The Roberts Linkage is a symmetric four-bar mechanism that traces an approximate straight line along a coupler point fixed to the centre of its connecting rod. Over its useful travel it holds the path within roughly 0.1% of stroke length deviation from a true line. It exists to generate straight-line motion without prismatic guides, eliminating slide friction and contamination paths. You will find the Roberts straight-line linkage in lower-limb prosthetic knees, optical scanners, and pen-plotter carriages where rail wear would corrupt repeatability.
Roberts Linkage Interactive Calculator
Vary stroke, side-link length, geometry factor, and allowable error to see Roberts linkage straight-line deviation and fit margin.
Equation Used
This calculator estimates the worst-case departure of the Roberts coupler point from its intended straight line. For the standard symmetric geometry, use k = 0.0025; increasing side-link length reduces deviation with the square of the stroke-to-link ratio.
- Symmetric Roberts linkage with equal side rockers.
- Standard proportions use base length = 2 * L_side.
- Apex point lies on the coupler perpendicular bisector.
- k = 0.0025 represents the standard Roberts geometry over useful travel.
The Roberts Linkage in Action
The Roberts Linkage, also called the Roberts straight-line linkage, is a four-bar mechanism with two equal-length rocker arms and a triangular coupler whose tip rides along an approximate straight line. The two ground pivots sit at the ends of a fixed base. The two side links are equal length, and the coupler triangle is isoceles with a tip point that hangs below the connecting rod by a defined distance. When you swing one rocker, the geometry forces the coupler tip to follow a path that stays within about 0.1% of stroke deviation from a true line over roughly the middle 60% of travel. Push past that range and the path curls upward at both ends — that's the rocker arms running out of useful angle.
Why build it this way? Because a prismatic slide costs you friction, dust ingress, and a wear surface that drifts over time. A Roberts straight-line linkage replaces all of that with four pin joints. Pin joints can run on sealed bearings or jewel pivots, last decades, and don't care about grit. The trade is geometric — the line is approximate, not exact. For most applications the approximation is irrelevant; for fringe-counting interferometry, it isn't.
Tolerances matter. If your two side links differ in length by even 0.2% on a 100 mm arm, the coupler tip path goes asymmetric and you'll see a sideways drift of around 0.05 mm at mid-stroke. The coupler triangle apex must sit on the perpendicular bisector of the connecting rod — if it's offset by 0.5 mm, the line tilts. Common failure modes are pin-joint slop opening up after 10⁶ cycles, the coupler triangle flexing under load if it's machined too thin, and ground-pivot mounting plates deflecting and shifting the base length. Any one of those turns a 0.1% straight-line path into a 1% banana.
Key Components
- Ground Link (Base): The fixed reference frame between the two ground pivots. Its length sets the scale of the mechanism — typical Roberts builds use a base length equal to 2× the side-link length. Tolerance on base length should be ±0.05 mm per 100 mm or the symmetry of the coupler curve breaks down.
- Side Links (Rockers): Two equal-length arms pinned to the ground at one end and to the coupler at the other. They must match within 0.1% of length — so on a 100 mm arm, no more than 0.1 mm difference. They typically swing through ±30° about a vertical centreline.
- Coupler (Triangular Connecting Link): An isoceles triangle linking the tips of the two rockers. The apex of the triangle is the coupler point that traces the straight line. The apex must sit exactly on the perpendicular bisector of the rod connecting the two rocker tips. Apex offset distance is normally equal to the side-link length.
- Pin Joints: Four revolute joints — two ground pivots, two coupler-to-rocker joints. Use bushings or sealed ball bearings. Radial play above 0.05 mm at any joint shows up directly in the output path as positional noise.
- Coupler Point: The tip of the coupler triangle that traces the straight-line segment. Output is taken from this point — you attach the load, pen, or sensor here. The straight-line region is roughly 1.4× the side-link length in span.
Who Uses the Roberts Linkage
The Roberts Linkage shows up wherever a designer needs straight-line travel without the wear, friction, or contamination penalty of a prismatic slide. It's quieter than a slide, runs without lubrication when fitted with bronze bushings, and survives hostile environments. It's not as accurate as a Peaucellier-Lipkin exact straight-line linkage, but it uses 4 bars instead of 8 — half the parts, half the joints, half the slop budget.
- Prosthetics: The Otto Bock 3R60 polycentric knee uses a four-bar linkage of the Roberts family to produce stance-phase stability and an approximate straight-line shank trajectory during the mid-swing phase, letting amputees walk with a natural gait pattern.
- Pen Plotters and Drafting: Vintage HP 7475A-style pen carriages and earlier mechanical chart recorders used Roberts-type linkages to drive the pen tip along a line without prismatic rails, keeping ink contamination out of bearing surfaces.
- Vehicle Suspension: Some classic British rear-axle locating linkages (notably on early Rover P6 prototypes) used Roberts-style four-bar geometries to keep the axle moving vertically without lateral drift, eliminating Panhard rod requirements in low-travel applications.
- Textile Machinery: Picanol rapier loom weft-insertion guides have historically used Roberts-style linkages to drive guide forks along an approximate straight line across the warp without sliding contact that would attract lint.
- Optical Scanners: Drum-type document scanners built by companies like Howtek used Roberts straight-line linkages to translate sensor heads parallel to the drum axis, where rail wear over millions of scans would have introduced banding artefacts.
- Walking Toys and Robotics: Theo Jansen-style classroom kinematic kits and small educational walking robots use Roberts and related approximate straight-line linkages to give the foot a flat-bottom trajectory during ground contact.
The Formula Behind the Roberts Linkage
The key design number for a Roberts Linkage is the deviation from a true straight line — how far the coupler point wanders off the ideal line over the useful travel. This deviation scales with stroke. At short strokes (say 20% of side-link length) the path looks dead straight to the eye. At nominal stroke (60-70% of side-link length) you hit the design sweet spot, where deviation peaks at about 0.1% of stroke. Push beyond 80% and deviation grows fast — 1% at 90% stroke, 5%+ above that as the rockers approach their angular limits.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Δmax | Maximum deviation of the coupler point from a true straight line over the useful stroke | mm | in |
| Lstroke | Length of straight-line travel measured along the nominal line | mm | in |
| Lside | Length of each side rocker link (both must be equal) | mm | in |
| k | Geometric constant for the symmetric Roberts configuration, approximately 0.0025 for the standard proportions (base length = 2 × L<sub>side</sub>, coupler apex offset = L<sub>side</sub>) | dimensionless | dimensionless |
Worked Example: Roberts Linkage in a precision film-advance mechanism
A 16 mm archival film transport workshop in Wellington is rebuilding the intermittent claw guide on a 1960s Bell & Howell film inspection machine. They want to replace a worn prismatic slide with a Roberts straight-line linkage that drives the claw tip along a 24 mm vertical path, and they need to know the worst-case lateral deviation across the stroke so they can verify the claw still engages film perforations to within ±0.05 mm.
Given
- Lside = 60 mm
- Lstroke,nom = 24 mm
- k = 0.0025 dimensionless
Solution
Step 1 — at the nominal 24 mm stroke (which is 40% of the 60 mm side link), compute the stroke ratio:
Step 2 — apply the deviation formula at nominal stroke:
That's about 10 µm of lateral wander across the full stroke — comfortably inside the ±50 µm perforation engagement budget. The claw will track perforations cleanly.
Step 3 — at the low end of the typical operating range, a 12 mm half-stroke (20% of Lside):
At this stroke the path is straight to within 1.2 µm — effectively a true line by any optical measurement you can do on the bench. You'd never see this deviation on a dial indicator.
Step 4 — at the high end, push the stroke to 48 mm (80% of Lside):
Now you're at 77 µm of wander — over the perforation tolerance budget. The claw would start clipping perf edges and tearing film at this stroke. The geometry has run out of useful range, and the rockers are approaching the dead-band where the curve folds back on itself.
Result
Maximum lateral deviation at the 24 mm nominal stroke is 0. 0096 mm — about 10 µm. That's well inside the perforation engagement tolerance and the claw will index film without scoring or tearing. Comparing across the operating range: at 12 mm stroke the path is straight to 1.2 µm (invisible to standard inspection), at 24 mm nominal it's 10 µm (clean engagement), and at 48 mm the deviation jumps to 77 µm and the mechanism stops being usable for film perforations. If a built mechanism measures higher deviation than predicted, look first at side-link length mismatch — even 0.1 mm difference between the two 60 mm rockers introduces a constant offset of around 25 µm. Second, check the coupler apex position against the perpendicular bisector of the upper rod; offset here tilts the line and shifts measured deviation by 30-50 µm depending on stroke direction. Third, pin-joint radial play above 0.03 mm at any of the four joints shows up as random positional noise that swamps the geometric deviation entirely.
Roberts Linkage vs Alternatives
The Roberts straight-line linkage is one of several approximate straight-line four-bar mechanisms. The two main alternatives are the Chebyshev linkage and the Hoeken linkage. All three use four bars; the differences lie in proportion, useful stroke length, and where the straight-line region sits in the cycle. Pick based on stroke ratio, output speed uniformity, and how much of the cycle you need straight-line behaviour for.
| Property | Roberts Linkage | Chebyshev Linkage | Hoeken Linkage |
|---|---|---|---|
| Straight-line accuracy (deviation as % of stroke) | ~0.1% at nominal stroke | ~0.25% at nominal stroke | ~0.05% at nominal stroke |
| Useful straight-line stroke (as % of side link) | ~60-70% | ~50% | ~50% with near-uniform velocity |
| Velocity uniformity along straight section | Variable, peaks mid-stroke | Variable | Near-constant velocity along straight portion |
| Mechanical complexity | 4 bars, 4 pin joints | 4 bars, 4 pin joints | 4 bars, 4 pin joints |
| Typical lifespan with sealed bearings | 10⁷-10⁸ cycles | 10⁷-10⁸ cycles | 10⁷-10⁸ cycles |
| Best application fit | Compact straight-line travel, prosthetics, scanners | Walking robots, classroom kits | Pick-and-place, constant-velocity drawing |
| Sensitivity to link-length error | High — 0.1% mismatch shifts path | Moderate | High |
Frequently Asked Questions About Roberts Linkage
Keep the two side rockers matched to within 0.1% of their nominal length. On a 60 mm rocker that's 0.06 mm. Above this, the coupler point develops a constant lateral offset that grows with stroke — at 0.5% mismatch on a 60 mm arm you'll see roughly 50 µm of asymmetric drift, which is enough to ruin precision work.
The cause is geometric: the symmetric Roberts configuration relies on the two rockers reaching mirror-image angles at every instant. Unequal lengths break the mirror, and the coupler apex traces a curved arc instead of a line. Cut both rockers from one bar in a single setup, then dowel-pin them to a common fixture before final reaming of the pin holes.
That's the geometry doing exactly what it's supposed to do. The Roberts straight-line approximation is only valid over roughly the middle 60-70% of the rocker swing. Outside that band, the coupler point lifts away from the line as the rockers approach their angular limits — the path curls upward symmetrically at both ends.
If you need straight-line behaviour all the way to the stroke ends, redesign with longer side links so the useful portion of the curve covers your required stroke. Rule of thumb: side-link length should be at least 1.5× the required straight-line stroke.
Pick the Hoeken. The Roberts straight-line linkage produces a straight path but the coupler point moves slowly at the path ends and quickly through the middle — velocity is sinusoidal-ish along the line. The Hoeken is specifically tuned to give near-constant velocity along its straight section, which is what you want when placing parts at uniform spacing.
Use the Roberts when you only care that the path is straight, not that the speed along it is uniform. Prosthetic knees, optical scan stages, and dispenser nozzles where dwell time is set by a separate timing mechanism all favour Roberts.
The coupler triangle is almost certainly flexing under inertial load. The triangle is loaded in bending at the apex — when the linkage runs at speed, the angular acceleration of the coupler whips the apex point through forces that scale with rotational acceleration squared. A coupler machined from 3 mm aluminium plate that works fine at 1 Hz can ring like a tuning fork at 10 Hz.
Diagnostic check: tap the coupler with a pen at rest and listen. If you hear a clear ring, it's underdamped and will vibrate in service. Fix by increasing thickness, adding a triangulating gusset, or switching to a stiffer material like 7075-T6 or carbon fibre laminate. Stiffness scales with the cube of thickness, so a small thickness increase pays off enormously.
It's never exact. The Roberts Linkage produces a path that's geometrically approximate — the deviation can be made small (sub-micron over short strokes) but never zero. If you need a true exact straight line, use a Peaucellier-Lipkin linkage or a Hart's A-frame inversor. Both produce mathematically exact straight-line motion at the cost of more bars and more joints.
For 99% of mechanical applications the Roberts approximation is fine. The exceptions are interferometry, lithography stages, and any application where deviation must be smaller than the wavelength of light. There, the extra mechanism complexity of an exact linkage is worth it.
Almost always pin-joint wear. The Roberts straight-line linkage is geometrically sensitive to radial play at any of its four pin joints. A bushing wearing from 0.02 mm to 0.08 mm radial clearance triples the positional noise at the coupler point, and unlike slide wear it shows up as random scatter rather than systematic drift.
Diagnostic: with the mechanism unloaded, grip the coupler apex and try to wiggle it laterally. Anything more than 0.05 mm of free play means at least one joint is past its useful life. Replace bushings as a matched set — replacing one and leaving the others worn just shifts the noise source. For high-cycle builds, specify sealed needle roller bearings at the ground pivots from the start.
References & Further Reading
- Wikipedia contributors. Straight-line mechanism. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
