How to Read a Coaxial Cable Datasheet Before You Buy an RF Cable Assembly

A robot program can lose weeks because one coaxial cable was approved like a catalog commodity instead of a controlled RF component. We see it when an AGV passes bench test, then drops GNSS accuracy near steel racking; when a vision system works in prototype, then shows intermittent video loss after the cable route tightens in production; or when a radar harness ships with the right connector and the wrong attenuation budget. The failure usually starts much earlier than the field return. It starts when the buyer reads only the part number and outside diameter, but not the datasheet lines that actually predict signal loss, routing risk, and validation scope.

One mobile-robot team came to us after losing 19 days between pilot build and customer sign-off. Their LTE antenna cable had a familiar coax family name and a clean continuity report. What it did not have was enough margin at the real installed length, enough bend allowance around the roof bracket, or enough shielding discipline beside inverter wiring. The published catalog looked acceptable at first glance. The installed result was not. That gap between "looks standard" and "works in production" is exactly why buyers need a practical method for reading a coaxial cable datasheet.

This guide is written for teams sourcing coaxial cable manufacturers, custom connector solutions, and custom cable assemblies for AGV & AMR platforms, logistics & warehouse robots, and other RF-enabled automation systems. The goal is simple: turn a dense supplier datasheet into a go or no-go decision before you release an RFQ, tooling, or pilot PO.

The first mistake: reading the cable datasheet without the application in front of you

A coax datasheet does not tell you whether the cable is good in the abstract. It tells you whether the cable is good for a specific impedance, frequency band, route geometry, temperature range, and connector system. Buyers get into trouble when they compare only headline values such as "low loss" or "high shield coverage" without writing down the actual installed conditions. A 50 ohm mini-coax that looks excellent in a catalog may still be the wrong choice if the route forces a 20 mm bend where the cable needs 35 mm, or if the link budget is already tight at 3.5 GHz.

Before you read the first table, put five application facts on one sheet:

1. System impedance: usually 50 ohm or 75 ohm.
2. Operating frequency or band: for example GNSS L1, LTE, Wi-Fi, HD video, or radar.
3. Installed length from connector reference plane to connector reference plane.
4. Real mechanical route: bend points, clamp points, and moving or static zones.
5. Acceptance method: continuity only, or continuity plus VSWR, insertion loss, TDR, dielectric, or shielding verification.

If those five points are vague, the datasheet review becomes guesswork.

"A coax datasheet only becomes useful when it is tied to the real route. I would rather review a rough installed sketch with frequency and length than a perfect PDF with no application context."

— Hommer Zhao, Founder, Robotics Cable Assembly

The 8 datasheet lines that actually change a buying decision

Many supplier sheets contain 20 or 30 values. In practice, most purchasing and NPI teams should focus on the eight lines below first.

Two additional lines often matter in high-reliability programs: dielectric withstanding voltage and conductor plating. Those values are not always the top decision drivers, but they become important when the route includes harsh temperature swing, long storage life, corrosion pressure, or strict validation documentation.

Start with impedance and frequency, because they eliminate bad options fast

The fastest way to reject the wrong coax is to check impedance and attenuation together. Coaxial cable systems are not forgiving when the cable family, connector family, and equipment ports assume different impedances. In robotics, 50 ohm is common for GNSS, LTE, Wi-Fi, telemetry, RFID, and radar-style links. 75 ohm appears more often in video and certain vision or broadcast-derived architectures. If the system path is 50 ohm, a 75 ohm cable is not a bargain. It is a mismatch cost that will show up later in reflection, loss, or unstable diagnostics.

After impedance, go straight to attenuation at frequency. Do not stop at a datasheet line that says "low loss." Find the value nearest your real operating frequency and convert it into the actual installed length. If the sheet publishes 48 dB/100 m @ 1 GHz, that becomes 0.48 dB/m @ 1 GHz. A 3.2 m installed run would then contribute about 1.54 dB of cable loss before connector effects. That number is still only one piece of the link, but it is the number that engineering can actually review.

The exact approval threshold depends on the radio, antenna gain, and full system margin. The point is that procurement should never compare coax only by unit price when one cable consumes 0.8 dB and another consumes 1.5 dB in the same route.

"When buyers convert attenuation into the real installed length, technical debate gets much cleaner. A number like 1.54 dB at 3.2 m is actionable. 'Low loss' is not."

— Hommer Zhao, Founder, Robotics Cable Assembly

Then read the mechanical lines like they are electrical risk

Robotics teams often separate electrical review from mechanical routing review. For coax, that is a mistake. Minimum bend radius, outer diameter, jacket material, and shield construction all affect whether the cable still behaves like controlled impedance after installation. If the route crushes the dielectric, kinks the braid, or forces repeated bending below the published limit, the electrical path changes even if continuity still passes.

That is why bend radius belongs in the same review as attenuation. A cable with the right loss number can still be wrong for a compact AMR roof route, a rotating sensor mast, or a cabinet exit beside a gland plate. For moving or repeatedly serviced paths, use the datasheet as a starting point, then ask whether the supplier has application-relevant flex data. Static bend radius from a catalog is not the same thing as validated life inside a moving robot route.

Also pay attention to shield construction wording. "95% braid" sounds strong, but it does not automatically describe transfer impedance, foil overlap, drain-wire strategy, or real EMI performance beside inverter cabling, motor leads, or high-current charging circuits. If your system has noisy surroundings, ask for construction detail and route the cable as part of the EMC plan, not as an afterthought.

The short checklist for dielectric, jacket, and environment

This is where many "equivalent" substitutions fail. Two cables can share impedance and similar attenuation while behaving very differently in heat, flex, or chemical exposure because the dielectric and jacket systems differ.

• PTFE or FEP constructions often buy temperature margin and cleaner high-frequency behavior, but they may cost more.
• PE or foam PE dielectric may help attenuation and velocity factor, but the route still has to protect the structure from crush.
• PVC jackets can be commercially attractive in static indoor runs, but they are often a poor fit for tighter temperature, abrasion, or cleaning exposure.
• PUR or similar tougher jackets may improve wear performance, but always verify the exact temperature and chemical limits instead of assuming.

If the application includes washdown, outdoor UV, charger heat, battery-acid risk, or repetitive service opening, the cable datasheet must be read beside the environment notes. This is also where standards language can help. References such as MIL-DTL-17 or IP code can clarify expectation, but they do not replace route-specific engineering review.

"Equivalent coax is one of the most expensive phrases in sourcing. If the dielectric, jacket, or shield build changes, the cable is only equivalent after the installed route and test plan say it is."

— Hommer Zhao, Founder, Robotics Cable Assembly

How to turn a datasheet review into a better RFQ

The best buyers do not send a supplier only a part number and target length. They send the evidence needed to confirm whether that part number survives real production use. After reviewing the datasheet, your RFQ should include:

• Drawing, route sketch, or photos showing connector orientation and clamp locations.
• BOM or approved cable family reference, including any alternates already under consideration.
• Operating frequency or band, system impedance, and estimated allowable link loss if known.
• Installed length, prototype quantity, annual volume, and target lead time.
• Environment details: temperature range, vibration, abrasion, moisture, chemicals, and nearby noise sources.
• Compliance target and documentation expectation, such as traceability, sample test report, or first-article package.
• Validation scope: continuity, pin map, VSWR, insertion loss, TDR, retention, dielectric, or environmental tests.

That package speeds quoting and reduces fake alignment. A supplier can quickly tell you whether the catalog cable is suitable, whether the connector termination changes the result, and whether a custom assembly is safer than a generic patch approach.

FAQ

Which three coax datasheet fields should I check first?

Start with impedance, attenuation at your operating frequency, and minimum bend radius. If those three items do not match the connector family, installed route, and signal budget, the rest of the datasheet will not save the design.

Is a 50 ohm coax datasheet always better for robotics?

No. 50 ohm is common for GNSS, LTE, Wi-Fi, telemetry, and radar links, while 75 ohm is common for video and some vision systems. The cable must match the system impedance end to end.

What attenuation number should procurement put into the RFQ?

Use the attenuation at the real operating frequency, not only the lowest published test point. Many buyers convert the datasheet value from dB/100 m into the actual installed length so engineering can review the real link loss directly.

Why does bend radius belong in the electrical review?

Because a cable that fits electrically can still fail mechanically. In robotic routes, repeated bending below the published limit distorts impedance, damages the dielectric, and shortens flex life.

Does shield coverage tell me everything about EMI performance?

No. Shield coverage is useful, but transfer impedance, braid construction, foil design, grounding method, and route separation also matter. A 95% braid figure alone does not guarantee quiet signal performance.

What should I send a supplier after reviewing the datasheet?

Send the drawing or route sketch, BOM, target quantity, environment, connector family, compliance target, and required test scope. That lets the supplier confirm whether the catalog cable really matches production risk.

Send the next package, not just the PDF

If you are qualifying a coaxial cable for an RF cable assembly, send the drawing or route sketch, BOM, quantity split, installed environment, target lead time, and compliance target next. Include the operating band, connector family, and any attenuation or test limit you already know. We will send back a manufacturability review, recommended cable-and-connector stack, risk notes on routing and shielding, a proposed validation scope, and a quote aligned to prototype and production builds.