# Balancing machine retrofit: old mechanics, new measurement system > You have a thirty-year-old balancing machine on the shop floor. The bed is sound, the supports are fine, the drive turns, but the instrument shows two numbers on analogue dials and cannot store a result or print a report. Replacing the whole machine is expensive and unnecessary. Below we go through how to fit this mechanics with a Balanset-1A OEM measurement core, what you need to check first, and where the method will not work. **In short:** A balancing machine retrofit means this: you keep the bed, the supports and the drive, and replace the entire measurement side with a modern one. The Balanset-1A OEM measurement core takes the place of the analogue instrument: two accelerometers on the supports, a laser phase sensor with a reflective marker, a two-channel USB module, and software on a laptop. The machine starts calculating the correction in one and two planes, storing influence coefficients (the machine's remembered response to a trial weight), splitting the weight across fixed positions, calculating a drilling correction, and printing a report. The Balanset-1A manual describes exactly this use: the instrument works as a measuring system for soft-bearing (above-resonance) balancing machines. Source: https://axiline.pt/en/articles/balancing-machine-retrofit/ Publisher: AXILINE · Vila Nova de Gaia, Portugal · +351 931 831 229 · axilinegeral@gmail.com ## What an OEM kit is, and how it differs from the full kit The full Balanset-1A kit is built for site visits. The case holds a two-channel USB module, two accelerometers, a laser phase sensor, a magnetic stand for it, electronic scales, the manual, and a flash drive with the software. You arrive on site, fit the sensors to the bearing housings, and balance the rotor in its own bearings. The OEM kit solves a different problem. It is not a stand-alone instrument in a case, but a measurement core built into your own structure. The case, the magnetic stand and the scales are mostly redundant here: the sensors sit permanently on the machine, the tachometer is fixed to a bracket, and the laptop or panel PC lives in a cabinet. We agree the exact OEM delivery contents for your specific rig, so the precise list is set by your design, not by a catalogue. ### Two-channel USB module Preamplifiers, integrators and an ADC in a single housing. Power and the computer link both run over USB, so no separate power supply is needed. Inputs X1 and X2 take the vibration sensors, input X3 takes the phase sensor. ### Two accelerometers Single-axis, housing up to 25×25×20 mm, mass up to 40 g. Nominal sensitivity of about 20–30 mV/(mm/s) is entered in the software and only changes when a sensor is replaced. On a rig, they are mounted rigidly on the supports and stay there. ### Laser phase sensor (tachometer) Provides the speed and the reference phase — the angular mark the instrument uses to work out where the unbalance sits on the rotor. It works off a reflective marker on the rotor or the arbor. Range 100–100,000 RPM, phase measurement error ±1°. Without the marker, the system cannot pick out the running-speed component — the vibration at the rotation frequency, which is what balancing actually works with. ### Windows software Calculates the solution for one and two planes, displays overall vibration and the 1x running-speed component with phase, plots the FFT spectrum and a polar diagram, and stores influence coefficients and an archive of reports. ### What you provide The frame or bed, the supports, the drive and the guarding, mounting pads for the sensors, a bracket for the tachometer, the computer and the control cabinet. The mechanical side stays with you, or with our joint design work. ## The mechanics are sound, the measurement side is outdated A balancing machine ages unevenly. The cast bed and supports outlive several generations of electronics. Rollers and bearings get replaced on schedule. But the measurement side fails first, and it cannot be repaired: the components are no longer made, nobody remembers the circuits, and the documentation went missing along with the previous chief mechanic. Here is what you end up with. The operator chases a needle on an analogue dial and judges the angle by eye. The result is not saved anywhere except a notebook. Two operators on the same rotor come up with different numbers, and there is nothing to argue the point with. The customer asks for a report with the residual unbalance, and you have neither an archive nor a printout. Replacing the whole machine costs as much as several machines and needs space, a foundation and approvals. On top of that, you pay a second time for the very mechanics that already work. A retrofit turns that logic around: you keep the hardware and replace only what has actually gone out of date. - No software and no calculation: the angle and mass are worked out by hand from nomograms or from memory. - No archive: the result lives in a notebook, and you cannot pull up the history for a specific rotor. - No reports: there is nothing to hand the customer for sign-off. - Poor repeatability: the spread between operators and between runs is not controlled. - No influence coefficients: every rotor, even a familiar one, gets balanced from scratch. - No spectrum: there is no way to tell whether 1x really dominates, or something else is getting in the way. ## Retrofit: what actually changes The idea behind a retrofit is simple. The mechanics produce the movement at the supports; the electronics measure that movement. You replace the second half without touching the first. The Balanset-1A manual describes exactly this use: the instrument with accelerometers works either for on-site balancing in a machine's own bearings, or as a measuring system for soft-bearing (above-resonance) balancing machines. 1. **Sensors on the supports** — Two accelerometers go on the machine's supports, close to the point the load passes through. The mounting is rigid: a stud on a machined pad is better than a magnet if the sensor is staying put. You choose the measurement direction by the highest vibration, usually horizontal-radial, and do not change it after that. 2. **Laser tachometer and marker** — You mount the phase sensor on a bracket so the beam lands reliably on the reflective marker. The marker is stuck to the shaft, the pulley or the arbor. A steady pulse from the marker sets both the speed and the reference phase, from which everything else is calculated. 3. **Two-channel module** — The module records both channels at once, so you see both supports in a single run. It is powered over USB. If the mains supply in the shop is noisy, the manual recommends running from the laptop's battery. 4. **Software on a laptop or panel PC** — The Windows software displays overall vibration velocity, the 1x running-speed component with phase, speed, the time waveform and the spectrum. All the balancing and the printed report come out of the same software. 5. **Setup and verification on a reference rotor** — You enter the sensors' actual sensitivity coefficients and set the plane geometry and radii. Then you take a rotor with a known weight and check that the machine repeats the result from run to run. This is your zero point for all the work that follows. > Sensor fixing deserves a separate note. Working from stored influence coefficients requires the vibration sensor and the phase sensor to sit exactly as they did during the first balancing run. On a site visit, that comes down to operator discipline. On a rig, the sensors are rigidly fixed and never repositioned, so the stored-coefficients mode delivers its full benefit: a familiar rotor gets balanced in a single run, with no trial weight. ## What you get after the retrofit The list below is not about interface convenience. Every item removes a specific operation the operator currently does by hand, or does not do at all. ### One and two planes A single-plane scheme for disc-shaped rotors, two-plane for elongated ones. The two-plane calculation accounts for how the planes affect each other, rather than treating them separately. ### Influence coefficients in the archive Coefficients obtained from trial runs are saved under the rotor's name. Next time, the same rotor runs in stored-coefficients mode: no trial weight is needed. ### Fixed positions and drilling The software splits the weight across the actual fixing points — blades or bolt holes, for example — and gives you a position number instead of an angle. Position one, Z1, matches where the trial weight went, and the numbering runs in the direction of rotation. If you are removing material instead, you switch the calculation to removal, and the angle turns 180° automatically. ### Trim balancing Missed tolerance on the first pass? The software works out how much mass to add to the weights already fitted. The successive-approximation method cancels out errors in how the weights were placed. ### Tolerance against the G grades The instrument calculates the permissible residual unbalance against the G grades — balance quality grades: the lower the grade number, the finer the tolerance. The manual references ISO 1940; ISO 21940-11 applies now. Check which part and edition of the standard applies to your machine. ### Archive, reports, polar diagram Results, vibration values, time waveforms and spectra are all saved. The balancing report is generated and can be edited in the built-in editor. The polar diagram shows the vector before and after. ### Arbor eccentricity calculation Index balancing separates the part's own unbalance from arbor run-out. For rigs where the part sits on a work-holding arbor, this removes a systematic error. ### Recalculating to other planes If the design does not allow a weight where the calculation asks for one, the software recalculates the correction masses for the planes that are actually accessible. ## What to check before the retrofit Before ordering the OEM kit, go through this list with a wrench in hand. Half of the problems that come up during installation have nothing to do with electronics — they come from having nowhere to fix a sensor, or a speed that will not hold steady. - [x] Access to the supports: can you reach both of the machine's supports with the guarding closed. - [x] Mounting pads for the sensors: is there a clean, flat surface for a stud or a magnet, and does the cable get in the way. - [x] Space for the tachometer: where the bracket will sit, whether the beam reaches the marker, and whether tooling blocks the view of it. - [x] The marker on the rotor: where the reflective marker will go, and whether it will survive the working cycle. - [x] Drive and speed: does the drive hold a steady rotation frequency, is there any belt slip, and does the speed fall within the 100–100,000 RPM range. - [x] Resonance: does the rig's operating speed coincide with a natural frequency of the frame or the supports. In resonance, the 1x phase drifts and the result does not repeat. - [x] Vibration level: the measurement range for RMS (root mean square) vibration velocity at 1x is 0.02–80 mm/s, and the RMS measurement band per the manual is 5–200 Hz. - [x] Power and workspace: a socket for the laptop, cable protection, space for the computer. If mains quality is poor, run from the battery. - [x] Conditions: temperature from +5 to +50 °C, humidity below 85% with no condensation, no strong electromagnetic fields or shocks. - [x] Safety: guarding, secure fixing of the trial and correction weights, an emergency stop for the drive. > If overall vibration on the rig is several times higher than the 1x running-speed component, a retrofit will not solve your problem. Look at the mechanics first: play in the supports, worn-out rollers, loose fasteners, rubbing. Balancing only brings down 1x, and the new instrument will simply show that more honestly than the old one did. ## A rig from scratch: OEM as the measurement core Sometimes there is no old machine, but there is a steady flow of rotors to balance. In that case the OEM kit becomes the core of a new rig, and the mechanics get designed around it. The key decision comes at the first step: a hard-bearing (below-resonance) or a soft-bearing (above-resonance) support scheme. The Balanset-1A measures support vibration with accelerometers. That is exactly the quantity a soft-bearing scheme works with, and it is what the manual lists as the standard application. For a hard, below-resonance scheme, where force sensors are built into the support's rigid link, the situation is different: the support barely moves, and there is nothing left to measure. We are saying this plainly so you do not design a hard-bearing rig around accelerometers and then get an unwelcome surprise. - One plane or two: work it out from the geometry of the parts that will go on the rig. An L/D ratio (rotor length to diameter) below roughly 0.5 allows single-plane work; an elongated rotor needs two. - Supports: type (roller, prism, or plain-bearing), the range of journal diameters, adjustment of the distance between supports. - Drive: belt, cardan shaft, or through a coupling. A belt drive is simpler, but watch for slip and speed stability. - Frame: stiffness and natural frequencies. Work these out so the operating range does not sit on a peak of the frame's response. - Guarding and interlocks: access to the correction planes with the rotor stopped. - Build space for the sensors and the tachometer into the drawing from the start, not after assembly. - The operator's workstation: computer, cable routing, printing reports. | Feature | Soft-bearing (above-resonance) scheme | Hard-bearing (below-resonance) scheme | | --- | --- | --- | | Operating speed | Above the supports' natural frequency | Below the supports' natural frequency | | What is measured | Displacement and velocity of the support's motion | Force in the support's rigid link | | Support stiffness | Low, the support can move | High, the support barely moves | | Setup for a rotor type | Needs a trial weight and influence coefficients for every new rotor type | Machine calibration, then calculation from geometry and mass | | Sensitivity | High, works well at small residual unbalances | Lower, but less dependent on rotor mass | | Passing through resonance | Start-up passes through resonance; you need to settle at a steady speed | Resonance sits above the operating range | | Balanset-1A OEM | Standard application per the manual | Needs separate verification: accelerometers need the support to move | ## OEM or the full kit: how to choose There is one more scenario worth keeping in mind. Balancing in a machine's own bearings accounts for the real supports, the foundation, the operating temperature and the machine's own regime. A rig does not reproduce any of that. So a rig makes sense where you balance parts before assembly: impellers, pulleys, fan wheels, screw conveyors, motor rotors after rewinding. For machines already sitting on their foundation, on-site balancing remains the main method. | Your situation | What to get | Why | | --- | --- | --- | | On-site balancing at customers, different machines | The full kit | The case, the magnetic stand and the scales are needed every day. Sensors move from machine to machine. | | Retrofitting an existing balancing machine | The OEM kit | The sensors and the tachometer are fixed permanently. The case and the stand are not needed, and rigid sensor mounting unlocks working from stored coefficients. | | Building a new rig for a production part | The OEM kit | The core is built into your own structure and cabinet. You design the mechanics around your own range of rotors. | | A repair shop: a rig plus visits around the plant | The full kit and OEM separately | One instrument travels, the other lives on the rig. That way you never reposition sensors and never lose repeatability. | | Still undecided whether you need a rig | The full kit | Start with balancing in the machine's own bearings. You will build up data across your range and see whether a rig is justified. | ## Metrology and paperwork: what we can honestly promise We are being deliberately careful here, because the wording decides whether your customer will accept the report. The software reports "within tolerance" when the residual 1x running-speed component has dropped below the target you entered yourself. That is not confirmation of a G grade under ISO 21940-11, and not an assessment of the machine under ISO 20816. Three criteria exist side by side and do not substitute for one another: residual 1x against your own setpoint, residual unbalance in g·mm/kg against the G grades, and overall vibration velocity in mm/s RMS against zones A–D (machine-condition zones running from "good" to "unacceptable"). Calibrating the measurement channel after a retrofit comes down to two actions: entering the sensors' actual sensitivity coefficients, and checking the machine against a reference rotor with a known weight. You set the frequency of that check in your own procedure. Metrological status for mandatory verification depends on the requirements of your metrology service, your industry and national legislation. We do not promise automatic compliance, and we recommend clarifying this question before you order. > A note on frequency band. The manual states an RMS vibration-velocity measurement range of 5–200 Hz. The condition-assessment criteria in ISO 20816 usually call for a broadband measurement of 10–1000 Hz. If contractual acceptance is tied to that band, check applicability beforehand and record the measurement points, the band, the operating condition, the support type, and the exact part and edition of the standard in the report. ## How AXILINE's engineers can help The Balanset instruments are designed and manufactured by the same engineers who use them for on-site balancing themselves. That is a real advantage when you are choosing a configuration: questions like "where does the sensor go on a support like this" or "is one plane enough for this impeller" get answered directly, not filtered through a salesperson. For a retrofit, we help you go through your setup piece by piece: check whether the machine's mechanics suit measuring support vibration, choose the sensors and how to mount them, settle on the number of planes and the radii, configure the software, enter the sensitivity coefficients, and run the first balancing job on a reference rotor together with your operator. Training your staff is part of the same work: the software is built for users with no specialist training in vibration diagnostics, but an operator's habits form in the first few days. Where to start the conversation. Send us the machine's type and layout, the range of rotor masses and diameters, the operating speed, the drive type, and photos of the supports. That is enough for us to tell you honestly whether your machine suits a retrofit, or whether it is simpler to build a new rig around the OEM core. If your machine is not a good fit for a retrofit, we will say so, rather than sell you a kit for a job it will not do. There is a separate option for anyone still on the fence: start with on-site balancing. We come and balance your rotors in their own bearings, you see the instrument working on your own range of parts, and you make the decision about a rig with real numbers in hand. Sources: [Balanset-1A operation manual](https://vibromera.eu/balanset-1a-operation-manual/) · [Balanset-1A manufacturer specification](https://vibromera.eu/product/balanset-1/) · [ISO 21940-11:2016](https://www.iso.org/standard/54074.html) ## Frequently asked questions **Will the Balanset-1A suit my balancing machine?** Check three things. First, the machine's supports need to move noticeably under unbalance, because accelerometers measure the support's motion, not force in a rigid link. Soft-bearing (above-resonance) machines fit as standard — that is stated in the manual. Second, the speed needs to sit in the 100–100,000 RPM range and stay stable; the drive must not "wander." Third, you need access to the supports for the sensor pads and to the shaft for the reflective marker. If the machine is hard-bearing and the supports barely move, the honest answer is: run a trial measurement on your rotor first, then decide. Sometimes the answer is converting the supports to a soft scheme; sometimes it is simpler to build a separate rig. **What sensors are needed, and can I fit my own?** The kit includes two single-axis accelerometers, housing up to 25×25×20 mm, mass up to 40 g, with nominal sensitivity of about 20–30 mV/(mm/s). The signal path is matched to them: the preamplifiers, integrators and ADC in the module are designed around these sensors. The manual does not claim universal compatibility with any IEPE/ICP accelerometer, so we recommend the standard sensors. The sensitivity coefficient is entered in the Settings window and only changes when a sensor is replaced. Connections: X1 and X2 for the vibration sensors, X3 for the laser phase sensor. **How many planes does the system support?** One or two. For disc-shaped rotors with L/D below roughly 0.5, one plane is usually enough. Elongated rotors need two, or couple unbalance — a tilt in the mass axis that a single weight cannot remove — will be left behind, and the vibration at the second support will not go away. One plane needs two runs: the baseline and one trial. Two planes need three runs: the baseline and two trials, one per plane. Two-channel acquisition covers both supports at the same time. **Does the machine's mechanics need any rework?** Often, but only a little. The typical scope: machining sensor pads on the supports, fitting a bracket for the laser sensor, applying a reflective marker or fixing a replaceable target, sorting out the drive and the belt, and checking the torque on the bed and the support-fixing bolts. If the machine's mechanics are worn out, a retrofit will not cure that. Worn rollers, play in the supports, a wandering drive all produce readings that will not repeat, and the new instrument will simply show you that scatter honestly instead of a tidy number. **What about verification and calibration after the retrofit?** Keep the two questions separate. Calibrating the measurement channel means entering the sensors' actual sensitivity coefficients and checking the machine against a reference rotor with a known weight. You do that yourself when commissioning the machine, and repeat it periodically. Metrological status for mandatory verification depends on the requirements of your metrology service, your industry and national legislation, and we do not promise automatic compliance. Clarify this before ordering, because the answer decides whether the reports will be accepted for contractual sign-off. **Can I keep the old drive and the old speed?** Usually yes, if the speed is stable and falls within the instrument's operating range. A belt drive is acceptable, but watch for slip: speed jumps break the phase and spoil the calculation. A VFD is more convenient, because it lets you return to exactly the same speed on every run and steer around the resonance zone. Balance at a steady speed, and always the same one, if you want to work from stored influence coefficients.