How to Spec Automotive Cables: The 5-Step Checklist I Use After $12,000 in Mistakes

If you're specifying automotive cables, connectors, fiber optics, or complete wiring systems—or you're working with a supplier like LEONI on a custom harness—this checklist is for you. I've been handling cable and connectivity orders for eight years, and I've personally made (and documented) enough specification mistakes to total roughly $12,000 in rework and wasted budget. Take it from someone who's paid for these lessons: this is the exact checklist I now run before every order.

Five steps. Each one takes a few minutes. Skipping any of them will eventually cost you more than the time you saved.

Step 1: Define the operating environment, not just the application

"It's for a car" is not a specification. I learned that the hard way in 2017, when I approved a standard PVC-insulated cable for what turned out to be an engine-bay application. The customer said "engine compartment," I heard "under the hood," and I never asked the obvious follow-up questions. Six months later, the insulation was cracking from heat cycling. Every single unit on that $2,400 order went to scrap, plus we ate a one-week production delay.

Before you finalize any automotive cable order, write down:

• Temperature range—ambient and continuous, with realistic peaks (engine bays can exceed 125°C)
• Fluid exposure—oil, fuel, coolant, brake fluid, road salt
• Vibration and abrasion—is the harness near sharp edges, moving parts, or in a high-vibration zone?
• Routing constraints—grommets, tight passages, exposure to the outside

When I compared our Q1 and Q2 spec sheets side by side that year—same supplier, different amounts of detail—I finally understood why the details matter so much. The cable that failed wasn't "cheap." It was specified for the wrong environment.

Step 2: Size the conductor by current and voltage drop—not "bigger is safer"

The "thicker wire has more headroom, so why not" thinking comes from an era when vehicle harnesses only carried a few basic loads. That era is gone. Today, with ECUs, sensors, and infotainment, every additional millimeter of bundle diameter affects weight, routing, and cost.

Here's what I do instead:

1. Calculate the actual current draw per circuit, including inrush where applicable.
2. Check the ampacity table for the cable family—LEONI automotive cables, for example, have published current ratings per cross-section.
3. Apply derating factors for bundling and ambient temperature.
4. Verify voltage drop over the full wire length. Maximum current isn't the only constraint.

People think bigger conductors are automatically better. Actually, an oversized cable is often worse: it's heavier, stiffer, harder to route, and it puts unnecessary strain on terminals and connectors. The assumption that "thick = safe" has the causation backwards. The right cross-section is safe. The wrong one—even with generous headroom—creates problems you didn't have before.

Step 3: Match the connector to the cable, and the terminal to the tooling

This is the one that hurt the most. In September 2022, a harness failed intermittently during testing. The cable spec was perfect. The connector was a well-known automotive-grade part. So why did the circuit keep dropping out?

Turns out the terminal was crimped onto a cable with the wrong insulation diameter, so it never fully seated in the connector housing. The cable was fine. The connector was fine. The combination wasn't. Nobody caught it because nobody cross-checked the terminal's crimp range against the cable's insulation diameter. That mistake scuttled a $3,200 order.

Never expected a failure to come from two good components that simply didn't belong together. The surprise wasn't the test failure—it was how invisible it was on paper.

Run through this every time:

• Terminal wire crimp range matches the conductor cross-section.
• Terminal insulation crimp range matches the insulation diameter—not just the wire gauge.
• Connector housing sealing, locking mechanism, and mating interface are verified against the chosen terminal.
• Crimp tooling and dies are qualified for that exact terminal. Lookalike dies produce lookalike crimps.

If you're unsure about compatibility, ask for the application engineer's recommendation, or reference USCAR-2 for connector performance and ISO 6722 for cable construction. These standards exist specifically to prevent this class of error.

Step 4: Respect the bend radius—especially for moving cables

This is the step most people ignore. A cable can have the perfect conductor, insulation, and connector—and fail in weeks if it's routed around a sharp corner or flexed beyond its rating.

People ask "why are phones so strong" when a device survives years of drops, yanks, and pocket abuse. The answer is the same reason a car's wiring harness can survive 15 years and 200,000 miles: someone specified the right materials, bend radius, and strain relief for the real world. The same discipline applies to automotive routing. And I do not mean just static routing—I mean the cable's whole life, including installation and maintenance.

Check these before ordering:

• Minimum bend radius for fixed installation versus dynamic routing.
• Expected flex cycles if the cable moves—doors, seats, trunk lids.
• Strain relief at connector entries and harness branch points.

For high-flex applications, standard cable isn't enough. That's where products like the LEONI DuraForce Pro 2 family come in—reinforced conductors and jacket materials designed for repeated movement. It costs more per meter. Replacing a failed harness mid-production costs a lot more.

Step 5: Build a physical prototype before you commit to production quantities

Here's the counterintuitive one. Buyers and engineers skip prototyping because they think it delays the timeline. In my experience, it does the opposite: a prototype moves failures to a point where they cost you two weeks instead of two months.

After the third rejection in Q1 2024, I made prototype approval mandatory for any new harness design. Since then, we've caught 47 potential errors before production. Not one of them would have been caught by "checking the drawings more carefully."

The rule: if you haven't built and tested one physical prototype, you haven't finalized the spec.

Ray Leoni, our senior application engineer, puts it perfectly:

"A two-week prototype saves you a six-week crisis."

Ray's been in this industry for over 20 years, and I've watched him be right a ton of times. Listen to your Ray.

Before you finalize: the mistakes I still see

Even with these five steps, I keep seeing the same errors. Some of them, I made myself.

• Copying an old order's spec without re-verifying. The old order was right for the old application. Yours is different.
• Trusting a datasheet over the product family documentation. Part numbers get superseded. Verify current specs.
• Comparing price per meter instead of total installed cost. A cheaper cable that needs extra protection, longer assembly time, or causes warranty claims isn't cheaper.
• Skipping the supplier's design review. At LEONI, we offer these as part of the quote process. They catch problems like the terminal mismatch in Step 3. Not using one is like refusing free insurance.

And one last tip: before you place a large order, pick up the phone and make a clear phone call to the application engineer. Not an email. Describe the application, read the spec back, and confirm every item on this list. Emails get skimmed. Phone calls get heard. That single habit has prevented more mistakes than any other item on this checklist.

The bottom line

There's something satisfying about a perfectly specified harness order. After a few years of rework and wasted budget, watching a production run come through clean and on schedule—that's the payoff.

The best part of finally systematizing this process: no more 3 a.m. worry sessions about whether we missed something. If you're about to quote automotive cables, connectors, or a complete wiring system from LEONI or anyone else, run through these five steps first. It'll save you the tuition I paid.