A Three-wiper Rectangular Frame Reciprocator is a cam mechanism in which a single rotating eccentric drives a rectangular frame, and three wiper plates riding inside that frame each pick off a phased linear stroke. The configuration appears in Henry T. Brown's 1868 catalogue "507 Mechanical Movements" as movement No. 191. The frame translates the eccentric's circular motion into pure reciprocation, while the wiper geometry taps three offset linear outputs from one input shaft. That gives you three synchronised strokes — typically 0°, 120°, and 240° apart — from one motor, which is why it still shows up on label applicators, ink-pad printers, and small dosing heads.
Three-wiper Rectangular Frame Reciprocator Interactive Calculator
Vary eccentric pin offset, shaft angle, and speed to see the resulting stroke and three phased wiper positions.
Equation Used
The article states that frame stroke is exactly twice the eccentricity: a 10 mm offset pin produces a 20 mm frame stroke. The three wiper positions are shown as ideal phased sinusoidal displacements from the mid-stroke position.
- Ideal scotch-yoke motion with no compliance or slip.
- Three wiper outputs are treated as 120 deg phased sinusoidal strokes.
- Displacements are measured from mid-stroke.
- Clearance, wear, and spring lift-off are ignored in the kinematic calculation.
Operating Principle of the Three-wiper Rectangular Frame Reciprocator
The mechanism starts with an eccentric pin on the input shaft. That pin sits inside a rectangular frame — sometimes called a wiper plate cam or scotch-yoke variant — and as the shaft rotates, the frame is forced to translate back and forth along one axis while the pin slides freely across the frame's internal width. Pure rotary-to-linear conversion, no rocker arms, no return springs. The stroke length equals exactly twice the eccentricity, so a 10 mm offset pin gives you a 20 mm frame stroke. Simple geometry, predictable kinematics.
The "three-wiper" part is where it gets useful. Three follower plates ride against three different working faces of the frame — typically the two short ends and one long edge — and each follower picks off motion at a different phase of the cycle. Two wipers at the short ends move 180° out of phase with each other (one extends while the other retracts), and the third wiper, riding the long face, sees a phased linear stroke that's geometrically derived from the frame's transverse position. You get three coordinated outputs from one motor, no gear train needed.
Tolerances matter more than people expect. The frame-to-pin clearance must sit around 0.05–0.10 mm — much tighter and the pin galls under thermal expansion; much looser and you'll hear an audible clack at top-dead-centre as the pin reverses against the frame wall. Wiper-to-frame clearance runs similar. If you let the wiper face wear past about 0.15 mm clearance, the follower output picks up backlash that shows as a visible jitter at stroke reversal, and on a printing application that translates straight to ghosted impressions. The most common failure mode is wiper-face wear caused by the eccentric pin running dry — keep the contact faces oiled, or specify a hardened wear plate.
Key Components
- Drive shaft with eccentric pin: Carries the offset pin that drives the frame. Pin offset sets the stroke — for a 25 mm output stroke the eccentricity is 12.5 mm. Shaft typically runs 30–300 RPM in industrial use; above ~400 RPM the frame's reciprocating mass starts producing measurable vibration that needs counterweighting.
- Rectangular frame (yoke): The closed rectangular slot the eccentric pin rides inside. Frame must be ground flat to within 0.02 mm across the working faces or you get uneven wiper contact. Material is usually case-hardened tool steel — hardness ≥58 HRC on the working faces, otherwise the pin tracks a groove inside 200,000 cycles.
- Three wiper followers: Spring-loaded or gravity-loaded plates riding against three faces of the frame. Each picks off a phased linear output. Contact faces should match the frame hardness within 5 HRC points to keep wear balanced — mismatch and one part wears out first.
- Output slides or pushrods: Carry the wiper motion to the working tools — print heads, label applicators, dosing nozzles. Slide guides need to be parallel to the wiper travel within 0.1°, otherwise side-loading on the wiper face accelerates wear and introduces stroke loss.
- Return spring or gravity bias: Holds each wiper against the frame face during the retract half of the cycle. Spring rate is sized so preload is at least 1.5× the friction force at maximum stroke speed — if it's too soft the wiper lifts off the frame at top speed and the output goes erratic.
Real-World Applications of the Three-wiper Rectangular Frame Reciprocator
The reciprocator earns its place anywhere you need three coordinated linear strokes off one motor, where a servo system would be overkill and a cam stack would take up too much space. You see it most in low-to-medium-speed packaging, printing, and inspection automation — applications where the phase relationship between the three strokes is fixed by design and never needs reprogramming. Its weakness is high speed: above roughly 400 RPM the reciprocating frame mass demands counterbalancing, and most builders move to a different mechanism rather than fight it.
- Packaging machinery: Three-station label applicator on a Herma 132M wet-glue labeller, where one shaft drives glue-pot wiper, label-pickup arm, and applicator pad in fixed phase.
- Printing and marking: Pad-printing ink transfer on a Comec ETP 200 tampo printer — wiper 1 floods the cliché, wiper 2 doctors the surface, wiper 3 lifts the pad.
- Pharmaceutical dosing: Three-head powder dosing on a Bonfiglioli BMS-style monoblock filler, dispensing 50–500 mg fills into vial trays at 60 cycles per minute.
- Inspection automation: Triple-probe gauging head on an automotive valve-seat checker, like the units Marposs builds for cylinder-head lines, where three measurement plungers contact the workpiece in sequence.
- Textile finishing: Three-roller starch wiper bank on a Karl Mayer warp-sizing machine, applying sizing solution in three coordinated passes per shaft revolution.
- Cosmetic filling: Lipstick capping line at a Marchesini ML 510 — three wiper plates close cap, knurl the closure, and eject the unit in one shaft turn.
The Formula Behind the Three-wiper Rectangular Frame Reciprocator
What you usually need to know before sizing the rest of the line is the linear velocity of the wiper output at any given input RPM. The frame moves sinusoidally — same as a scotch yoke — so the wiper sees a peak velocity that depends on shaft speed and eccentricity. At the low end of the typical operating range (around 30 RPM) the peak wiper velocity is gentle and you can hand-build the contact faces from mild steel. At the nominal 60–120 RPM you're in the sweet spot where standard hardened steel and a basic oil bath handle everything cleanly. Push above 250 RPM and the peak velocity climbs to where wiper bounce becomes the limiting factor — you'll need stiffer return springs and balanced frame mass.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| vpeak | Peak linear velocity of the frame (and hence the wiper output) | m/s | in/s |
| N | Drive shaft rotational speed | RPM | RPM |
| e | Eccentricity of the drive pin (½ of total stroke) | m | in |
| S | Total wiper stroke (= 2 × e) | m | in |
Worked Example: Three-wiper Rectangular Frame Reciprocator in a triple-head ampoule labelling station
You are sizing the three-wiper rectangular frame reciprocator that drives the glue wiper, label pickup, and pressure pad on a Herma-style three-head ampoule labelling station running 90 ampoules per minute. Eccentricity is 15 mm (giving 30 mm total stroke), and the line runs nominally at 90 RPM with operating range 30–250 RPM depending on label size and ampoule diameter.
Given
- e = 0.015 m
- S = 0.030 m
- Nnom = 90 RPM
- Nlow = 30 RPM
- Nhigh = 250 RPM
Solution
Step 1 — at the nominal 90 RPM, compute the angular velocity in rad/s:
Step 2 — compute peak wiper velocity at nominal speed:
That's a brisk but controllable stroke — you can watch each wiper move clearly with the naked eye, and a hardened-steel wiper face running in an oil mist sees no measurable wear over a million cycles. This is where most three-wiper frame reciprocators are designed to live.
Step 3 — at the low end of the operating range, 30 RPM:
At 47 mm/s the motion looks almost languid. Glue pickup on the wiper plate is generous, label registration is forgiving, and you can run mild steel wear plates with hand-applied grease. The downside: throughput drops to 30 ampoules per minute — fine for short runs, useless for production.
Step 4 — at the high end, 250 RPM:
At nearly 400 mm/s the frame is reversing direction 500 times per minute and the inertial loads on the wiper return springs climb roughly 8× over the nominal case (acceleration scales with N²). In practice you'll see wiper bounce at top-dead-centre unless the springs are upgraded, and the frame itself needs counterweighting or you'll feel the whole machine shaking at 8.3 Hz. Most builders cap this style of mechanism around 200 RPM in production.
Result
Peak wiper velocity at nominal 90 RPM is 0. 141 m/s — fast enough to give clean label transfer but slow enough that hardened wear plates last the life of the machine. The full operating range runs from 0.047 m/s at 30 RPM (visible, gentle, low-throughput) up to 0.393 m/s at 250 RPM (where inertial bounce becomes the dominant failure mode), with the design sweet spot sitting between 60 and 120 RPM. If you measure peak velocity below the predicted figure, the most likely causes are: (1) eccentric pin slip on the drive shaft — check the keyway or grub screw, a 0.5 mm slip on a 15 mm eccentric loses 7% stroke; (2) frame-to-pin clearance opened past 0.15 mm letting the pin lose contact at reversal, which you'll hear as a tick at TDC; or (3) drive-side belt slip if you're running a timing belt off the gearmotor, which presents as a velocity that drops only at higher RPM.
Three-wiper Rectangular Frame Reciprocator vs Alternatives
The three-wiper frame reciprocator earns its keep when you need three rigidly phased strokes off one shaft and you don't need to reprogram them. Compare it to a cam stack (separate disk cams stacked on one shaft) and to three independent servo slides — the comparison comes down to RPM range, cost, and how often the phasing needs to change.
| Property | Three-wiper Rectangular Frame Reciprocator | Stacked Disk Cam Set | Three Independent Servo Slides |
|---|---|---|---|
| Maximum practical RPM | ~200 RPM (frame inertia limited) | ~600 RPM (lighter follower mass) | Limited only by servo bandwidth, typically >1000 equivalent cycles/min |
| Phase reprogrammability | Fixed by geometry, not adjustable | Re-keyable on shaft (manual) | Software-adjustable in real time |
| Capital cost (3-output system) | Low — single eccentric, one frame, three wipers | Medium — three precision-ground cams | High — three servos, drives, controller |
| Stroke accuracy and repeatability | ±0.05 mm typical, governed by frame clearance | ±0.02 mm with ground cams | ±0.005 mm closed-loop |
| Maintenance interval (typical) | Wiper face inspection every 500k cycles | Cam profile inspection every 2M cycles | Servo brush/encoder check annually |
| Best application fit | Mid-speed packaging, printing, dosing with fixed phasing | High-speed presses where each output needs a unique motion profile | Flexible lines running multiple product formats per shift |
| Footprint | Compact — single shaft, single frame block | Wider — cam stack along shaft | Largest — three actuator envelopes plus cabling |
Frequently Asked Questions About Three-wiper Rectangular Frame Reciprocator
Because the third wiper rides the long face of the rectangular frame, not the short ends. The two short-end wipers see the full ±e displacement directly, but the long-edge wiper sees a derived motion — its actual stroke depends on where along the frame's long axis the contact point sits, and on the geometry of the wiper's own pivot or guide.
If you want all three outputs to have identical stroke length, you have to either move all three wipers to symmetrical short-edge positions (which limits you to two truly independent phases) or design the long-face wiper with a mechanical advantage ratio that compensates. A common error is assuming the long-edge wiper sees the same stroke as the short-edge ones — it usually doesn't, and a quick dial-indicator check on each output before you commit to the rest of the build saves a lot of grief.
If the phasing is fixed and you'll never need to change it, the frame reciprocator wins on three fronts: it can't drift out of phase (the geometry enforces timing), it has no air consumption, and it costs roughly half what three solenoid-valved cylinders cost once you include the manifold and controls.
Pneumatics win if you need to skip strokes, change dose volumes on the fly, or run different products through the same head. At 60 cpm both will work — the decision is purely about whether the process is rigidly synchronous (frame wins) or flexible (cylinders win).
This is almost always thermal expansion in the frame closing up the pin clearance. The drive shaft heats from bearing friction, the pin grows slightly, and if your initial cold clearance was on the tight side of the 0.05–0.10 mm window the pin starts dragging on the frame at temperature.
Drag manifests differently on each wiper because the long-edge wiper sees the smallest direct contribution from the pin and is the most sensitive to frame distortion. Check pin-to-frame clearance at operating temperature with a feeler gauge — if it's under 0.03 mm you need to either open the bore by 0.02–0.03 mm or move to a lower-friction bushing material. Bronze impregnated with PTFE solves this on most builds.
Geometrically yes — you have four faces on a rectangle. Practically the fourth output duplicates the phase of one of the other three (the two short edges are 180° apart, the two long edges are 180° apart). So a four-wiper version gives you two pairs of opposing strokes, not four independent phases.
If you genuinely need four independent phases off one shaft, you're better off with a stacked disk cam set or stepping up to a globoid indexer with auxiliary cams. The rectangular-frame topology fundamentally caps you at three useful phases.
Run a quick A/B test. Mark the drive shaft and the frame independently with paint marks, then jog the shaft through one full revolution by hand and measure each wiper's stroke with a dial indicator. If the strokes match the predicted 2e value within 0.1 mm, the kinematics are clean and your problem is on the drive side — belt slip, coupling slop, or motor torque dropping at speed.
If the manual stroke is already short, the frame or pin is worn. The diagnostic giveaway: drive-side problems get worse with RPM, kinematic problems are constant across RPM. A velocity that's 10% low at 30 RPM and still 10% low at 200 RPM points at the frame, not the motor.
Rule of thumb: counterweight when peak frame acceleration exceeds about 50 m/s². For a 15 mm eccentricity that crosses the threshold around 175 RPM. Below that the unbalanced reciprocating mass typically produces vibration the machine frame can absorb without complaint.
Above that threshold you'll either feel it through the floor or see it in the inspection-camera image as a periodic blur. The fix is a counter-rotating balance mass on the same shaft, sized to the frame's reciprocating weight times its eccentricity. On a small dosing head this is often just a brass slug bolted to a crank web — five minutes of work that cleans up the vibration completely.
References & Further Reading
- Wikipedia contributors. Scotch yoke. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
