Multi-axis motion stages · accuracy versus repeatability, unidirectional versus bidirectional figures, ISO 230-2 test points, screw lead, speed and resolution
An XYZ positioning stage stacks linear axes, such as ball-screw or linear-motor driven slides on linear guides with servo or stepper motors, to move a tool or workpiece to programmed points in three directions, for dispensing, placement, inspection and assembly. Ask three things first: how the maker measured accuracy and repeatability (where, how many points, which direction), what speed and resolution the drive gives with your load, and how mounting and heat will change the result in your machine.
Multi-axis XY tables and XYZ stages built from ball-screw, belt or linear-motor modules with motors and controllers for automation, dispensing, placement and inspection. Single-axis linear guides and ball screws as components, piezo nanopositioners and full gantry robots for heavy loads are outside this page except where compared.
01
Novanta defines accuracy as the difference between the position commanded by the controller and the load position measured by an independent instrument such as a laser interferometer, and notes that accuracy varies along the travel because of machining, mounting and encoder errors. Averaging scattered measurements, as with RMS or 3-sigma methods, can make a datasheet figure look better than the stage performs; Novanta follows ISO 230-2:2006, which requires 20 target points plus the home position over the full travel. [1]
Repeatability is the ability to return to a commanded position. Novanta says unidirectional repeatability figures are better than bidirectional ones, that buyers should ask which is quoted, and that “typical” values may come from controlled tests rather than ordinary installation. [1] Dover Motion likewise notes that bidirectional variance is higher and that quoting unidirectional repeatability alone can mask large backlash; it measures with a laser interferometer over repeated approaches from both directions. [2]
Editorial note: for a stacked XYZ stage, ask for accuracy and bidirectional repeatability of each axis and, if the process depends on it, of the combined tool point, measured on stages taken from stock and mounted as you will mount them.
Give the supplier the travel on each axis, load and its centre of gravity, required accuracy and bidirectional repeatability, speed and acceleration, cycle (points per minute), mounting surface and orientation, environment (temperature, dust), controller interface, cable management and safety requirements.
02
Dover Motion explains that leadscrew-driven stages generate frictional heating, causing thermal expansion and positional drift that depends on speed and duty cycle, and that where very high repeatability is needed a linear motor with a linear encoder in closed loop removes leadscrew effects, limited then by encoder resolution and thermal effects. From laser interferometer testing it reports its anti-backlash leadscrew stages at below 0.5 µm unidirectional and below 1 µm bidirectional repeatability (typ.). [2]
Editorial calculation: a ball screw with 10 mm lead turned at 3,000 rpm moves the carriage 10 × 3,000 / 60 = 500 mm/s; with a motor encoder of 10,000 counts per revolution, one count is 10 mm / 10,000 = 1 µm of travel, before screw lead error, backlash and thermal growth. Halving the lead to 5 mm doubles resolution and thrust per torque but halves top speed. Lead, speed and encoder counts are assumed examples; use the stage maker's figures and your load.
03
ALIO Industries' specification for its AI-XBS-10000-UII 100 mm ball-screw linear stage shows the set of figures a buyer can ask any supplier for. It is driven by a frameless torque motor and precision ball screw with an anti-backlash nut, with a non-contact optical linear encoder and dual recirculating ball rail bearings. Footnotes on the sheet tie performance values to stated conditions. [3]
The sheet lists accuracy, repeatability, straightness, flatness and angular errors separately, and gives the bidirectional (not only unidirectional) repeatability, which is what the measurement guidance above asks buyers to check. [3] [1] [2]
Editorial calculation: ±12 arc-sec of pitch over a 100 mm offset from the stage (Abbe arm) is 100 mm × 12 / 206,265 ≈ 5.8 µm of potential error at the tool point, larger than the ±2 µm accuracy of the upgraded grade.
| Specification (ALIO AI-XBS-10000-UII) | Standard | Ultra |
|---|---|---|
| Travel | 100 mm (±50 mm) | 100 mm |
| Linear displacement accuracy | ±5 µm | ±2 µm |
| Bidirectional repeatability | ±1.0 µm | ±1.0 µm |
| Backlash | 0 nm | 0 nm |
| Resolution | About 4.88 nm after 4096 interpolation (1–5000 nm digital optional) | Same |
| Straightness in Y / flatness | ±3 µm / ±3 µm | ±2 µm / ±2 µm |
| Pitch, yaw, roll | ±12 arc-sec each | ±12 arc-sec each |
| Maximum velocity / peak acceleration | 0.3 m/s / 0.5 g | Same |
| Payload horizontal / vertical | 150 kg / 80 kg | Same |
| Operating temperature / humidity | 0–50 °C / 10–80% non-condensing | Same |
Ask each supplier for the same lines: travel, accuracy and repeatability (bidirectional, with test method), backlash, resolution, straightness, flatness, pitch/yaw/roll, speed, acceleration, payload by orientation and environment.
04
Zaber Technologies lists its XY and XYZ systems as families built from its own linear stages, each with a maximum travel, a payload for the XY configuration, a lower payload for the XYZ configuration and a maximum speed. Families range from screw-driven compact and steel-bodied stages to direct-drive linear-motor stages, with linear encoder or dust cover options on some. [4]
Editorial calculation: the XYZ payload is 20% of the XY payload for the LSQ and LRQ families (10 kg of 50 kg), 5% for the LRM (2.5 kg of 50 kg) and 75% for the LRT (120 kg of 160 kg), so a vertical axis can set the load limit of the whole stage. The LDA and LDM linear-motor families list Z travel of 25 mm and 60 mm against 150 mm and 210 mm in X and Y. [4]
Give the supplier the payload mass and centre of gravity on the top axis, whether the Z axis lifts it, the travel on each axis and the speed needed, and ask which payload figure (XY or XYZ) its quotation is based on. [4]
| Zaber family | Max travel X × Y (× Z) (mm) | XY payload up to (kg) | XYZ payload up to (kg) | Speed up to (mm/s) |
|---|---|---|---|---|
| LSM (compact) | 200 × 200 (× 200) | 10 | 2.5 | 104 [4] |
| LSQ | 600 × 600 (× 600) | 50 | 10 | 1,000 [4] |
| LRM (steel) | 200 × 200 (× 200) | 50 | 2.5 | 100 [4] |
| LRQ | 600 × 600 (× 600) | 50 | 10 | 840 [4] |
| LRT (high thrust) | 1,000 × 1,000 (× 1,000) | 160 | 120 | 700 [4] |
| DMA (linear motor) | 10 × 10 (× 10) | 4.5 | 1 | 365 [4] |
| LDA (linear motor) | 150 × 150 (× 25) | 9.1 | 1.6 | 800 [4] |
| LDM (linear motor) | 210 × 210 (× 60) | 14.7 | 5.5 | 1,200 [4] |
05
The listing reviewed covers XY tables and XYZ stages built from ball-screw and belt-driven linear modules, with linear motor modules and motion controllers. Ask for each axis's accuracy and bidirectional repeatability, test method and load rating.
06
Drift starts in the drive row: Dover Motion attributes it to frictional heating of the leadscrew and removes it with a linear motor and linear encoder in closed loop. [2]
| Part | Role | Fix in the order |
|---|---|---|
| Linear guides | Carry and guide each axis. | Rail size, preload, straightness. |
| Ball screw or linear motor | Drives the carriage; screw heating causes drift. [2] | Lead, accuracy grade, or motor force. |
| Motors and encoders | Set speed and resolution. [2] | Motor type, encoder counts. |
| Linear encoder (option) | Closed-loop position on the load. [2] | Resolution and accuracy. |
| Controller | Commands motion and coordinates axes. | Interface and error compensation. |
| Cable carriers | Route cables through moving axes. | Size and life. |
07
Ask for accuracy with its method (ISO 230-2, 20 points plus home) and repeatability as a bidirectional figure; Z travel, load capacity, motor type, humidity, IP rating, supply and weight have no published value in this table. [1]
| Field | What it means and published values | Ask the supplier for |
|---|---|---|
| Travel Range (X/Y) (mm) | Accuracy varies along travel. [1] | Travel and test points. |
| Travel Range (Z) (mm) | Vertical axis travel. | Travel and brake. |
| Positioning Accuracy (mm) | Commanded vs interferometer-measured position; ISO 230-2. [1] | Method and maximum error. |
| Repeatability (mm) | Ask uni- or bidirectional; e.g. <1 µm bidirectional (leadscrew stage). [1] [2] | Bidirectional value. |
| Maximum Speed (mm/s) | Lead × rpm (editorial calculation above). | Speed at your load. |
| Maximum Load Capacity (kg) | Load and moment limits. | Load by orientation. |
| Drive System | Ball screw, belt or linear motor. [2] | Type and accuracy grade. |
| Motor Type | Servo or stepper. | Motor and encoder. |
| Operating Temperature Range (°C) | Thermal growth affects position. [2] | Range and compensation. |
| Relative Humidity (Non-condensing) (%RH) | Environment. | Range. |
| Ingress Protection Rating | Dust and fluid protection. | Covers or bellows. |
| Supply Voltage (V AC) | Drives and controller. | Supply. |
| Weight (kg) | Machine base design. | Weight. |
| Angular errors and straightness | ±12 arc-sec pitch/yaw/roll; ±2–3 µm straightness and flatness on one 100 mm stage. [3] | Angular and straightness specs per axis. |
08
Paste it into your enquiry below or into an email to a maker, then replace every bracket.
Travel: [X; Y; Z] Load: [mass; centre of gravity; orientation] Accuracy: [per axis; at tool point] Repeatability: [bidirectional] Speed: [speed; acceleration; points per minute] Environment: [temperature; dust] Control: [controller; interface] Test data: [method; number of points]
1 company lists this product among what they make. Company figures are quoted from each company's own website; every card states where the relationship came from.
Lists XY tables and XYZ positioning stages built from ball-screw and belt linear modules, with linear motor modules and controllers.
Product scope checked 2026-10-11
Explore company information and manufacturing capabilities, then discuss the scope of your project with a relevant supplier.
Novanta defines accuracy as commanded versus independently measured position, and repeatability as the ability to return to a commanded position. [1]
Dover Motion says unidirectional figures alone can mask large backlash; Novanta says unidirectional figures are the better ones. [2] [1]
ISO 230-2, as applied by Novanta, requires 20 target points plus home over the full travel. [1]
ALIO's 100 mm stage sheet, for example, lists bidirectional repeatability (±1.0 µm), backlash, resolution, straightness and flatness (±2–3 µm), pitch, yaw and roll (±12 arc-sec), speed, acceleration and payload by orientation. [3]
In Zaber's published families it does: LSQ and LRQ systems list up to 50 kg as XY stages and up to 10 kg as XYZ stages; the LRT family lists 160 kg and 120 kg. [4]
Sources checked 2026-10-12. Novanta, Dover Motion, ALIO Industries and Zaber Technologies are motion-stage makers; their notes and product pages explain how they specify and test their own stages. ISO 230-2 is cited as Novanta applies it.
Ask for use case, specification boundaries, supplier type, and RFQ preparation information for this product.
Compare manufacturer profiles with relevant product and process capability.