RJ-45 plug and contains exactly four pairs of insulated copper wires — eight conductors in total. Within each pair the two wires are twisted around each other at a carefully chosen rate (a different number of twists per metre for each of the four pairs), and that simple twist is the whole trick.

A durable plastic jacket surrounds the four pairs. UTP has no metal shielding beyond the twisted conductors themselves, which keeps it cheap, thin, and easy to terminate. Because it is the default cabling for Ethernet, the CCNA expects you to understand both what it is and why it is built this way.

Definition — twisted pair. A cable in which two insulated copper conductors are twisted together; the two wires carry a differential signal (equal and opposite voltages), and the twist makes the pair resistant to outside noise and to interference from the neighbouring pairs.

Conductors

Four pairs, eight insulated copper wires, each pair a separate twisted circuit.

Twist rate

Different twists per metre for each pair, so the pairs do not line up and reinforce each other's crosstalk.

Termination

All eight conductors land in the 8P8C RJ-45 plug, which is why gigabit and 10G can use every pair at once.

Part 2

Why twisting works

Two enemies attack signals on copper: crosstalk and electromagnetic interference (EMI).

Crosstalk
A signal on one pair induces a small unwanted voltage in an adjacent pair. It is worst when two wires run parallel over a long distance.
EMI
Outside sources — motors, fluorescent ballasts, power cables, radio transmitters — inject noise onto the wires.

The twist defeats both. Because the two wires of a pair are twisted, they constantly swap positions relative to a noise source, so any noise or crosstalk couples into both wires almost equally. The receiver reads the difference between the two wires, and the common noise cancels out — a property called common-mode rejection. The tighter and more uniform the twists, and the more different the twist rates between pairs, the better the cancellation. Twisting also reduces how much radio energy leaks out, which is why UTP is less of a security and interference problem than flat untwisted pair.

Memory hook. A pair is two wires carrying equal and opposite signals. Twist them and both wires see the same noise; subtract the two at the receiver and the noise disappears. That is the entire reason twisted pair won the cabling war.
Part 3

Shielding: UTP, FTP, STP, S/STP

When a cable runs through an electrically noisy environment (a factory floor, next to motors or power distribution), a metal foil or braid can be added. The four common variants differ only in what is shielded.

TypeShieldingOne-line summary
UTPNoneUnshielded twisted pair; four pairs in a jacket. The default, cheapest, most common Ethernet cable.
FTPOverall foil shield around all pairsFoil Twisted Pair; one shared foil screen protects all four pairs from outside EMI.
STPShield per pair (plus often an overall shield)Shielded Twisted Pair; each pair is individually screened, giving the best crosstalk protection.
S/STPBoth: pair shields and an overall braid/foilScreened Shielded Twisted Pair; per-pair shields inside an overall screen — maximum protection, highest cost.
Note. Terminology drifts between vendors (STP is sometimes used loosely for any shielded cable), but the exam idea is simple: more shielding = more EMI resistance, more cost, more bulk, and the shield must be grounded properly. A shield that is not grounded can act as an antenna and make things worse.
Pick the shielding variant · select one

Part 4

568A vs 568B pinouts

The RJ-45 plug has eight pins and the cable has four pairs. The TIA/EIA-568 standard defines two ways to map those pairs onto the pins: 568A and 568B. They are almost identical — the only difference is that pairs 2 and 3 are swapped. In both standards:

In 568B the orange pair is on pins 1–2 and the green pair on pins 3–6. In 568A that is reversed: green on 1–2 and orange on 3–6.

Exam tip. A cable is straight-through if both ends use the same standard — 568A on both ends, or 568B on both ends. It does not matter which one you pick, as long as you pick the same one. 568B is the more common in practice. A cable is a crossover if one end is 568A and the other is 568B. Remember the pair on pins 1,2 and 3,6 is the one that moves.
Pinout explorer · select the wiring standard

Part 5

Straight-through, crossover, rollover

Ethernet historically used two different transmission pins, so the wiring of the cable depended on the devices at each end. Three cable functions matter:

CableWiringConnects
Straight-throughSame standard both ends (A–A or B–B)Different device types: PC ↔ switch, switch ↔ router
CrossoverA on one end, B on the otherSame device types: switch ↔ switch, PC ↔ PC, router ↔ router
Rollover (console)Pin order fully reversed (1↔8, 2↔7 …)A PC’s serial/USB port to a device’s console port

The logic behind straight-through versus crossover is that a transmitter must reach a receiver. A PC’s transmit pins are 1 and 2 while a switch’s receive pins are 3 and 6, so a straight-through cable already crosses them over internally at the two device pinouts. Two PCs both transmit on 1,2, so connecting them directly needs a crossover to route 1,2 into the other side’s 3,6. The rollover cable is different in purpose: it is not for data at all but for out-of-band management, letting a technician open a console session on a switch or router even before it has any IP address.

Auto-MDIX. Modern switches support auto-MDIX (automatic medium-dependent interface crossover). The port senses which pins are transmitting and electrically swaps its own transmit/receive pair to match, so either a straight-through or a crossover cable works. This is why crossovers are increasingly rare. On the exam, still answer with the classic rule unless the question explicitly mentions auto-MDIX.

Which cable for which job?

Exam scenario questions usually ask you to pick the correct cable between two named devices. Use this decision table:

ScenarioCableReason
PC → switch access portStraight-throughDifferent device types
Switch → router Ethernet portStraight-throughDifferent device types
Switch → switch (older, no auto-MDIX)CrossoverSame device types
PC → PC directlyCrossoverSame device types
Router → routerCrossoverSame device types
PC → device console portRollover (console)Out-of-band management
Anything → modern switchStraight-through or crossoverAuto-MDIX makes either work
Cable chooser · pick the scenario

Part 6

Network interface cards, speed, duplex, and autonegotiation

A Network Interface Card (NIC) is the hardware that connects a host to the network medium. In a desktop PC it is a card or a chip on the motherboard; in a laptop it is built in; in a phone it is a chip on the board. There are two broad kinds:

A NIC works at two OSI layers at once, and the exam likes that fact:

Exam tip. If asked which layers a NIC operates at, the answer is Layers 1 and 2. Its MAC address is a Layer 2 identity; its signalling and connectors are Layer 1.

Speed, duplex, and autonegotiation

The link between a NIC and a switch port is governed by three settings:

SettingMeaningNotes
SpeedBits per second the link carriesCommon values: 10 Mbps, 100 Mbps, 1 Gbps, 10 Gbps. The two ends must agree.
DuplexWhether both directions can be used at onceHalf-duplex = send or receive, collisions possible; full-duplex = send and receive simultaneously, no collisions.
AutonegotiationAutomatic agreement of speed and duplexDevices exchange capability signals so the best common settings are chosen automatically.

Full-duplex is the normal mode on modern switched Ethernet: each pair (or each direction on the wire) is used for one direction only, so collision detection is turned off and throughput can roughly double. Half-duplex still uses the older CSMA/CD rules to arbitrate who may transmit.

Watch out — duplex mismatch. A common real-world fault is a duplex mismatch, where one end is hard-set to full-duplex and the other negotiates to half-duplex. The link comes up but runs badly: late collisions, CRC errors, and very poor throughput. Best practice is either to let both ends autonegotiate, or to hard-set both ends to the same speed and duplex. Never hard-set one side only.
Which layer does this NIC job belong to? · select a job

Part 7 · Your turn

Review

Check what you have learned with these review questions.

Review · 6 questions

Sources

Sources & references

  1. ANSI/TIA. TIA/EIA-568 Commercial Building Telecommunications Cabling Standard (568A and 568B pin/pair assignments). tiaonline.org.
  2. IEEE. IEEE 802.3 (Ethernet; twisted-pair PHYs and the MDI/MDIX interface). standards.ieee.org.
  3. Cisco. CCNA (200-301) Exam Topics — Network Fundamentals (cabling and interfaces). cisco.com.
  4. Cisco. CCNA: Introduction to Networks — “Network Media” module (UTP construction, straight-through vs crossover). netacad.com.

Target exams: CCNA 200-301 v1.1 · CyberOps Associate 200-201. Last content verification: 2026-10-11.