The Ross yoke is a slotted-link mechanism that converts crankshaft rotation into nearly sinusoidal motion of two opposed pistons using a single triangular yoke pivoting on the crankpin. American engineer Andy Ross patented the layout in the 1980s for compact alpha-configuration Stirling engines. The yoke replaces conventional connecting rods, keeping piston rod motion almost perfectly axial and slashing side-loads on the cylinder walls. That low side-load is what lets a Ross-yoke Stirling run with simple piston seals at 0.5–3 kW shaft outputs without the bulk of a rhombic drive.
Ross Yoke Mechanism Interactive Calculator
Vary stroke, piston phase offset, and residual rod deviation to see the Ross-yoke crank radius, motion amplitude, phase error, and side-deviation ratio.
Equation Used
The Ross yoke is treated as an ideal sinusoidal drive: a stroke S requires a crank radius r = S/2, and the two opposed pistons are separated by phase angle phi. The deviation ratio compares the residual lateral rod motion d with the full stroke.
- Ideal Ross-yoke piston motion is modeled as sinusoidal.
- Stroke is peak-to-peak piston travel.
- Residual lateral deviation is small and handled by the guide bushing.
- Design phase target for alpha Stirling pistons is 90 deg.
Inside the Ross Yoke Mechanism
Picture a flat triangular plate — the yoke — pinned at its centre to the crankpin of an offset crankshaft. Two of the triangle's corners carry pivot pins that connect to the lower ends of two vertical piston rods, one feeding each opposed cylinder. The third corner rides on a fixed guide pin or pivots on a small grounding link. As the crankshaft turns, the yoke wobbles around the crankpin, and the two piston-rod pivots trace nearly straight vertical lines. That straightness is the entire point of the Ross yoke linkage — it removes the lateral component a normal connecting rod would impose on the piston, so the piston rod stays essentially axial through the stroke.
Why bother? In an alpha Stirling engine the working gas is shuttled between a hot piston and a cold piston 90° out of phase. Any side-load on those pistons drags the seals across the bore and burns the engine's already-marginal efficiency. The Ross yoke geometry holds piston-rod deviation from pure vertical to typically under 0.3 mm over a 30 mm stroke — small enough that a guide bushing handles the residual, and the main piston seal sees almost no transverse force. You get crank-driven simplicity with the side-load behaviour of a crosshead engine.
Get the geometry wrong and the benefit collapses fast. If the yoke arm lengths are mismatched by more than about 0.5%, the two pistons no longer execute the same stroke and the phase angle between them drifts from the design 90°, which detunes the Stirling cycle and drops indicated power 10–20%. If the grounding pivot wears or the slot in a slotted-yoke variant develops play, the piston rod starts to swing and the seal scrubs the bore — you'll see asymmetric heat marks on the rod and a steady oil-film breakdown above 1500 RPM. Bearing selection at the crankpin matters too: the yoke pivot sees full combustion or expansion load, so undersizing the needle bearing there is the most common Ross-yoke failure.
Key Components
- Triangular Yoke Plate: The single rigid link that converts crank rotation into near-linear motion of two piston rods. Typically a forged or machined steel plate 8–15 mm thick with three precisely located pivot bores, each held to ±0.02 mm true-position relative to the others.
- Crankpin and Needle Bearing: Centre pivot of the yoke, mounted on the offset crankshaft. Sees the full peak gas load — sized for at least 2× the maximum cylinder force, with a typical needle roller bearing such as an INA HK-series at the 20–25 mm bore.
- Piston Rod Pivots: Two pin joints at the upper corners of the yoke connecting to the lower end of each piston rod. Bushed with bronze or DU-type bearings; clearance held below 0.05 mm to prevent the rod from cocking under load.
- Grounding Link or Guide Pin: The third corner of the yoke is constrained either by a short grounding link to the crankcase or by a slot riding on a fixed pin. This is what kinematically forces the piston-rod pivots to track straight lines.
- Piston Rod Guide Bushing: A bushing in the cylinder gland that absorbs any residual side-load. Handles the small leftover lateral motion (typically <0.3 mm) so the piston seal itself sees only axial force.
- Counterweighted Crankshaft: Offset crank with integrated counterweights to balance the wobbling yoke mass. Without proper counterweighting a Ross-yoke engine vibrates noticeably at 1× crank speed because the yoke centre of mass orbits the crankpin.
Who Uses the Ross Yoke Mechanism
The Ross yoke shows up almost exclusively where an alpha Stirling engine needs to be small, low-vibration, and serviceable without the part count of a rhombic drive. You'll find it in solar dish-Stirling units, niche cogeneration plants, and a handful of cryocoolers running the cycle in reverse. It's not a mainstream IC-engine linkage — it's a specialty mechanism for builders who care more about piston-rod axiality than about power density.
- Solar Power Generation: Stirling Energy Systems' SunCatcher 25 kW dish-Stirling units used Ross-yoke-derived linkages in their kinematic Stirling engine to drive opposed pistons from concentrated solar heat.
- Combined Heat and Power: Whispergen MkV residential micro-CHP units shipped with a four-cylinder double-acting Stirling using Ross-yoke linkages, producing roughly 1 kW electric and 8 kW thermal output.
- Educational and Research Engines: Sunpower Inc. and various university Stirling research rigs at Ohio University used Ross-yoke test engines because the mechanism is simple enough to instrument and teach with.
- Marine Auxiliary Power: Small experimental Stirling gensets for sailboat house-bank charging in the 0.5–2 kW range, where the low-vibration signature of a Ross yoke is more valuable than peak efficiency.
- Cryocoolers: Reverse-cycle Stirling cryocoolers for laboratory liquefaction, where Ross-yoke variants drive opposed pistons at 30–50 Hz to reach 80 K cold-head temperatures.
- Demonstration and Hobby Engines: Andy Ross's own ST-5 demonstration engine and several published model-engineer Ross-yoke builds running on propane or alcohol burners at 500–1500 RPM.
The Formula Behind the Ross Yoke Mechanism
The number that matters most for a Ross-yoke designer is piston-rod lateral deviation — how far the piston rod swings sideways during a full crank revolution. This sets whether you can run a simple single-piston seal or whether you need a separate crosshead. At low crank angles near top-dead-centre the lateral deviation is near zero and the linkage is at its kinematic best. Through the mid-stroke region around 90° the deviation peaks. Push the stroke length too high relative to the yoke arm length and the deviation grows past what a guide bushing can absorb cleanly - the sweet spot for most Stirling builds sits at a stroke-to-arm ratio of roughly 0.6 to 0.8.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| δmax | Maximum lateral deviation of the piston-rod pivot from a straight vertical line over one revolution | mm | in |
| S | Piston stroke (twice the crank throw) | mm | in |
| Lyoke | Effective yoke arm length from crankpin centre to piston-rod pivot | mm | in |
Worked Example: Ross Yoke Mechanism in a 1 kW residential micro-CHP Stirling
You are laying out the Ross yoke for a 1 kW alpha Stirling micro-CHP unit similar in scale to a Whispergen MkV cylinder pair. The design stroke is S = 30 mm, target nominal speed is 1500 RPM, and you need to confirm the piston-rod lateral deviation stays below 0.4 mm so a single PTFE-graphite piston seal can run without a separate crosshead. The proposed yoke arm length is Lyoke = 45 mm.
Given
- S = 30 mm
- Lyoke = 45 mm
- N = 1500 RPM
Solution
Step 1 — at the nominal yoke arm length of 45 mm, compute the maximum lateral deviation:
That's far above the 0.4 mm seal tolerance — the back-of-envelope formula is telling you the simple yoke geometry alone won't carry this stroke. In a real Ross yoke the grounding link or slot pulls that deviation down by roughly an order of magnitude through kinematic constraint, so the effective figure is closer to 0.25 mm. The formula gives you the unconstrained worst case — useful as a sanity check on yoke arm sizing.
Step 2 — at the low end of typical Stirling-engine stroke-to-arm ratios, take a shorter stroke build at S = 20 mm with the same 45 mm arm:
The unconstrained deviation drops by more than half. With the grounding link engaged you'd see effective lateral motion under 0.12 mm — comfortably inside what a graphite-impregnated PTFE seal handles for 5,000+ hours of service.
Step 3 — at the high end, push to S = 40 mm on the same 45 mm arm to see what happens when a builder gets greedy with displacement:
That doubles the unconstrained deviation versus nominal. Even with the grounding link doing its work, effective lateral motion climbs past 0.45 mm — the seal scrubs, the rod heats locally, and you'll see oil-film breakdown within the first 100 hours. This is why experienced Ross-yoke designers cap stroke-to-arm ratio at about 0.7.
Result
Nominal unconstrained deviation lands at 2. 5 mm, which the grounding link reduces to roughly 0.25 mm at the piston rod — comfortably inside the 0.4 mm seal envelope. At a 20 mm short-stroke build the effective lateral motion drops to about 0.12 mm and seal life climbs past 5,000 hours; push the stroke to 40 mm and effective lateral motion exceeds 0.45 mm, which scrubs the seal and burns it within 100 hours of running. If you measure higher deviation than predicted on a built engine, the most common causes are: (1) wear in the grounding-link bushing letting the third yoke corner walk laterally, which directly multiplies piston-rod swing; (2) yoke-arm asymmetry from a machining error of more than ±0.05 mm between the two upper pivot bores, which cocks the yoke under load; or (3) crankpin needle-bearing radial play above 0.03 mm, which lets the entire yoke orbit eccentrically and shows up as a 1× crank-speed wobble on a dial indicator at the rod gland.
Choosing the Ross Yoke Mechanism: Pros and Cons
The Ross yoke competes head-on with the rhombic drive and the conventional connecting rod for opposed-piston Stirling layouts. Each has a clear engineering signature on side-load, part count, vibration, and cost. Pick based on what your engine actually needs to do, not on novelty.
| Property | Ross Yoke | Rhombic Drive | Conventional Connecting Rod |
|---|---|---|---|
| Piston-rod side load | Very low (<5% of gas load) | Essentially zero (perfectly balanced) | High (15–25% of gas load) |
| Part count per cylinder pair | 1 yoke + 1 crank + 1 grounding link | 2 connecting rods + 2 gears + 2 cranks | 1 connecting rod + 1 crank |
| Typical operating speed | 500–2500 RPM | 500–1800 RPM | 1000–6000 RPM |
| Vibration signature | Low, 1× crank if balanced | Near zero, inherently balanced | Moderate, 1× and 2× components |
| Manufacturing cost | Medium — one precision plate | High — twin gears + twin cranks | Low — proven supply chain |
| Bearing maintenance interval | 3000–5000 hours (crankpin needle bearing) | 5000–10000 hours | 5000+ hours |
| Best application fit | 0.5–3 kW alpha Stirling, micro-CHP | Beta Stirling with displacer + power piston on one axis | Mainstream IC engines, high-RPM compressors |
Frequently Asked Questions About Ross Yoke Mechanism
The yoke itself has mass that orbits the crankpin — the centre of mass of the triangular plate doesn't sit on the crank axis, so as the crank spins the yoke acts like a rotating eccentric. Pistons being correctly phased balances the gas-pressure forces, not the inertial forces of the linkage.
Fix it by either adding crankshaft counterweights sized to the yoke's polar moment, or by machining the yoke itself with mass distributed symmetrically around the crankpin bore. A rough rule: counterweight mass × counterweight radius should equal yoke effective mass × distance from crankpin to yoke centroid, within 5%.
It comes down to engine architecture. A rhombic drive only makes sense when you have a displacer and a power piston on the same axis — that's a beta or gamma Stirling. A Ross yoke is for alpha-configuration engines with two separate cylinders running 90° out of phase.
If you're building alpha because you want simpler heat exchanger geometry and easier service access, the Ross yoke is the natural choice. If you've already committed to a beta layout with concentric displacer and power piston, the rhombic drive's perfect balance is hard to beat and the Ross yoke isn't really an option.
The grounding link or slot constraint is almost never perfectly perpendicular to the cylinder axes once the engine is assembled. A few tenths of a degree of misalignment between the grounding pivot axis and the line connecting the two piston-rod pivots causes the yoke to favour one side through the wobble cycle, and the result is a stroke difference of 0.1–0.3 mm between the two cylinders.
Check it with a dial indicator at top-dead-centre on each piston while turning the crank by hand. If the strokes differ by more than 1%, shim the grounding-link mount until they match. Don't try to compensate by adjusting yoke pivot positions — that introduces a new error.
Aim for 0.6 to 0.7. Below 0.5 you're wasting linkage envelope — the yoke is bigger than it needs to be and you carry extra reciprocating mass. Above 0.8 the lateral deviation grows quadratically and seal life drops off a cliff.
Andy Ross's own published demonstration engines sit close to 0.67. That's a good starting point unless you have a specific reason to deviate, such as packaging constraints or a pre-existing crankshaft you're working around.
In principle yes, but the inertial loads on the yoke pivots scale with speed squared, and the crankpin needle bearing sees both gas force and yoke-mass centripetal force. At 4000 RPM the bearing load roughly triples versus 1500 RPM operation.
Practical Ross-yoke builds top out around 2500 RPM. Above that you need a substantially heavier crankpin bearing, careful yoke balancing, and the side-load advantage starts to erode anyway because oil-film behaviour at the guide bushing shifts. For high-speed work, a conventional connecting rod with a crosshead is the better tool.
Phase-angle drift is the usual culprit. The Stirling cycle is steeply sensitive to the angle between the two pistons — design phase is 90°, and every 5° of error costs roughly 7–10% of indicated power. On a Ross yoke, phase angle is set by the geometric placement of the two piston-rod pivot bores on the plate.
If those bores are off by 0.5 mm in the wrong direction, you've shifted phase by several degrees. Measure the actual phase angle by dial-indicating both pistons through a full crank revolution and noting the crank angle at each TDC. If you find the phase off, the fix is a new yoke plate — you can't shim your way out of bore-position error.
References & Further Reading
- Wikipedia contributors. Stirling engine. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
