A phase accuracy of ±1° is the practical specification carried by mainstream portable two-plane dynamic balancers, and the value is matched by the 5% amplitude / 1° instrumentation envelope that B&K documented as the threshold below which repeat trial-runs become unnecessary [S2][S5].
Phase here means the angular position of the 1× rotation-frequency vibration vector relative to a once-per-revolution reference pulse, normally read from an optical tachometer or photoelectric trigger; that angle is the coordinate the operator uses to place a correction weight, so its error maps directly into a placement error of the heavy-spot [S2][S5].
What ±1° phase accuracy actually means at the correction plane
A 1° placement error on a 100 mm correction radius at 3000 rpm is a tangential position error of about 1.75 mm, which translates into a residual unbalance vector that grows with rotor mass and speed; portable field balancers therefore quote ±1° in the same data sheet that lists a residual unbalance as low as 2 mg·mm for small rotors and 200 mg·mm for large rotors [S3][S5]. On a soft-bearing bench the same 1° target is achieved without per-rotor calibration, because the suspension itself is treated as an uncalibrated mechanical stage and the influence coefficients are extracted from a single trial run or a rotor photograph plus one reference dimension [S3].
Brüel & Kjær's own 2013 application note put the practical bar at 5% amplitude combined with 1° phase, and stated that crossing that envelope removes the need for a third trial run on rigid rotors; the note also recommended a photoelectric once-per-revolution trigger and dual accelerometers near the bearings as the minimum instrumentation [S2]. The same ratio logic is why SAE ARP4162 proving rotors and test weights are still used to verify any production balancing machine before a run is accepted [S4].
Why high-speed and dual-speed balancers push phase accuracy tighter
Centrifugal force scales with the square of angular velocity, so a higher balancing speed magnifies the same mass eccentricity into a larger, easier-to-measure vibration vector; this is the engineering reason a high-speed permanent-magnet rotor balancer published in 2026 uses dual-speed control to keep amplitude high enough to identify unbalance while avoiding resonance crossings [S1]. The same paper reports a measured vibration reduction of 15.8% (mean σ = 10.2%) for the optimized method against 19.7% (σ = 12%) for a baseline, a difference the authors treat as significant at the 96% confidence level [S1].
Phase error budgets tighten on these machines because the influence-coefficient method treats amplitude and phase as a vector pair; an unaccounted phase lag in the sensor chain or in the optical trigger conditioning directly biases the calculated correction mass, and at 10 000 rpm and above that bias becomes the dominant residual once amplitude noise is filtered out. For flexible or coreless rotors the practical answer is to balance at the operating speed rather than at a lower convenience speed, so the 1° phase spec is only meaningful if it is held at the rotor's rated rpm, not at a slow crawl [S1].
Soft-bearing vs hard-bearing benches: where phase accuracy comes from
On a soft-bearing balancer the supports are designed to move with the rotor, so what the displacement sensor reads is the response of a low-stiffness suspension, not the centrifugal force directly; the machine is calibrated per rotor family, which is why historically these benches were the most accurate but also the most labour-intensive [S3][S4]. Hard-bearing benches measure centrifugal force through stiff supports and are permanently calibrated, so a single set of "abc" dimensions (radius plus axial distance to each correction plane) and a proving run with a test weight is enough to bring the machine to spec, removing the per-rotor calibration overhead [S4].
The practical accuracy floor is set by the unbalance reduction ratio (URR): a 95% URR machine can in theory take a rotor from 20× its tolerance to within tolerance in one run, while a 90% URR machine only achieves a 10× reduction per run and so demands more iterations, especially for overhung correction planes, closely spaced planes, or large-diameter rotors where plane separation is poor relative to the support span [S4]. Phase accuracy interacts with URR because each iteration re-reads the residual vector; a tighter phase spec means fewer iterations are needed to converge on the same residual unbalance, which is the real cost driver on a production floor [S2][S4].
Selection criteria: phase accuracy in context
For a portable field balancer in a service workshop, the binding specs are usually rotor mass range (for example 50 g to 50 t for the soft-bearing ERBESSD family), residual unbalance in g·mm, URR, and software features such as influence-coefficient calculation and FFT spectrum display, with phase accuracy as a check that the instrument chain is healthy [S3][S5]. A two-channel portable instrument that reads amplitude and phase at both bearings simultaneously, like the Balanset-1A, targets the same ±1° / 5% envelope in the field, with verified operation on rotors up to 24 000 kg [S5].
For a production balancing machine the binding specs are URR (commonly 90–95% per single run), plane separation limits, maximum rotor diameter, and calibration interval per ISO 21940-11 or SAE ARP4162, with phase accuracy reported on the type-test certificate rather than on every day-shift log [S4]. A flexible or high-speed rotor pushes the choice toward a hard-bearing bench with rotor-specific calibration features and a built-in trial-weight wizard, because the alternative, a soft-bearing bench calibrated for one rotor family, cannot follow a frequent model changeover without re-calibration downtime [S3][S4].
The cross-reference between dynamic balancing instruments and adjacent process instrumentation is direct: the same once-per-revolution optical trigger used to read phase on a balancer is also the reference channel for a vibration analyzer running order tracking, so a phase-accurate balancer doubles as a diagnostic tool once the unbalance has been removed. For three-phase drive trains the unbalance source is often the rotor of the driving three-phase asynchronous motor, which is why balancing specs are quoted on motor data sheets as a balancing quality grade G (formerly Q) at a stated rotation speed [S6].
Limits, failure modes, and what ±1° does not fix
Phase accuracy is meaningless if the reference trigger is misread; common field failures are reflective-tape misses on a dirty shaft, electrical noise on long trigger cables, and a photodiode mounted at the wrong angle so the pulse timing shifts with runout, all of which manifest as a phase reading that drifts by several degrees run-to-run [S5]. The mitigation is a magnetic or optical pickup with a clear once-per-revolution event, accelerometers torqued to a known stud, and a verification run with a trial weight whose position is read back before any production correction is trusted [S2][S5].
Balancing cannot remove bearing play, misalignment, cracks, a bent shaft, or resonance; field data summarized in 2025 places these mechanical faults as the root cause in roughly 90% of "the balancer would not converge" cases, leaving true unbalance as a minority of vibration service calls [S5][S8]. Where balancing is part of a wider condition-monitoring programme, the same sensor chain feeds both the 1×-tracking phase reading and a broadband FFT, and the two together are how an experienced analyst separates unbalance from looseness, misalignment, or a developing bearing fault [S8][S9].
Standards, proving, and how phase accuracy is documented
ISO 21940-11 is the standard that defines balancing quality grade G (formerly Q) as a velocity limit in mm/s for a given rotor type, and the grade is only valid at the specific rotation speed the rotor will run at in service, not at an arbitrary balancing speed [S5][S6]. ISO 10816 is the companion standard for evaluating the resulting vibration severity on the machine bearings, and the two together define the acceptance envelope that any phase-accurate balancer is measured against [S5].
The proving artefact is a test rotor with a known test weight run through three measurement sequences (reference, left-plane weight, right-plane weight), and the resulting unbalance reduction ratio is the number that, together with the ±1° phase spec, is recorded on the machine's calibration certificate per SAE ARP4162 [S4]. For high-speed research balancers the same proving procedure is repeated at each operating speed, because phase error, sensor dynamics, and structural modes all change with rpm and the certificate is only valid at the speeds actually verified [S1][S4].
Resolution in g·mm, the companion spec to phase accuracy, is set by the smallest unbalance the instrument can read at a given rotor speed and correction radius, and it is the figure that drives the floor of achievable residual unbalance on a real shop floor. For readers cross-checking spec sheets, the dynamic balancing instrument resolution thresholds and what drives them breakdown is a useful adjacent reference on the same instrument family.
Trackable signals for the next reporting cycle: any ISO 21940-11 revision that re-tables grade G limits for high-speed PM rotors, and any OEM disclosure of URR versus phase accuracy on hard-bearing production balancers, since URR is the figure that actually drives throughput on a balancing cell, not the ±1° phase number alone.