Alarmas, cableado e interferencias

70 notas de campo sobre este tema

Ninguna nota de este tema coincide con esa palabra.

• On the height controller panel press F2 then 6 to open the network settings and confirm the addressing.
• Make sure no two network adapters share an IP — every adapter has to be checked, not just the obvious one.
• Open the platform configuration tool and confirm the configured height-controller IP matches the controller's actual IP.
• Unplug the height controller's network cable: the adapter that reports disconnected is the one wired to the controller.
• Check the IP settings of the adapter that connects to the laser.
• Check the PC host's own network configuration.
• Check the expansion board's supply — the BCL4516E series needs 24 V.
• Expansion board IN lamp: steady yellow means physically connected, flashing green means actually communicating. Both are needed.
• If the network lamp misbehaves, swap in this order: cable, then network adapter, then expansion board.
• These expansion boards all sit on the EtherCAT bus — any communication failure shows up as an EtherCAT alarm.
• If the slave scan errors, connect a single expansion board or a single servo and scan that. If it appears, add slaves back one at a time until the faulty one shows up, then check that slave's settings and power.
• When the scan count does not match, find the missing slave and connect only that one. If it still is not found, check its configuration, wiring and power.
• If the addressing is correct and the timeout alarm persists, swap the cable, then the adapter, then the height controller.
• Bus network alarm with the cable reported as not connected — physically reseat both ends before anything else.
• Check whether the PC and the drive network ports have been swapped.
• Network cable pins bent, deformed or broken, or the cable body kinked at a sharp angle.
• Cable is good: pins straight and intact, no sharp bends. Rule it out and move on.
• Enable lamp not lit means communication between the software and the 4562 terminal board has failed. Unplug the leftmost 62-pin cable, inspect the pins for bending, and use a multimeter to check pin-to-pin continuity at both ends.
• If the signal does not conduct, the link between the 4562 terminal board and the drive is broken — the enable signal never arrives, so check the servo drive cable.
• If the signal does not conduct, the servo drive cable between the height controller and the drive is faulty — the enable signal never arrives.
• If the signal conducts but the drive still will not enable, the wiring is fine and the problem is supply-side: measure the 4562 board's input voltage for stability.
• No continuity between pin 6 and pin 15 means the height controller's servo output is faulty. Unplug everything except the power supply from the back of the controller and measure pins 6 and 15 again.
• Continuity between pins 6 and 15 means the enable output is good — next check the enable signal at the drive end of the servo cable.
• If it still does not conduct with everything unplugged, power-cycle the height controller and measure its supply for stability. If the supply is stable and pins 6-15 still show no continuity with output 6 lit in the test screen, the controller itself has failed — raise a service case.
• Inspect the 62-pin cable for abnormal pins.
• Check whether the encoder feedback pair in the 15-pin cable was ever soldered.
• Check the 15-pin cable for dry joints.
• Likely poor contact in the servo cable or a dry joint at the axis connector. Change the servo cable first, then the drive.
• Read the white label on the servo drive cable: if the brand or the control type printed there does not match what is actually installed (closed-loop), the cable is wrong — swap it.
• On a semi-finished servo drive cable the customer solders themselves, re-check every pin definition and look for dry joints and shorts.
• For any non-finished cable, verify each pin is soldered correctly with no dry joint or poor contact.
• Check whether the drive is showing a non-bus alarm such as overload or encoder fault — that changes the whole diagnosis.
• Check that the servo type is configured correctly; a wrong type mimics a wiring fault.
• If jogging works, the servo receives the motion command but no encoder signal comes back — the fault is inside the drive, not the command path.
• Encoder feedback is abnormal — treat that as the root cause rather than a symptom.
• Take the axis motor off and spin it free: if the alarm persists, it is electrical, not mechanical.
• If the ejector pin is straight, reseat the 62-pin cable; if the X-axis servo alarm remains, the unit has to go back for repair.
• If everything above checks out and homing still misbehaves, open the platform configuration tool and untick the Z-phase signal as a test.
• The origin and limit switches reach the control card correctly — now confirm the signal actually toggles while the machine is moving.
• If the limit input port itself is faulty, replace the adapter board.
• Cut mains power to the servos and the laser, then start the software again — this clears a whole class of stale-state faults.
• If a restart does not clear it, the network port is probably being disturbed from outside: tidy the cabinet wiring and check machine earthing and mains interference.
• That places the interference at the servo or another external load: check their earthing, fit a magnetic ring on the servo motor power cable, and earth the FG terminal of the height controller's supply.
• After enabling the motors, mains interference reaches the IPC's USB port and the keyboard/mouse receiver stops getting through.
• The PC is smooth until the cutting software starts, then the mouse stutters or freezes — look at USB interference and service load, not at the PC.
• Nominal maximum is around 24 V — set the meter range accordingly before measuring.
• Check that the licence dongle driver is installed correctly.
• A licence-expired prompt means the registered code needs re-entering; restart the software afterwards and it runs again.
• No licence found, or the licence has expired — same fix as above.
• Software reports no motion control card detected — check the card seating and the PCIe slot before reinstalling anything.
• If it will not clear, try a different PCI slot, then a different PC.
• If that still fails, replace the motion control card.
• If it still will not clear, replace the network cable, then the 4562 card.
• If it still will not clear, work through the network adapter, the PC host and the camera in that order.
• That software version cannot talk to the slave at all — update to a newer version rather than fighting the wiring.
• If the firmware upgrade still fails, open the network adapter list and check for multiple active ports.
• If a network upgrade is not possible, upgrade from a USB stick instead.
• Copy the upgrade file to the root of the USB stick, not into a folder on it — the controller only looks at the root.
• System delay is measured from File, then Diagnostic tools, then Axis monitor, then System delay measurement.
• Find the axis or axes whose delay is out of line and pull them back towards the others. If Z sits at 7999 while the rest are near 5500, bring Z down to about 5500.
• A delay that is too large is reduced by raising the servo's speed feedforward; a delay that is too small needs it lowered.
• Check the speed gain against the drive parameters. Yaskawa example: with the height controller set to 500 r/min/V, Pn300 should be 6.
• Check the pulses-per-revolution setting against the drive. Yaskawa example: with 10000 set in the height controller, Pn212 should be 2500.
• The parameter must match four times the servo's single-turn feedback pulse count, then be corrected for the actual screw lead.
• Check that Pn000 is set to speed control with zero-point clamp, and that the Pn501 zero-clamp value is normal.
• Zero-speed clamp differs between servos — the principle, the I/O signal logic and the control parameters all change with the brand.
• Where rows of relays sit side by side, leave 20 cm between them or separate them completely — inductively coupled coils are a classic interference source.
• Add earth wires at several points on the bed — near the Y-axis origin of the tube section, mid-bed, and near the middle chuck. On badly disturbed machines 4 to 5 wires at each of those points is what actually brings the interference down.
• Bond the tube-cutting and plate-cutting sections of the bed and the feet together with plate and copper wire, using multiple contact points. Strip paint or powder coating at each contact first — a bonded connection through paint does nothing.
• A DIF value above 30 on the tube section proves interference there. Check whether the plate-cutting section shows it too before you start adding earths.