When a machine misses tolerance, the slide table is usually blamed before the tool, the fixture or the temperature. Sometimes the slide is at fault and sometimes it is not, but the argument only ends when you can name the error source and its magnitude. There are five, and they behave differently.
1. Screw lead error — fixed, predictable, compensable
A ball screw is ground (or rolled) to a tolerance class. C7 rolled screws, common in budget automation, allow ±0.05 mm per 300 mm. C5 ground screws allow ±0.023 mm per 300 mm. C3 allows ±0.008 mm but costs several times more and needs a temperature-controlled environment to be worth it.
Lead error is deterministic: it repeats every revolution. That makes it compensable. A control with screw mapping, or a linear scale, removes most of it. What cannot be removed is the periodic error within one revolution, which on a C5 screw is typically 5–8 µm.
| Grade | Lead error / 300 mm | Typical use |
|---|---|---|
| C7 rolled | ±0.05 mm | Transfer axes, general automation |
| C5 ground | ±0.023 mm | Machine tool axes, precision positioning |
| C3 ground | ±0.008 mm | Metrology, semiconductor, high-end grinding |
2. Rail parallelism — machined in, not adjusted out
Two rails must be parallel in both planes, and the carriage must not be over-constrained. Our specification is 0.005 mm per 300 mm in the horizontal plane and 0.01 mm per 300 mm vertically. If the rails are machined in separate set-ups, that figure is met by shimming, which holds for a year and then relaxes.
This is the single strongest argument for buying a slide table as an assembly rather than building one from rails, screw and plate. Assembly alignment depends on the skill of whoever bolts it together; a machined-in datum does not.
3. Backlash — the reversal error
Backlash is lost motion when the axis reverses. On a ball screw it comes from the nut, not the screw, and it is removed by preload. A single nut with a spring preload removes 0.02–0.03 mm. A double nut with mechanical preload holds lost motion at the level quoted for the model.
Preload is not free: it adds drag torque and generates heat. Too much preload is a common cause of a screw that runs hot and a nut that wears out in eighteen months. The correct preload is the smallest one that removes backlash at your maximum axial load.
Measure it, do not assume it
Backlash is measured with a dial indicator on the table and a commanded reversal of 0.1 mm. Repeat five times in each direction. A single reading understates the problem because preload settles.
- Indicator resolution
- 0.001 mm
- Reversal step
- 0.1 mm
- Accept
- at or below the quoted figure
- Reject
- > 0.01 mm, re-preload nut
4. Base stiffness — where aluminium runs out
A slide table base deflects under load, and the deflection is not uniform: it is largest at the centre of travel when the carriage is furthest from a mounting foot. On an aluminium frame the centre of travel is the soft spot, and the deflection grows faster than the load and faster than the span — which is why a cast iron column range sits alongside the aluminium frames.
Stiffness also governs vibration. A cast iron frame damps far better than an aluminium plate, and on a cutting axis that shows up in surface finish before it shows up on any positioning report. If the part looks worse than the tolerance says it should, suspect stiffness rather than accuracy.
5. Thermal growth — the slow error
Steel expands about 11 µm per metre per °C; aluminium about 23 µm. A 1 m aluminium base that warms 5 °C during a shift moves roughly 0.1 mm at its free end. This is why an axis that is accurate at 08:00 is out of tolerance at 14:00, and why the error is invisible on a short test cycle.
There are three defences. Reduce heat input by sizing the motor for the duty rather than the peak. Reduce sensitivity by using cast iron or by fixing only one end of the base. Or measure the real position with a scale and let the control compensate. The third is the only one that works when the environment is out of your control.
Where to spend money first
For most machine builders the order of return is: preload first (cheap, removes reversal error), rail and datum machining second (built into the assembly), screw grade third, scale fourth, cast iron frame fifth. Buying a C3 screw on an aluminium frame with a spring-preloaded nut is spending money in the wrong order.
State the tolerance you need, the stroke over which you need it, and the temperature range, and we will tell you which combination reaches it — including when a standard model is enough. Every assembly ships with a measured inspection report so the first argument on the shop floor is settled by data.
Common questions
Only with a temperature-controlled environment and a scale, or when the process genuinely needs sub-10 µm positioning over a long stroke. Otherwise C5 with good preload gives more practical benefit.