What determines cable assembly loss?

Total insertion loss has three components: conductor loss (skin effect in center conductor and shield, proportional to sqrt(f)), dielectric loss (proportional to f, dominant at higher frequencies), and connector loss (typically 0.05 to 0.3 dB per connector, depending on type and frequency). Conductor loss: the center conductor and shield carry RF current in the skin depth (delta = sqrt(2 rho / (omega mu))). Smaller cables have higher conductor loss because the current-carrying area is smaller. At 18 GHz: RG-405 (0.085 inch semi-rigid) has approximately 0.8 dB/ft conductor loss. Dielectric loss: PTFE (Teflon) is the standard dielectric with tan delta approximately 0.0002 at 10 GHz. Low-loss cables use expanded PTFE or air-spaced dielectrics. Total: a 1-meter SMA semi-rigid cable at 18 GHz has approximately 2 to 3 dB insertion loss. Phase-stable test cables use loose-fitting outer conductors to maintain phase within plus or minus 2 degrees over thousands of flex cycles.

What cable types are available?

Semi-rigid: solid copper outer conductor, PTFE dielectric, copper or silver-plated center conductor. Best electrical performance (lowest loss, highest shielding greater than 100 dB, best phase stability). Can be bent once to shape but cannot be flexed repeatedly. Sizes: 0.020, 0.034, 0.047, 0.085, 0.141, 0.250 inch OD. Conformable (hand-formable): corrugated or helically-wound outer conductor allows repeated bending. Performance close to semi-rigid but with slightly higher loss. Good compromise for internal equipment routing. Flexible: braided outer shield, allows repeated flexing. Test-grade cables use armored construction with phase-stable design. Shielding: 60 to 90 dB depending on braid density. Ultra-flexible: multiple braids or spiral wrap, designed for moving applications (antenna positioners, robotic test systems). Higher loss but survives millions of flex cycles.

How are cable assemblies tested?

Every cable assembly is tested before shipment. Standard tests include: insertion loss (S21) measured versus frequency from DC to the rated maximum. VSWR or return loss (S11) at both ends: must be below specification (typically VSWR less than 1.3 or RL greater than 17 dB) across the entire band. Phase versus frequency: for phase-matched pairs, phase difference must be within specification (typically plus or minus 2 to 5 degrees). Shielding effectiveness: transfer impedance or shielding measurement per MIL-DTL-17. Power handling: tested at maximum rated power with temperature monitoring. Phase stability: for test cables, the change in phase with flexing is measured (typically less than 5 degrees for a 90-degree bend). Mechanical: pull force on connectors (typically greater than 20 lbs), torque on coupling nuts. Test data is often provided as a certificate with each cable, showing the actual measured performance.

RF Interconnect

Cable Assembly

/kay-bul uh-sem-blee/

Pre-terminated coaxial cable with tested connectors. Loss: conductor (∝√f) + dielectric (∝f) + connector (0.05-0.3 dB each). Semi-rigid: solid Cu outer, >100 dB shielding, best performance. Flexible: braided, 60-90 dB shield, repeated flex. Conformable: hand-formable, near semi-rigid. Phase-stable test: ±2° over 1000s of flex. Tested: IL, RL, phase, shield, power, mechanical.

Shield: >100 dB

VSWR: <1.3

Phase: ±2-5°

Understanding Cable Assemblies

The cable assembly is the most common and most underappreciated component in any RF system. A bad cable can ruin months of circuit design work: a loose connector creates intermittent faults that are nightmare to debug, a phase-unstable cable makes VNA measurements meaningless, and a poorly shielded cable leaks signal that corrupts adjacent channels.

Professional RF engineers learn to treat cables with respect: always torque connectors to specification, never exceed the minimum bend radius, use phase-stable test cables for repeatable measurements, and replace cables at the first sign of degradation. The cable is the weakest link in any measurement chain.

Cable Loss Equations

Total insertion loss:
IL = α_c√f + α_df + IL_conn
α_c: conductor loss coefficient
α_d: dielectric loss coefficient

Skin depth:
δ = √(2ρ/(ωμ))
Cu @ 10 GHz: δ = 0.66 μm

Example (0.085" semi-rigid, 1m):
@1 GHz: 0.3 dB + 0.2 dB conn = 0.5 dB
@10 GHz: 0.8 dB + 0.2 dB = 1.0 dB
@18 GHz: 1.2 dB + 0.3 dB = 1.5 dB

Cable Type Comparison

Type	IL @18GHz/ft	Shielding	Flex Cycles	Application
Semi-rigid 0.085"	0.8 dB	>100 dB	1 (form once)	Internal routing
Conformable	0.9 dB	90-100 dB	100+	Prototype, lab
Phase-stable test	1.0 dB	80-90 dB	100K+	VNA, bench test
Flexible (RG-402)	1.2 dB	60-80 dB	10K+	Field install
Ultra-flex	1.5 dB	60-70 dB	1M+	Positioner, robot

Common Questions

Frequently Asked Questions

Loss sources?

Conductor (∝√f, skin effect), dielectric (∝f, tanδ), connectors (0.05-0.3 dB each). Smaller cables: higher conductor loss (less area). PTFE: tanδ ≈ 0.0002. 1m SMA semi-rigid @18GHz: ~1.5 dB total. Phase-stable: loose outer conductor maintains ±2° over flexing.

Types?

Semi-rigid: solid Cu outer, best performance, form once. Conformable: corrugated outer, repeated bending, near semi-rigid. Flexible: braided, test bench, field. Ultra-flex: millions of cycles, positioners. Phase-stable test: armored, ±2° guaranteed. Size: 0.020"-0.250" OD.

Testing?

Every assembly tested: IL (S21 vs freq), VSWR (S11, both ends, <1.3), phase vs freq, shielding (transfer impedance), power handling (temp monitored), phase stability (flex test), mechanical (pull force >20 lbs, torque). Certificate with actual measured data. Phase-matched pairs: ±2-5° spec.

Related Terms

← Butler Matrix Cable Loss →

RF Cables

Request a Quote

Need custom cable assemblies, phase-matched pairs, or test cables? Contact our team.

Get in Touch