Preparing the Machine for Balancing

← Balanset-1A Knowledge Base — all chapters

Preparation is half the battle. Most “failed balancings” fail exactly here: the device honestly measures the vibration of a machine with loose bearings or a cracked frame, and its calculations won’t converge.

Universal checklist (any machine)

1. Bearing play. Remove the belts / disengage the drive and rock the rotor with a pry bar. There must be no perceptible play. Play = replace the bearings first, then balance. Real case: 2 mm of play made balancing impossible until the bearings were replaced and the seat restored (build-up welding + re-machining of the journal).
2. Cracks. Inspect the housing, frame, weld seams, pulleys. A crack “breathes” under load and makes the system nonlinear. Weld cracks up. There is a known case where a crack in a pulley blocked balancing.
3. Fasteners. All support, motor and frame bolts tightened.
4. The rotor in working condition. All blades/hammers/knives in place. Replace worn elements in pairs, opposite each other (180° apart), otherwise you create unbalance yourself.
5. The rotor is clean. Caked dirt, material, ice — that is unbalance too, and a “wandering” one.
6. Stable RPM. The drive must hold its speed; between runs of a series no more than ~100 rpm difference.

Mulcher and mower specifics

The most common machinery among Balanset-1A users. The specifics:

• Lift the mulcher (machine in the air on the hydraulics) — the rotor must spin freely; the machine must not rest on its working body.
• The front flap/curtain and the pusher frame. Freely hanging parts rattle and spoil the readings. They must be tack-welded to the housing through metal spacers (for the duration of balancing) or removed. This is a recurring support recommendation: an unsecured hydraulic flap of a forestry mulcher wrecks the balancing. In experience, welding the flap “solves the problem in nearly half of the cases”.

• Hammers/knives installed (preferably new or in equal condition). Balancing a “bare” rotor without the working tools is a mistake: with knives fitted the balance will be different.
• Free-swinging knives (chain, hinged) don’t interfere with balancing as long as they are not seized. A seized knife = unbalance.
• Sensors — on top or on the side of the bearing housings, not on the rotor end face.
• Mind the weld-seam mass: welding a weight on adds ~100 g of weld metal — noticeable at the fine-trim stage.

Where to mount the sensors

• The sensor goes on the bearing housing, as close to the bearing as possible, on a clean flat spot (the magnet must sit flush).
• Measuring axis — radiālajā (usually horizontal, across the rotor axis). The horizontal direction is normally “softer” and gives a bigger signal.
• X1 — plane 1 (e.g. the left support), X2 — plane 2 (the right one). Remember which is which and don’t swap them until the job is done.
• Secure the cables so they can’t be dragged into the rotor — that is the most common way to kill a sensor.

Overhung rotors (fans on the motor shaft)

On an overhung rotor (the impeller “hangs” on the shaft end, as on most centrifugal fans) the correction planes are on the impeller itself — the front and back discs — while the sensors go on the motor’s bearing housings. The sensor-to-plane pairing is shown by colour in the diagram: sensor 1 works with plane 1 (the outer disc), sensor 2 with plane 2 (the inner disc, nearer the motor).

Tachometer and the mark

• Stick a piece of reflective tape on the hub or shaft — not on a blade and not on a shiny surface.
• Tachometer on the magnetic stand, beam onto the mark at a slight angle, distance per the manual. Put the stand on a non-vibrating surface.
• Check: the RPM in the program must match reality and be stable. The usual traps (sun, glare, double marks) — chapter 08.
• Once the measurement series starts, do not re-stick the mark or move the tachometer.

A test run in vibrometer mode

Before balancing, start the machine and watch vibrometer mode:

1. RPM reads and is stable.
2. Overall vibration and the 1x component are stable from run to run (amplitude scatter < 10–20%, phase < 10–20°).
   

   

3. Assess the starting level: if overall vibration is huge (30–50+ mm/s) there is most likely a mechanical problem or resonance; above 50 mm/s support recommends balancing in two stages, starting at reduced RPM.
4. Look at the spectrum: 1x dominating is good news — it’s unbalance, balancing will help. 2x and higher dominating, or peaks away from the rotation frequency — find the cause first (chapter 07).

Choosing the balancing RPM

• Ideally — the working RPM. But if the working RPM sits on a resonance or the vibration is off the scale, start lower: for mulchers typically 600–900 rpm, then work your way up.
• The RunDown (coast-down) chart helps find “clean” RPM: resonance peaks are immediately visible — balance between them. Real example: test-rig resonances at ~420 and ~644 rpm → balancing at ~844 rpm.

  

  

• All runs of one series must be at the same RPM.

Choosing the trial weight

The trial weight must change the vibration, otherwise the device has nothing to compute from. The criterion: a 20–30% change of amplitude or a 20–30° change of phase after the trial weight is installed.

Field reference points:

• mowers, light mulchers: 200–300 g;
• large forestry mulchers: 500–700 g;
• a ~600 kg rotor: 400–600 g;
• very large rotors (tonnes): kilograms (a 1.7 t rotor took ~1 kg).

The estimation formula support gives (rotor mass Mr in kg, mounting radius Rp in m, RPM N):

Mp ≈ Mr / (Rp × (N/100))²   [kg]

The software also has a trial-weight calculator (vibromera.eu/content/trial-weight-calculator/). A trial weight that is too mazais — the change drowns in noise (real case: 3–7 g on a turbine changed almost nothing; 5 g worked once the background noise was reduced); too big — pushes the system into nonlinearity.

Fix the weight securely: a clamp, a bolt, a magnet at low RPM — but so it cannot fly off. Weigh it on the kit scales.