Linear guide noise: five causes and what each one tells you

2026-11-06

# Linear guide noise: five causes and what each one tells you

A customer writes that the guide is "noisy". That single word covers five different faults, and the fix for each is different. The good news: the sound tells you which one it is, if you know what to listen for.

1. Constant hiss or rush, proportional to speed

Cause: ball recirculation. Every ball passes through the end cap and turns 180°. At speed this is audible and it is normal. It rises with velocity, not with load.

Test: run the block slowly by hand. If it goes quiet, it is recirculation, not damage.

Fix: none needed — unless the customer wanted silence, in which case the answer is a different model, not a repair. On our range that is the Q1 type, where a synchronising cage keeps the balls apart. Order the model as QHW — writing "quiet version" in the notes on an HGH order does not produce a Q1 block.

2. Click or knock once per pass

Cause: a damaged ball or a dent in the raceway. One defective ball makes one click per circulation.

Test: the click is periodic and repeats at the same position. Slide the block slowly and mark the rail where it clicks.

Fix: replace the block. A dented raceway is not repairable. If the dent came from debris, the wiper is not doing its job either — check it.

3. Grinding or roughness that gets worse with load

Cause: insufficient or wrong lubrication. Metal-to-metal contact under load.

Test: clean and relubricate. If the noise drops noticeably within a few strokes, it was lubrication.

Fix: establish a relubrication interval based on travel distance, not calendar time. A block that travels 20 km a month needs attention far more often than one that moves once a week.

4. Intermittent binding, worse at one end of the stroke

Cause: misalignment between the two rails, or rail height difference. The block is being forced through a narrowing gap.

Test: loosen the mounting bolts on one rail and see whether the noise changes. If it disappears, it is alignment, not the guide.

Fix: align to the specified parallelism and height tolerance, then torque in the documented sequence. This is the most common cause on machines assembled in the field.

5. Squeal or chatter only during cutting

Cause: preload too low for the cutting force, so the block lifts and re-seats. It is not a noise problem, it is a rigidity problem.

Test: it appears only under load and disappears in air cuts.

Fix: this is where the instinct is to raise preload. Check the size first. Going from ZA to ZB on the same size adds a little rigidity and costs life. Going up one size adds roughly half again the load rating. Raise the size before raising the preload.

What to send us when you report a noise

Four things make it diagnosable without shipping the part back:

1. The model, and the preload written as ZA (HG, 0.05–0.07C)
2. When it happens — always, at speed, only under load, only at one end
3. How many hours or kilometres it has run
4. Whether the other axis on the same machine sounds the same

With those four we can usually tell you which of the five it is before you take anything apart.