NUM750 Error: Hardware Mismatch, Not Excess Motion Axes

After the correction, the NUM750 should complete its restart without displaying the “too many axes or spindles” message. Start with the recovered cards and the configured hardware count. A RAM reset exposes mismatches between the installed card set and the configuration; repeatedly changing isolated parameter values does not correct incompatible or incorrectly identified hardware.

Stop Trying the Wrong Fixes

The fault appears during restart, before a test rack can behave like a complete CNC. These common responses waste time:

• Repeating the RAM reset: Another reset clears state again but does not make the recovered cards compatible with one another.
• Copying parameter values without rebuilding the hardware inventory: Values from another arrangement may describe axes, spindles, slots, or resources that this rack does not contain.
• Changing one parameter until the message disappears: That can hide one count mismatch while leaving the underlying card assignment wrong.
• Troubleshooting machine wiring: This installation is a card-test rack, not a machine. Motors, feedback devices, limits, and field wiring are not the first check for a resource-count message raised during CNC initialization.
• Assuming every recovered card belongs in the rack: A card can fit mechanically yet still be the wrong type or revision for the intended configuration.

Start here: inventory the installed cards before editing the parameter set again.

Read the Message as a Configuration Conflict

The controller builds an internal hardware model during startup. It must reconcile the installed cards with the configured axes, spindles, and their assignments. The message means that reconciliation failed; it does not prove that the rack physically contains too many motion devices.

The reported settings are P2 N0=0, P2 N1=7; P6 N0=01, P6 N1=01; P9 N0=0, P9 N1=02, P9 N2=10, with the remaining entries at FF; and P53 N0=2, P53 N1=9. Do not assign meanings to those fields from memory. Read each field against the NUM750 documentation for the exact CPU and card set, then compare the requested resources with the rack inventory.

Prove the Hardware Mismatch First

1. Remove power using the servicing procedure for the rack. Record every installed card, its complete part identification, revision marking, and slot position.
2. Compare that list with the hardware combination intended for the test rack. Check the CPU, axis-control cards, spindle-related cards, and any card that contributes motion resources.
3. Look for duplicate functions, missing companion hardware, and cards recovered from a different CNC arrangement. Treat an unidentified card as incompatible until its documentation says otherwise.
4. Inspect connectors, backplane seating, and card-retention hardware. A card that is present but not recognized can make the configured count exceed the detected count.
5. Restart with the smallest documented compatible card set. Add optional recovered cards one at a time, restarting after each addition. The card that introduces the message narrows the fault to compatibility, slot assignment, card condition, or the configuration required by that card.

Do not substitute cards while energized. Also separate a card-recognition problem from a parameter problem: if the controller does not detect an installed card, no axis-count edit can repair the electrical or compatibility failure.

Rebuild the Configuration That Matches the Rack

1. Keep only the minimum compatible cards required for the controller to initialize.
2. Obtain the correct initialization data or parameter record for that exact hardware combination. A parameter set from a complete machine is not automatically valid for a bench rack.
3. Check every populated entry in P2, P6, P9, and P53 against its documented function. Confirm whether each entry declares a quantity, assignment, slot, or logical resource before changing it.
4. Disable only unused entries through the documented unused-value convention. The supplied P9 list uses FF for remaining entries, but apply that value only where the parameter definition identifies it as valid.
5. Configure no more logical axes or spindles than the recognized card set supports. Remove assignments for hardware that is absent from the rack.
6. Save the complete parameter set, restart once, and record the first message shown. Do not make several unrelated edits between restarts; one change per test preserves the diagnostic trail.

If the minimum card set still raises the message, swap only the suspected card with a known-compatible, known-working card of the documented type. A random recovered replacement adds another unknown and proves nothing.

Verify More Than a Cleared Message

• Run multiple controlled restarts. The message must remain absent without another RAM reset or parameter intervention.
• Confirm that the controller recognizes the expected cards and no unexpected motion resources appear.
• Re-read P2, P6, P9, and P53 after restart. Verify that stored values match the approved rack record.
• Add the remaining cards individually. Record the slot, card identity, parameter change, and result for each addition.
• Test the rack only within its bench purpose. Clearing initialization does not prove that a card can safely control a real axis or spindle.

A disappearing message after removing one card is useful but incomplete. Reinstalling that card with its correct documented configuration distinguishes a bad card from a configuration mismatch.

Avoid the Recurring Bench-Rack Pitfalls

• Label recovered cards with their full identification and last known status. “Recovered” is not a compatibility classification.
• Maintain a parameter backup for each validated rack composition. Do not use one universal file for changing card sets.
• Photograph or diagram slot order before disassembly. Slot-dependent recognition faults often look like count faults.
• Change one variable per restart. Simultaneous card moves and parameter edits destroy the evidence needed to isolate the cause.
• Keep unknown, suspect, and validated cards physically separated. A test rack must not silently promote an unverified card to a reference component.

FAQ