Cams on shaft to alternating rod motion is a mechanism where two or more cams keyed to a single rotating shaft drive separate follower rods that reciprocate out of phase with each other. Typical industrial shafts run 60-300 RPM and produce stroke accuracy within ±0.05 mm when the cam profile is ground rather than milled. The arrangement converts one rotary input into multiple timed linear outputs, which is what packaging lines, textile looms, and engine valve trains depend on to coordinate motion without separate motors.
Cams on Shaft to Alternating Rod Motion Interactive Calculator
Vary keyway slop, cam radius, shaft speed, and phase offset to see timing error and alternating rod motion.
Equation Used
This calculator uses the article's keyway-slop timing relation. Tangential looseness divided by cam radius gives a small angular error in radians, then converts it to degrees. The same phase error is converted into milliseconds using the shaft RPM.
- Keyway slop is treated as tangential clearance at the cam radius.
- Small-angle relation is used for phase shift.
- Timing error is based on one full shaft revolution at the entered RPM.
- Positive slop reduces the effective phase separation shown in the diagram.
Operating Principle of the Cams on Shaft to Alternating Rod Motion
Mount two cams on the same shaft, key them at a phase offset (commonly 180°, but it can be anything the timing diagram demands), and each cam pushes its own follower rod up and down as the shaft turns. While one rod is on its rise stroke, the other is on its fall — that is the whole point. You get coordinated alternating motion from a single drive, with no electronic sync, no second motor, and no clutching. The cam profile itself sets the velocity, dwell, and acceleration of each rod, so the lift profile is doing the work that a servo would otherwise need to do.
The cam follower roller rides the cam profile under spring or gravity preload. If the preload drops below the peak inertia force at top of return stroke, the follower lifts off the cam — what we call follower jump — and you get a hammering noise plus stroke length errors of 0.5 to 2 mm. Tolerance on the keyway is critical too. A keyway slop of 0.1 mm at a 30 mm cam radius shifts the phase by roughly 0.2°, which sounds small until you realise the two rods are now overlapping in the dwell window where they should be alternating. On a high-speed packaging cam running 250 RPM, that phase error scuffs product against the rising rod.
Common failures are predictable: pitting on the cam flank from undersized roller followers, fretting on the keyway from undersized key cross-section, and bent follower rods from side-loading when the rod axis is not perpendicular to the cam face. The fix on all three is geometry — get the alignment right at build, not in service.
Key Components
- Drive Shaft: Common rotating shaft carrying both cams. Typically 20-40 mm diameter hardened steel for industrial work, supported on two bearings spaced wider than the cam pack to keep deflection below 0.05 mm at peak load.
- Cams (Two or More): Profiled disks keyed to the shaft at a defined phase offset. Ground cam profiles hold ±0.02 mm; milled-only profiles drift to ±0.1 mm and produce audible follower clatter above 150 RPM.
- Cam Follower Rollers: Sealed needle-bearing rollers that ride the cam profile. Roller diameter must be at least 1/3 the minimum cam radius of curvature or you get pitting on the cam flank within a few hundred hours.
- Follower Rods: The reciprocating output rods. Must run in linear bushings or guide bearings — never cantilevered — or side-load from the cam contact bends the rod within weeks of service.
- Return Spring or Gravity Preload: Keeps the follower roller pressed against the cam. Spring force at the top of return stroke must exceed peak follower inertia (m × a) by at least 30% or the follower jumps off the cam.
- Key and Keyway: Locks each cam to the shaft at the correct phase angle. Keyway tolerance under 0.05 mm — slop here directly translates into phase error between the two rods.
Industries That Rely on the Cams on Shaft to Alternating Rod Motion
This arrangement shows up anywhere you need two or more linear motions timed against each other from a single drive. It is the quiet workhorse behind countless industrial machines because it eliminates the cost and failure modes of separate actuators. You see it in printing presses, where ink rollers and form rollers must alternate against the plate cylinder; in textile machinery, where heddle frames lift and fall in defined sequence; and in any compact machine where adding a second motor is not an option.
- Printing: Heidelberg Speedmaster offset presses use multi-cam shafts to time ink fountain rollers against form rollers, alternating contact with the plate cylinder at 18,000 sheets per hour.
- Textile Machinery: Picanol air-jet looms drive the heddle frame lift mechanism from a multi-cam shaft, alternating up to 8 frames in defined sequence at 1,000 picks per minute.
- Packaging: Bosch cartoning machines use cams on a common shaft to drive the carton-erecting fingers and product-pushing rod in alternating phase at 200 cartons per minute.
- Internal Combustion Engines: Camshaft on a Cummins ISX diesel — intake and exhaust cams on the same shaft alternate to open valves in the four-stroke sequence at up to 2,100 RPM.
- Sewing Machines: Industrial Juki lockstitch heads use a top shaft with two cams driving the needle bar and presser foot in alternating phase at 5,000 stitches per minute.
- Bottling: Krones rotary fillers use shaft-mounted cams to alternate fill-valve open and bottle-clamp engagement, coordinating motion across 80 fill stations.
The Formula Behind the Cams on Shaft to Alternating Rod Motion
What you usually want to know up front is the linear velocity of each follower rod at any shaft angle, because that velocity sets your inertia loads, your contact stress, and whether the follower will jump off the cam. At the low end of the typical operating range — say 30 RPM on a slow indexing machine — follower velocity is gentle and follower jump is not a concern. Push the same cam to 200 RPM and peak follower velocity quadruples, peak acceleration quadruples again (it scales with the square of speed), and now you have to size the return spring and the cam profile around peak acceleration, not peak velocity. The sweet spot for most ground steel cams with needle-roller followers sits between 60 and 180 RPM.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| vf | Instantaneous follower rod velocity | m/s | in/s |
| ω | Shaft angular velocity (= 2π × N / 60 for N in RPM) | rad/s | rad/s |
| dh/dθ | Slope of the cam lift profile at angle θ (rate of follower lift per radian of shaft rotation) | m/rad | in/rad |
| θ | Shaft angle from cam reference position | rad | rad |
Worked Example: Cams on Shaft to Alternating Rod Motion in a Heidelberg-style ink roller drive
Sizing the follower velocity for an ink-distribution shaft on a sheet-fed offset press, where two cams 180° out of phase drive two ink-form rollers against the plate cylinder. The cam has a 12 mm rise over 90° of shaft rotation using a cycloidal profile. The press runs nominally at 120 RPM, with a typical operating range of 60 to 240 RPM. We want to know peak follower rod velocity at each end of that range so we can size the return spring and check follower jump.
Given
- Cam rise (hmax) = 12 mm
- Rise angle (β) = 90 deg
- Profile type = Cycloidal —
- Nnom = 120 RPM
- Nlow / Nhigh = 60 / 240 RPM
Solution
Step 1 — for a cycloidal cam profile, peak slope (dh/dθ)max occurs at the midpoint of the rise and equals 2 × hmax / β. Convert β to radians:
Step 2 — at nominal 120 RPM, convert to angular velocity and compute peak follower velocity:
That is fast enough to need a real return spring but slow enough that follower jump is not yet the limiting factor — the press operator will hear a clean rhythmic tap at the cam contact, not a hammer.
Step 3 — at the low end, 60 RPM:
At this speed the rollers transfer ink smoothly, the follower preload is barely tested, and you could honestly run a much weaker spring. This is the regime classroom demo presses live in.
Step 4 — at the high end, 240 RPM:
Peak acceleration scales with ω2, so going from 120 to 240 RPM quadruples the inertia force on the follower. A return spring sized for the nominal case will be marginal here — you will see follower jump and the press will print streaks because the ink roller stops following the cam profile cleanly at top of return stroke.
Result
Peak follower velocity at the nominal 120 RPM is 0. 192 m/s. In practice, that is the velocity at which a properly tensioned return spring still keeps the roller glued to the cam and the press lays down even ink — you can hear it as a steady tick rather than a clatter. At 60 RPM the rod cruises at 0.096 m/s and any reasonable spring works; at 240 RPM the rod theoretically reaches 0.384 m/s but follower jump kicks in around 200 RPM with a stock spring, which is why production presses sit in the 100-180 RPM band. If your measured stroke comes up short of the 12 mm cam rise, the most common causes are: (1) a worn cam flank with pitting on the rise side, knocking 0.3-0.8 mm off the effective rise, (2) keyway slop on the cam-to-shaft fit letting the cam shift phase by 1-2° under load, or (3) a follower roller diameter under 1/3 the cam's minimum radius of curvature, which lets the roller dig into the profile rather than ride it cleanly.
Choosing the Cams on Shaft to Alternating Rod Motion: Pros and Cons
Cams on a shared shaft are not the only way to get alternating rod motion. The two main alternatives are eccentric (or crank) drives on a shared shaft, and independent linear actuators electronically synchronised. Each has a clear regime where it wins. Here is how they stack up on the dimensions that actually matter when you are sizing a machine.
| Property | Cams on Shaft | Eccentrics on Shaft | Synced Linear Actuators |
|---|---|---|---|
| Max practical shaft speed | 60-300 RPM (ground profile) | 300-1500 RPM | Limited by actuator: 60-200 cycles/min |
| Stroke profile flexibility | Any profile — dwells, asymmetric rise/fall | Pure sinusoidal only | Fully programmable |
| Stroke accuracy | ±0.02 mm ground, ±0.1 mm milled | ±0.05 mm typical | ±0.05 mm with closed-loop control |
| Cost per axis | Low — one motor, mechanical timing | Lowest — simplest geometry | Highest — motor + drive + controller per axis |
| Phase change at runtime | Not possible without re-keying cam | Not possible | Trivial — change a parameter |
| Lifespan at rated load | 50,000+ hours with sealed followers | 100,000+ hours (no sliding contact) | 10,000-30,000 hours actuator-dependent |
| Best application fit | Coordinated multi-rod motion with custom profile | High-speed simple sinusoidal motion | Machines needing recipe changes between products |
Frequently Asked Questions About Cams on Shaft to Alternating Rod Motion
Phase error rarely comes from the keyway angle itself — it comes from keyway slop combined with cam profile asymmetry. If your cam has a steeper rise on one side than the other, the follower's effective contact point shifts under load, which can move the apparent phase by 2-4° even with a perfectly cut keyway. Check the cam fit on the shaft with a dial indicator: rotate the shaft slowly, and if the cam shifts more than 0.03 mm radially, your key is undersized.
The other common culprit is shaft torsional wind-up. On a long shaft with cams at both ends, the trailing cam can lag the leading one by 1-2° under peak load. Stiffen the shaft or move the cams closer to the drive end.
Pick on the basis of what you are trying to control. Cycloidal gives the smoothest acceleration curve with zero jerk at start and end of rise, so use it when shaft speed is high and you need to minimise vibration — anything above 150 RPM benefits noticeably. Harmonic (simple cosine) is fine below 100 RPM and is the cheapest to grind. Modified trapezoidal gives the lowest peak acceleration for a given rise and time, so use it when the follower is heavy and inertia is your limit, even though it has slightly higher jerk than cycloidal.
Rule of thumb: if your peak follower acceleration calculation comes out above 50 m/s², switch to modified trapezoidal. If your machine vibrates audibly at speed, switch to cycloidal.
Three likely causes, in order of frequency. First, your cam profile slope is being eroded at the contact point — a roller follower that's too small (under 1/3 the minimum radius of curvature) digs slightly into the cam, effectively reducing the slope it sees. Replace the follower with a larger roller and the velocity will come up.
Second, follower rod guide friction. If the rod runs in plain bushings rather than linear ball bearings, breakaway friction can be high enough that the follower lags the cam profile slightly during fast accelerations. Measure with the rod removed and the follower running free — if the velocity matches prediction, the rod guide is your problem.
Third, drive shaft elasticity. A long, thin shaft winds up under load and the cam doesn't actually reach the commanded angular velocity at the moment of peak slope. Stiffen the shaft or shorten it.
Switch when one of three conditions is true. (1) You need to change the phase or stroke between product runs — a cam shaft locks you into one timing, a servo lets you change a parameter. (2) Your machine has more than 4 alternating axes — beyond that, the cam shaft gets long, torsionally soft, and hard to time accurately. (3) Your throughput requirement sits below 60 cycles per minute — at that speed cams are overkill and servos are easier to specify.
Stay with cams when throughput is above 100 cycles per minute, the motion profile is fixed for the life of the machine, and you have more than two coordinated axes. The mechanical timing is essentially free at that point and servos cost you 10x more per axis.
That tick is almost always follower jump — the roller lifting off the cam profile briefly and then slamming back down. It happens at the angle of peak negative acceleration, which on a symmetric cycloidal cam is at the top of the rise where the follower transitions from accelerating to decelerating. The return spring force at that exact angle is less than the inertia trying to throw the follower off the cam.
Diagnose by slowing the shaft to half speed. If the tick disappears, you have follower jump and need a stiffer return spring (target preload ≥ 1.3 × peak follower mass × peak acceleration). If the tick persists at all speeds, you probably have a chip or burr on the cam flank at that angle — inspect with a fingernail or a fine indicator.
You can usually add one, but check two things first. Bearing span — adding a cam in the middle of an existing shaft increases the bending load between bearings, and shaft deflection scales with the cube of the unsupported length. If your existing shaft already deflects 0.03-0.04 mm at peak load, a third cam will push you past 0.1 mm and you'll see follower contact issues.
Second, torsional stiffness. The drive end has to deliver peak torque to all three cams simultaneously when their lift profiles overlap. Sum the peak torques and check the shaft can deliver it without winding up more than 0.5°. If you're tight on either count, step the shaft up one diameter rather than trying to retrofit.
References & Further Reading
- Wikipedia contributors. Cam (mechanism). Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
