Problem Details
Symptom on a Fadal VMC 15 with a Baldor spindle drive: commanding spindle rotation produces a few revolutions forward, then a few reverse, repeating indefinitely until the operator hits E-stop. After the drive was rebuilt and a new encoder and spindle motor were installed, the fault not only persisted but degraded — the spindle now oscillates forward/reverse rapidly rather than slowly hunting.
Two observations define the failure signature and drive the whole diagnostic path:
- The spindle never reaches commanded speed. It stays in a low-speed, bidirectional hunt. That is characteristic of a position loop chasing a target it cannot satisfy, not a velocity loop failing to accelerate.
- Replacing drive, encoder, and motor did not clear it. When three serially-replaced components in the same power/feedback chain fail to change the outcome, the fault almost certainly lives upstream of the drive (command source, controller card, supply rails, wiring) or in a mechanical/pneumatic interlock, not in the amplifier itself.
Root Cause Analysis
Forward/reverse hunting at low speed with no spin-up is the classic footprint of a spindle orientation failure. During orientation the control closes a position loop on the spindle encoder or orientation sensor and drives the spindle in whichever direction reduces error. If the feedback never reports the in-position window, the loop oscillates around the target forever — exactly what is observed.
The plausible causes, in order of likelihood given that the drive/encoder/motor were already replaced:
| Cause | Mechanism | Why it fits |
|---|---|---|
| Orientation sensor fault or misalignment | Sensor above the main pulley never asserts the marker/in-position signal, so the position loop hunts | Directly produces bidirectional low-speed motion; independent of drive rebuild |
| Air detent cylinder not engaging | The stop/detent cylinder must lock before the sequence completes; without air the control never sees the lock confirmation | Explains why the sequence never releases into normal spin-up |
| Spindle controller card | Card generating the analog speed/direction command or processing orientation feedback is faulty | Only remaining element in the chain after drive, encoder, motor were changed |
| Incoming line / PSU rail problem | Sagging or missing logic rail causes erratic command output and marginal card operation | Cheap to verify, explains "got worse after service" |
| Card seating / connector oxidation | Intermittent feedback or command path through a backplane or edge connector | Common after cabinet work; symptom changed after service |
| Drive rebuild quality / configuration | Rebuilt drive returned with incorrect tuning, jumper, or backup battery state | Symptom worsened rather than cleared after the rebuild |
A second, non-exclusive contributor is the drive's own retained configuration. If the drive stores parameters in volatile memory backed by a battery, a depleted battery can leave it running defaults or corrupted values after any power cycle. Verify the battery condition and the parameter set as part of the drive evaluation rather than assuming a rebuilt unit is correct out of the box.
Diagnostic Procedure
Step 1 — Separate orientation from normal rotation
This is the fastest, highest-value test and costs nothing.
- Power up the machine and clear all faults. Do not command M3 or M4.
- Issue
M19(spindle orient) in MDI. - Observe the behavior.
Result of M19
|
Interpretation |
|---|---|
| Spindle hunts forward/reverse exactly as it does on M3/M4 | Orientation loop is the failing subsystem. Focus on orientation sensor, its alignment and wiring, the air detent, and the controller card handling orientation. |
| Spindle orients cleanly and locks, but M3/M4 still hunts | Orientation feedback is healthy. Move to the speed command path: controller card analog output, drive configuration, drive enable/direction logic. |
| No motion at all on M19 | Enable or command path is dead; check drive enable, run permissive, and interlocks before chasing feedback. |
Step 2 — Pneumatic and detent check
- Verify incoming shop air pressure at the machine regulator against the machine nameplate/regulator setting. Low or fluctuating supply will drop the detent mid-sequence.
- Manually actuate the solenoid that feeds the spindle orientation/detent cylinder (most solenoid valves have a manual override) and confirm the cylinder strokes fully and the detent engages.
- Confirm the solenoid coil is being commanded — measure coil voltage at the connector during the orientation attempt. No command means the fault is in the control output; command present with no motion means a stuck valve, blocked line, or seized cylinder.
- Check any limit/proximity switch that confirms the detent is locked. A miswired or failed confirmation switch will hold the sequence in the hunting state indefinitely.
Step 3 — Orientation sensor and alignment
- Locate the orientation sensor above the main spindle pulley.
- Inspect for coolant ingress, chip packing, cable chafe, and a loose mounting bracket.
- Verify the air gap and radial alignment to the target/magnet on the pulley. If any work was done in this area recently — motor swap, encoder swap, belt service — assume alignment was disturbed until proven otherwise.
- With power on and the spindle rotated slowly by hand (safely, drive disabled), watch the sensor output transition at the target. No transition, or a marginal/noisy transition, is the fault.
- Ohm the sensor cable end-to-end and check shield continuity and grounding at one end only. A shield grounded at both ends can inject enough noise into a low-level feedback line to destabilize the loop.
Step 4 — Electrical baseline
Do this before condemning any card. A marginal supply will mimic a dozen different logic faults.
| Check | Method | Accept criterion |
|---|---|---|
| Incoming line voltage | Measure phase-to-phase at the disconnect, machine idle and during the spin-up attempt | Within the machine nameplate tolerance; no collapse under load; balanced between phases |
| Control transformer secondary | Measure at the transformer output terminals | Matches the tap/nameplate value; stable during the fault |
| PSU logic rails | Measure each DC rail at the supply and again at the far end of the backplane | Within the supply's stated tolerance, low ripple, no sag when the spindle command is issued |
| Grounding | Verify machine ground and drive/motor ground bonds are tight and low-impedance | Clean bonds; motor frame ground intact after the motor swap |
Step 5 — Cards and connections
- Power down, lock out, and let the drive bus discharge.
- Unseat and reseat every card in the control rack, including the spindle controller card. Work the card fully out and back in to wipe the edge contacts.
- Inspect edge fingers and connector pins for oxidation, burning, or bent pins. Clean oxidized contacts with an appropriate contact cleaner; do not use abrasives on gold-plated fingers.
- Reseat all ribbon and terminal connections on the spindle drive, particularly the encoder feedback and the analog speed command / direction lines.
- Confirm the new encoder's wiring landed on the correct terminals with correct channel polarity. Swapped A/B or inverted channels invert the feedback sense and produce runaway or oscillation instead of controlled rotation.
Step 6 — Drive verification
- Check the drive's backup battery. If the drive retains parameters in battery-backed memory, a dead battery loses the configuration on every power cycle.
- Compare the drive's parameter set against the documented values for this machine/motor combination. A rebuilt drive may be returned with generic or bench-test settings.
- Verify the drive's jumper/DIP configuration matches the original unit — command source, feedback type, and direction logic.
- Read any fault or status indicators on the drive during the hunting event. If the drive itself is not faulting, it is likely faithfully executing an oscillating command from upstream.
- If the drive is receiving an analog speed reference, meter that reference during the fault. A command that swings positive/negative in step with the spindle motion proves the controller card is the source; a steady command with oscillating motion proves the drive/feedback loop is the source.
Isolating Command Source vs. Feedback Loop
The single measurement that splits the problem in half is the speed/direction command going into the drive:
| Command signal at drive input | Spindle behavior | Conclusion |
|---|---|---|
| Oscillates in sync with spindle | Hunting | Controller card / orientation logic is commanding the oscillation. The drive is innocent. |
| Steady, non-zero | Hunting | Drive or feedback loop is unstable. Suspect encoder wiring/polarity, drive tuning, or drive hardware. |
| Zero or absent | Hunting | Drive is running open on a corrupted reference or the wrong command source is selected in the drive configuration. |
Given that drive, motor, and encoder have already been replaced without change, the highest-probability outcome of this test is the first row — the spindle controller card, or the orientation feedback it depends on, is generating the oscillating command.
Repair and Verification
- Correct the identified fault: realign or replace the orientation sensor, restore air supply and detent operation, reseat or replace the spindle controller card, or reload/repair the drive configuration.
- Re-run
M19alone. The spindle must rotate in one direction, decelerate to the orient position, lock, and hold with no hunting. - Command
M3at a low speed. Confirm smooth acceleration to commanded RPM with no reversal. - Ramp the commanded speed in steps across the usable range. Watch for oscillation appearing only at specific speeds — that pattern points to a tuning or resonance issue rather than a feedback dropout.
- Command
M4and confirm correct reverse rotation and correct RPM. Wrong direction of actual rotation versus command indicates a motor phase or encoder polarity swap from the motor replacement. - Run a full toolchange cycle. The toolchange depends on successful orientation, so a clean change validates the orientation loop end to end.
- Cycle machine power and repeat the M19 and M3 tests. If the fault returns only after a power cycle, suspect the drive's parameter retention/battery.
FAQ
Why does my Fadal spindle rock forward and backward instead of spinning up?
That signature indicates a position loop hunting rather than a velocity loop failing. The control is attempting spindle orientation and never receives the in-position confirmation, so it drives back and forth around the target. Check the orientation sensor above the main pulley and the air detent cylinder first.
How do I test whether the problem is orientation-related?
Power up and issue M19 in MDI instead of M3 or M4. If the same forward/reverse hunting appears on M19 alone, the orientation subsystem is the fault. If M19 completes and locks cleanly, the problem is in the speed command path instead.
I replaced the drive, encoder, and motor and the fault is still there. What is left?
The spindle controller card that generates the command, the orientation sensor and its alignment, the air detent solenoid and cylinder, the incoming line and PSU rails, and the card/connector seating. Measure the speed command at the drive input during the fault to determine whether the oscillation originates upstream of the drive.
Can low shop air cause the spindle to oscillate?
Yes. If the detent/stop cylinder does not engage and lock before the spindle sequence proceeds, the control never sees the lock confirmation and the orientation loop keeps hunting. Verify regulator pressure and manually override the orientation solenoid to confirm the cylinder strokes fully.
Should I suspect the rebuilt drive's backup battery?
Check it. If the drive stores its parameters in battery-backed memory, a depleted battery loses the configuration at every power cycle and the drive runs on defaults or corrupted values. A fault that returns only after power cycling strongly suggests parameter retention loss.