AutomateDirectUSAMotion LibraryShop the parts catalog

Checklist

Identifying Salvage Robot Servo Motors Before Drive Selection

Spare parts, obsolescence and MRO sourcing Published 8 min read

Short answer

A salvage robot servomotor earns a drive only after three gates close: motor technology confirmed, winding ratings pinned down with the current type defined, and the feedback device identified and traced pin by pin with written supplier confirmation. Pole counts prove nothing, and an unsupportable geared resolver makes the motor a parts donor.

On this page
  1. The bench checklist
  2. What the nameplate decodes and what it never will
  3. Confirm the motor technology
  4. Settle the current rating before sizing the drive
  5. Worked example: screening the 1.41 kW motor
  6. The feedback device decides the project
  7. Pin counts prove nothing
  8. Screen the drive and get the answer in writing
  9. When to walk away
  10. Questions asked

A salvage robot servomotor earns a drive only after three gates close: the motor technology is confirmed, the winding ratings are pinned down with the current type defined, and the exact feedback device is identified and traced pin by pin with written supplier confirmation. The nameplate's 1.41 kW, 84 V, 11 A and 3000 rpm buy you a first screen and nothing more. Connector pole counts — 11 pins on one shell, 9 on the other — carry zero signal information. Geared resolvers on robot motors of this generation often are not standard resolver interfaces, and that is where most of these retrofits die. This checklist runs the bench work in order of cost, cheapest first, and ends with the walk-away criteria.

The bench checklist #

Work top to bottom. Each step is cheaper than the one after it, and any failure should end the project before money goes into a drive.

  • Photograph and transcribe the whole nameplate: 1.41 kW, 84 V, 11 A, 3000 rpm, and the type and serial strings.
  • Decode only what the standards define — protection, mounting, insulation class — and pass maker codes to the supplier verbatim.
  • Confirm the motor technology at the bench; three balanced phases means an AC synchronous servo that needs sinusoidal commutation.
  • Measure phase-to-phase resistance on all three pairs; a spread beyond a few percent (rule of thumb) fails the motor.
  • Measure insulation to the frame at 500 VDC with the feedback and temperature sensor disconnected; reject below 1 MΩ and investigate single-digit MΩ.
  • Settle whether 11 A is RMS or peak with the arithmetic screen, then the winding data.
  • Establish the voltage needed at 3000 rpm with a spin test and convert it into a drive bus-voltage class.
  • Open the rear cover and identify the feedback: resolver or encoder, geared to the shaft or not.
  • Trace every connector pin by continuity and resistance; assign nothing from pole count.
  • Screen the candidate drive on technology, bus class, continuous current and documented feedback support.
  • Require the supplier's written confirmation covering motor and feedback before ordering or applying power.

What the nameplate decodes and what it never will #

Three markings decode against standards. IP64 is the IEC 60034-5 protection code, built on IEC 60529: the 6 requires a dust-tight enclosure, the 4 protection against splashing water. That suits a dry machine floor and not a washdown, and an IP code never converts into a NEMA enclosure type, because the two systems test different things. IMB14 is an IEC 60034-7 mounting code for a face-flange-mounted machine: your bolt pattern lives on the flange face, not on feet. I.CL. F, on a German-built motor of this age, reads as insulation class F — the 155 °C thermal class of IEC 60085 — and it bounds the winding temperature, not the feedback electronics in the same housing.

What the nameplate never gives you: the definition of 84 V, the current type behind 11 A, pole count, winding resistance and inductance, and anything at all about the feedback. The type and serial strings are what a supplier uses to look the winding up, so quote them whole; a request built from 1.41 kW and 3000 rpm alone gets you a guess back.

Confirm the motor technology #

The plate says AC servomotor, which on a robot axis of this age means a permanent-magnet synchronous motor with three phases. Confirm it at the bench: three power conductors with near-equal phase-to-phase resistance, and a rotor that cogs as you rotate it by hand, magnets passing the stator teeth. Two heavy terminals and brush gear mean brushed DC and a different drive path; a motor with many identical small windings is a stepper. The decision rides on this: a synchronous servo runs only from a drive that commutates it from rotor feedback, so the feedback question can no longer be skipped.

Settle the current rating before sizing the drive #

Servo current comes in two currencies: AC RMS per phase on motor nameplates, and sometimes DC bus current on drive datasheets. Mixing them sizes the drive wrong by half or double. Run the arithmetic screen on the nameplate, taking 84 V as the phase-to-phase figure.

√3 × 84 V × 11 A = 1600 VA of apparent power against 1410 W of rated output, so combined power factor and efficiency sit at 1410 / 1600 = 0.88 — credible. Now read 11 A as peak instead: RMS becomes 11 / √2 = 7.78 A, and apparent power falls to √3 × 84 V × 7.78 A = 1131 VA, below the rated output, which would require better than perfect conversion. Commit to RMS per phase, and let the winding datasheet close the question. If a candidate drive quotes its output in DC bus current, get the conversion from its supplier rather than dividing by a factor yourself; the relationship depends on the drive's modulation.

Worked example: screening the 1.41 kW motor #

Calculated example with the nameplate values as inputs; no measured result is claimed.

Torque at the rated point: T [N·m] = 9550 × P[kW] / n[rpm] = 9550 × 1.41 / 3000 = 4.49 N·m. Plan around 4.5 N·m continuous at 3000 rpm for gearbox and belt sizing, and take the exact continuous torque from the datasheet, since servo makers sometimes rate it at a speed below maximum.

Bus voltage from a spin test: back-drive the shaft with a drill at a measured speed and read phase-to-phase voltage on a true-RMS meter. Ke = measured voltage divided by speed in krpm. Example inputs: 18.6 V at 1000 rpm gives Ke = 18.6 V/krpm line-to-line. At 3000 rpm the back-EMF is 3 × 18.6 V = 55.8 V RMS, or 55.8 × √2 = 78.9 V peak. A bus of 1.2 × 78.9 V ≈ 95 VDC — the 20 percent being a rule-of-thumb margin for winding resistance and inductance at rated current — puts such a motor on a low-voltage drive around the 100 VDC class, not behind the 325 VDC bus that 230 VAC mains rectify to. Your own spin test decides your number; the motor's Ke is an unknown until you measure it.

The feedback device decides the project #

Robot servomotors of this generation carry resolvers, and the rear cover on this type shows a gear train driving toothed feedback wheels — a geared resolver stack, not a bare resolver on the shaft. The distinction is decisive. A resolver reports angle as analog sine and cosine signals against a reference excitation supplied by the drive, and the two must agree on excitation frequency and voltage and on the transformation ratio. A gear between shaft and resolver breaks the assumption that feedback angle equals rotor angle, so the drive must know the ratio and phasing or it commutates at the wrong angle — and some robot resolvers use interfaces that are not standard resolver inputs at all. The teardown path through the housing, gearing and wiring is covered in identifying geared resolver and encoder feedback on old servos.

Commit to this gate: no traced pinout plus no documented drive support equals no power. A wrongly commutated first spin can throw the shaft violently, so the first power-up is unloaded, current-limited, with a hand on the disable.

Pin counts prove nothing #

The two connectors carry 11 and 9 poles, and that is the whole of what the counts tell you: how many conductors exist. Nothing about which is a phase, which is the brake, which is the temperature sensor, which is excitation. Robot motors commonly pack three phases and earth, a brake coil, a temperature sensor and the feedback signals into one or two shells, and the split between 11 and 9 pins varies by maker and year. Assign functions only by measurement: the three phase pairs show equal low resistance, the temperature sensor a stable reading orders of magnitude higher, the brake coil a resistance consistent with its rated voltage and power. The feedback pins make sense only against a traced interface. The discipline is the same as identifying unknown terminals on a NEMA size 1 motor starter: finish the map before anything is energized.

Screen the drive and get the answer in writing #

Once the feedback is pinned down, the drive question becomes the one in selecting a replacement servo drive for resolver motors. The candidate must clear four bars: sinusoidal commutation of a synchronous motor; a bus class that covers the measured back-EMF at 3000 rpm; continuous current at or above the motor's 11 A RMS per phase; and documented support for the exact feedback — excitation frequency and voltage, transformation ratio, gearing. Send the complete nameplate and your traced pinout, and require the confirmation in writing, naming the motor technology, the winding limits and the feedback interface. A supplier who will not put it in writing has answered the question.

When to walk away #

  • Phase resistances unbalanced beyond a few percent, or insulation below 1 MΩ at 500 VDC: the motor is a parts donor.
  • The feedback cannot be characterized — cover sealed, no drawing, no supplier decoding it: the motor is a parts donor. The unsupportable geared resolver is the most common dead end on salvage robot axes.
  • Two supplier queries without written confirmation: stop at two; a third query is hope, not engineering.
  • Any plan that involves finding out by applying power: the cost of a wrong-commutation crash exceeds every hour saved.

Characterized, a salvage servo is a cheap axis; uncharacterized, it is a set of magnets, a brake and two connectors. Buy the drive only after all three gates close.

Questions asked

Can a budget servo drive run a salvaged robot motor?

Sometimes, but the feedback is the usual blocker. Budget drives typically support a standard resolver or a standard encoder input, and geared or non-standard robot resolvers fall outside that list. Run the full identification first, then require the supplier's written answer for your exact motor and feedback. If no supplier will confirm it, no budget drive will safely run it.

Is the 11 A nameplate current RMS or peak?

Read it as RMS per phase and confirm against the winding data. Taking 84 V as phase-to-phase, three-phase apparent power is 1.732 × 84 V × 11 A, about 1600 VA, against 1410 W of rated output, which works at roughly 0.88 combined power factor and efficiency. Treating 11 A as peak would drop apparent power to about 1130 VA, below the rated output, which is impossible.

What do the 11-pin and 9-pin connectors tell me?

Nothing about signal assignment. Pin count never identifies functions, and robot motors mix phases, brake, temperature sensor and feedback in one shell. Assign functions only from continuity and resistance measurements: equal low resistance marks the three phase pairs, a much higher stable reading marks the temperature sensor, and the feedback pins must match a traced interface before anything is energized.

Can I replace an unsupported resolver with an encoder?

Yes, and it is the standard exit when no drive supports the original feedback. Mount an encoder on the free shaft end and choose a drive that commutates from it. You lose whatever multi-turn or gearing information the original stack carried, you add mechanical work, and the holding brake may already occupy the rear shaft.

Do I need to control the motor's holding brake?

Yes. Most robot axes use a spring-applied brake that releases only when its coil is powered, commonly at 24 V DC, verified from the coil data. If your drive cannot switch it, the axis stays locked and the drive trips on overcurrent. The brake holds position only; it is not a personnel safety device, which is a separate risk-assessment question.

Keep reading

  • How-to

    How to Identify Geared Resolver and Encoder Feedback on Old Servos

    Treat the resolver plus its geared 1:4 and 1:80 channels as one integrated coarse-and-fine, multi-turn position package: the resolver reads absolute angle within one shaft revolution, and the geared channels count revolutions above it. Identify each channel by resistance checks, ratio counting and waveform capture before choosing a replacement drive, because the resolver model number alone establishes neither commutation nor multi-turn position recovery.

    Sensors, encoders and machine vision on the machine

  • Selection guide

    Selecting a Replacement Servo Drive for Resolver Motors

    Screen the resolver interface before the power ratings: excitation frequency and voltage, transformation ratio between 0.2 and 1 V/V, pole relationship, pinout, and whether the resolver carries commutation. Then check the supply input range and continuous current against the motor nameplate. For a 1.3–1.5 kW axis the current screen passes easily; the resolver match decides the drive.

    Servo and stepper motion

  • Troubleshooting

    Brushed DC Servo Tach Dropout Diagnosis: Why It Jumps

    Jump at one shaft angle plus a speed-synchronous lope after tach work is a dropout in the tach's voltage, not an encoder or alignment fault. Prove it by spinning the feedback shaft and looking for a once-per-revolution notch in tach voltage; a cracked armature conductor needs a replacement armature or feedback unit, not another coat of insulating varnish.

    Servo and stepper motion