Troubleshooting CAN FD Networks: A Practical Diagnostic Workflow Using Raptor Breakout Boxes and Kvaser CAN Interfaces

CAN FD's higher data-phase bit rates make a network faster, but they also make it far less forgiving of the wiring faults, bad terminations and marginal transceivers that a classic CAN bus would simply shrug off. At EMB Power, New Eagle's exclusive UK and EU partner for the Raptor toolchain and a Kvaser qualified sales representative, we support controls engineers troubleshooting exactly this problem — on the bench, in-vehicle and on off-highway machinery — using Raptor breakout boxes and Kvaser's CAN interfaces. Here's the systematic workflow we use.

Why CAN FD Networks Fail Differently to Classic CAN

CAN FD (Flexible Data-Rate) raises the data-phase bit rate well beyond classic CAN's 1 Mbit/s ceiling, which is what makes it attractive for dense, high-bandwidth vehicle and machine networks. The trade-off is that signal reflections, marginal terminations and wiring faults that a 500 kbit/s classic CAN bus tolerates can produce bit, form or CRC errors at CAN FD data-phase rates. Three things make CAN FD troubleshooting its own discipline rather than a faster version of classic CAN diagnostics: the higher data rates expose signal integrity issues that were previously invisible; mixed networks combining legacy CAN 2.0 and CAN FD nodes can introduce protocol mismatches; and precise timing and termination become critical rather than merely good practice.

Build the Right Toolkit Before You Start

A systematic CAN FD diagnosis depends on three pieces of hardware working together, each doing a different job:

• A Raptor breakout box with test lead kit — gives physical access to CAN network pins for safe probing and selective isolation, without cutting into a harness. New Eagle produces these per control module (for example, the Raptor GCM48 Breakout Box with Test Lead Kit for a GCM48), so the right box depends on which Raptor module is on the network.
• A Kvaser CAN FD interface — the Kvaser USBcan Pro 4xHS is the tool for this job specifically because it captures CAN FD traffic (up to 8 Mbit/s in the data phase, subject to the physical layer being implemented correctly), with four channels, galvanic isolation and microsecond-level timestamping across channels. A classic-CAN-only interface such as the single-channel Kvaser Leaf Light R v2 still has a place on these networks — for probing a classic CAN segment, or bench work where FD capture isn't the point — but it won't show you FD data-phase traffic, so don't rely on it for the FD-specific steps below.
• An automotive-grade multimeter and oscilloscope — for resistance, voltage and waveform checks directly from the breakout box's test points.

A Five-Step Diagnostic Workflow

Work through these steps in order — each one rules out a category of fault before you move to the next, which is what keeps troubleshooting time-bounded instead of open-ended.

1. Power down and check the physical layer. With the network unpowered, use the breakout box to check resistance across CAN_H and CAN_L at several points along the bus. A healthy, correctly terminated network reads close to 60Ω (two 120Ω termination resistors in parallel); a reading below ~50Ω suggests a short, and above ~70Ω suggests a missing or failed termination. Inspect connectors, harnesses and breakout box pins visually for damage or corrosion while you're there.
2. Power up and confirm idle bus behaviour. Reconnect power and measure voltage at CAN_H and CAN_L using the breakout box's test points — this disturbs the rest of the network far less than tapping into the harness directly. A significant or asymmetric deviation from the expected recessive-state levels usually points to a transceiver fault or a wiring problem rather than a software issue.
3. Capture CAN FD traffic with the Kvaser USBcan Pro 4xHS. Connect it to the breakout box and log bus traffic in silent mode so you observe without disturbing the network. Watch for recurring bit, form, ACK or CRC errors, especially around the arbitration-to-data-phase transition — that boundary is where CAN FD-specific faults tend to surface first. Raptor CAL is useful alongside this for confirming how the ECU itself is responding post-capture.
4. Isolate the fault using the breakout box. Selectively disconnect segments or nodes until the error frames or bus-off condition clears. This is the most reliable way to find an intermittent fault caused by a single marginal device, since swapping or removing one node at a time narrows the search far faster than analysing traffic alone.
5. Check mixed CAN/CAN FD segments separately. A legacy CAN 2.0 node in an error-passive or bus-off state can disrupt timing for CAN FD nodes sharing the same physical segment. Use the analyser's listen-only mode to observe without further loading a vulnerable legacy node, and use the breakout box to physically segment CAN FD-only regions where the architecture allows it.

Reading the Waveform When the Trace Alone Isn't Enough

When error frames persist but their cause isn't obvious from the trace, connect an oscilloscope through the breakout box to inspect the physical waveform directly. At CAN FD data-phase rates, reflections from a missing or incorrect termination show up clearly as double pulses or standing waves that are easy to miss in a purely digital trace. Synchronising the oscilloscope capture with the Kvaser software's error-frame log lets you correlate a specific electrical anomaly with the exact frame it corrupted, rather than guessing at the connection.

Fault Patterns and What They Usually Mean

Keeping Diagnostics Reliable Over Time

A few habits make CAN FD troubleshooting repeatable rather than a fresh investigation every time: record measured resistances and idle voltages at initial build and after every modification, so you have a baseline to compare against later; always use the breakout box's test points rather than tapping into the harness directly, which keeps the harness intact and avoids introducing new faults while you're looking for existing ones; and account for electromagnetic interference when testing in-vehicle — shielded cabling, tight routing and proper chassis grounding all reduce false positives during a capture.

Why Engineering Teams Work With EMB Power

EMB Power is New Eagle's exclusive UK and EU supplier of Raptor control hardware and the Raptor-Dev model-based development toolchain, and now of the OpenECU range following New Eagle's 2026 acquisition of Pi Innovo. That gives UK and EU controls teams direct access to the right breakout box for their specific Raptor module, Kvaser's CAN FD-capable interfaces, and engineering support from a team that works with this toolchain daily — rather than a generic reseller handoff when a network issue turns out to be harder to find than expected.

FAQ

Can I use a classic-CAN-only interface like the Kvaser Leaf Light R v2 to troubleshoot a CAN FD network?

It has a role for probing classic CAN segments or basic bench connectivity, but it won't capture CAN FD data-phase traffic. For the FD-specific steps in this workflow, use an FD-capable interface such as the Kvaser USBcan Pro 4xHS.

Do I need a different breakout box for every Raptor control module?

Yes — New Eagle's breakout boxes are built per module (GCM48, GCM70, RCM80, RCM112 and so on), matched to that module's pinout and connector. EMB Power can confirm the right breakout box for your specific controller.

What's the single most common root cause behind CAN FD bus errors?

In our experience, incorrect or missing 120Ω termination at the ends of the bus accounts for the largest share, followed by baud rate mismatches between nodes and connector or wiring faults that only show up once vibration or thermal cycling is in play.

Specifications based on published New Eagle and Kvaser product data as of October 2026; confirm current specification and availability with EMB Power before finalising a design.

If you're specifying a CAN FD network, or need support getting to the bottom of an intermittent fault on one, contact EMB Power to talk through the right mix of Raptor and Kvaser diagnostic hardware for your project.