When a machine misbehaves at 2am, this is the skill that saves you. Online monitoring, the event log, and EtherCAT bus diagnostics that point straight at the failed node. Let's hunt a dropped slave.
EtherCAT is a daisy-chain: the frame visits each slave in order. When one drops, everything after it goes dark too — so the first "unreachable" node is your culprit. Press the fault, then click the node you think failed.
Controller event log
🔍 Online + breakpoints. Watch live variable values as the program runs; set a breakpoint to freeze on a condition and step line by line. Force a value to test a branch without rewiring.
📈 Data Trace samples variables every task cycle into a scope view — the only way to see a glitch that's gone before your eye catches it.
📋 Event log + error codes. The controller timestamps every fault with a code you can look up. Read it first — it usually names the unit and the cause.
🧭 The method: reproduce → read the log → narrow with trace/breakpoints → fix → confirm. Don't guess; let the controller tell you.
On an EtherCAT chain, slaves 3, 4 and 5 all report unreachable at once. Where's the most likely fault?