A vibrating mulcher is easy to underestimate — the machine still cuts, the season is short, and the shaking gets filed under "they’re all like that". Meanwhile the vibration is quietly billing you: bearings and bearing seats hammer themselves loose, cracks open in the body, hydraulic fittings start weeping, bolted joints back off, fuel burn rises, and the operator absorbs the rest through the seat. The end state is always the same — the machine stops for repairs in the middle of the season, when downtime is most expensive.
The cause is almost always rotor imbalance: knives and hammers wear unevenly, a flail breaks off, a repair weld adds mass on one side. This guide covers how to balance these rotors where they work — on the machine, in their own bearings — and the agro-specific traps that make mulchers harder than fans.
Why the knife-edge trick doesn’t work here
Every workshop knows the traditional method: rest the rotor on horizontal guides, let the heavy side roll down, weld a counterweight on top, repeat until it stops rolling. That is static balancing, and for narrow, disk-shaped rotors — brake discs, single pulleys, grinding discs — it is perfectly adequate.
A mulcher rotor is the opposite shape: long, with mass spread along its length. Picture a heavy spot at the top on the left end and another at the bottom on the right end. On knife edges the two cancel — the rotor sits still and looks balanced. At working speed each heavy spot generates its own centrifugal force at opposite ends of the shaft, and the pair of forces rocks the machine violently. That is couple imbalance, a condition that only shows up in rotation: it cannot be seen statically, and it can only be corrected by measuring in two planes and placing weights at both ends. There is no way around instrumented two-plane balancing for long rotors.
Inspect before you measure
On agricultural rotors more than anywhere else, balancing fails because the machine has a defect that no weight can compensate. Check, with the rotor stopped:
- Knives and flails — complete set, matching weights, free on their pins. Replace missing or grossly worn knives in opposing pairs before balancing; a fresh knife set changes the mass distribution completely.
- Bearings and seats — play, overheating, hammered-out housings.
- Shaft — visible bend or wobble.
- Body — cracks, loose front curtain, loose pusher-frame and attachment bolts.
- Inside the tube — debris. Sand or soil packed inside the rotor tube shifts on every start; the symptom is a vibration phase that wanders from run to run. Spin the rotor several times: if readings refuse to repeat, open and clean the tube first.
Each of these shows up later as a balancing session that will not converge. Fixing them first is faster.
The field procedure
The kit: two vibration sensors, laser tachometer on a magnetic stand, reflective tape, the measuring unit and a laptop. Then:
- Mount the sensors at the two ends of the rotor, as close to the bearing units as possible, perpendicular to the rotation axis.
- Stick the reflective mark on the rotor or pulley; aim the tachometer at it. Shield the optics from direct sun.
- Select two-plane balancing in the software (F3 Two-plane).
- Run 0 — measure the baseline vibration at working speed.
- Weigh the trial mass, enter it in the software with its mounting radius, and fix it in plane 1 (the sensor-1 end). Size it by the rule below — light enough to stay safe, heavy enough to move the needle.
- Run 1 — the amplitude must change by at least 20–30%, or the phase by 20–30°; if not, the weight is too light. Go heavier, but stay under the force ceiling in the next section: past it, the trial weight itself becomes the hazard.
- Take the trial weight off plane 1 and put it on plane 2. Run 2. If it cannot be removed — a tacked-on weld, an awkward position — tick Remain trial weight for that plane instead and use a second trial weight on plane 2; the software then measures plane 2’s influence relative to Run 1 rather than Run 0. There is one such checkbox per plane, and the software treats it as a fallback, not as the normal route.
- The software returns a mass and an angle for each plane — angles are counted from the trial-weight position, in the direction of rotation. Remove any trial weight you did not tell the software to leave on.
- Weld (or bolt) the correction weights at the computed positions. On mulchers, welding is the norm — clean the spot, weld properly; a weight that flies off at working speed is a projectile.
- Run T (Trim) — the verification run. If the software asks for a small extra correction, add it and re-check.
Drive speed matters: the rotor must spin at the same speed on every run. Hold the tractor PTO or hydraulic drive steady, and let the reading settle before recording.
Sizing the trial weight
Too light and the run tells you nothing; too heavy and you have bolted a hazard onto a machine you already suspect. The working rule is to keep the trial weight’s centrifugal force F = m·R·ω² inside 5–10% of the rotor’s weight, which rearranges to:
m = k · M · g / (R · ω²) k = 0.05 … 0.10
with the rotor mass M in kilograms, the mounting radius R in metres, g = 9.81 m/s² and ω in radians per second (ω = 2πn/60 for speed n in rpm). The result comes out in kilograms — multiply by 1000 for grams. The software’s own estimate uses k = 0.07 for rigid supports; on soft supports it works from the measured amplitude instead and uses the centrifugal figure as an upper limit. Note how hard the speed term bites: the same rotor at twice the speed takes a quarter of the trial mass, so re-do the arithmetic whenever the working speed changes.
If your unit runs the unified balancing form, it can do this for you. That form is not the default: it lives behind F4 Settings → Balancing form → Modern, needs streaming firmware (v5+), and is off on a fresh install — until you switch it on, F3 opens the classic two-plane window, which has no estimator. In the unified form, open First trial weight estimate, type in the rotor mass and the mounting radius (it returns nothing without them), take Run 0, and it suggests a mass in grams plus a mounting angle. That angle is measured from the tachometer mark, not from the trial-weight position that every later correction angle uses, and it only means anything when the sensors are mounted coaxially with the tachometer.
When the numbers refuse to converge
If vibration drops weakly, or each run contradicts the previous one, stop adding weights and find the real cause. The usual suspects, in order of frequency:
- The machine moved — a support placed under the mower mid-session, a removed cover, retightened mounts. Any stiffness or mass change between runs invalidates the calibration.
- Unsteady speed — each run at a slightly different rpm.
- Resonance — the working speed sits on a structural resonance and readings swing wildly; changing the test speed slightly tells you quickly. Units with streaming firmware (v5+) also have an experimental Bump test tab — switch it on in F4 Settings → Experimental, then tap the stopped machine and read the natural frequencies off the response.
- Sensor or tachometer faults — sensor on dirt or loose magnet, tachometer nudged between runs, sunlight in the optics.
- A real defect — bearings, cracked body, debris in the tube (see the checklist above).
The full diagnostic tree lives in When Balancing Doesn’t Help.
What a finished job looks like
A worked example with numbers and timing: mulcher rotor balanced in the field in 40 minutes, from 12.4 mm/s down to an acceptable level without removing the rotor. As a rule of thumb, a mulcher that used to shake the tractor should come back to a reading your hand on the frame can barely feel — and the bearing budget feels it first.
One habit pays for the instrument faster than anything else: balance after every knife replacement, not when the vibration becomes alarming. A 40-minute session at knife change time is cheaper than a single hammered-out bearing seat. For picking the instrument itself, start with Choosing a Field Balancer.