Independent CNC repair and technical reference Contact

MR-J4-A command and feedback counts disagree: electronic gearing and the five-digit display

Published Updated

P counts commands before electronic gearing; C counts motor encoder feedback. Check the active ratio, counter origins and five-digit display limits before treating their difference as droop E or lost pulses.

Contents
  1. 1. Four displays answer different questions
  2. 2. PA06 and PA07 alone do not establish the active ratio
  3. 3. A count example with explicit assumptions
  4. 4. Five digits hide higher places; illuminated dots matter
  5. 5. Form a defensible maintenance conclusion

Do unequal cumulative command pulses P and cumulative feedback pulses C on an MR-J4-A panel prove lost pulses or an encoder fault? First establish where each count is taken and its scale. Based on Mitsubishi Electric SH(NA)030107ENG-V (November 2025), this article covers MR-J4-_A_(-RJ) models rated 100 W or more, rotary servo motors and semi-closed-loop position control. It interprets existing records obtained safely, without requesting new motion commands.

1. Four displays answer different questions

SymbolManual definition and unitKey limitation
PCumulative command pulses, pulsePosition command input count, before multiplication by electronic gear CMX/CDV
CCumulative feedback pulses, pulseMotor encoder feedback count; do not assume the same scale as P
EDroop pulses, pulseDeviation counter value, displayed in encoder pulse units
nCommand pulse frequency, kpulse/sInput frequency before electronic gearing, not directly motor speed

These definitions come from §4.5.3 (4-27). P represents accumulated input commands, n their frequency, C motor encoder feedback and E internal following deviation. Two readings labeled pulse do not necessarily support direct subtraction. In particular, P−C is not a valid expression for actual E.

2. PA06 and PA07 alone do not establish the active ratio

Section 5.2.1 (5-16) defines PA06 as electronic gear numerator CMX and PA07 as denominator CDV. The same page shows the electronic gear selection in PA21: 0 selects ordinary electronic gearing, 1 uses command input pulses per revolution, and 2 and 3 select different compatibility modes. These numbers refer to the leftmost hexadecimal digit of PA21, not its whole value. The page footnote explicitly limits mode 3 (J2S compatibility) to amplifier software version B3 or later; verify applicability to the full amplifier model and software version. Reading PA06=4 and PA07=1 is therefore insufficient to establish an effective ratio of 4.

Also check which numerator was selected when recording. Section 3.6.1(4) (3-43) assigns CM2/CM1 as follows: 0/0 selects PA06, 0/1 PC32, 1/0 PC33 and 1/1 PC34. The manual says the numerator changes as soon as the selection signals change. On machines using this function, preserve the contemporaneous CM1/CM2 states and corresponding parameters, rather than just PA06 after maintenance. Do not switch these signals or change gearing merely to make the two displays agree.

3. A count example with explicit assumptions

This is a teaching example, not a field case or test result. Assume ordinary electronic gearing is confirmed in PA21, CM1/CM2 are both off, PA06=4 and PA07=1. Throughout the observation interval there is no ratio switching, counter reset, direction reversal or display truncation, and both records share the same start and end times.

If P increases by 2000 input pulses, the command increment after gearing is 2000×4÷1=8000 encoder pulse units. If C also increases by 8000, the unequal raw increments do not mean the motor “moved 6000 pulses too far.” Matching increments also do not prove zero following deviation throughout the move. Examine E and relevant operating records on a common time base for intermediate deviation; two endpoint totals cannot reconstruct the whole trajectory.

Even after ratio conversion, an arbitrary difference between cumulative readings is not E. The two displays may have different reset origins, and manually changing pages is not synchronous sampling. Distinguish “matched count increments over the same interval” from “synchronized deviation recording” in the maintenance report.

4. Five digits hide higher places; illuminated dots matter

Section 4.5.3 says P, C and E can count values exceeding ±99999, but the five-digit panel shows only the lower five digits. A sudden smaller reading therefore does not establish a reset or reveal complete travel. For example, a positive total of 99990 increased by 20 becomes 100010, whose lower five digits are 00010. Calculating 10−99990 would mistake a positive increment of 20 for a large reverse change. This is an arithmetic illustration of the display rule. With only two photographs and unknown intervening motion, the number of crossings of the five-digit boundary cannot be determined.

The manual uses illuminated decimal points in the upper four digits for a negative C value, reverse-direction counting in P and reverse pulses in E. Preserve all LED segments and dots in photographs, rather than transcribing only five numerals. Do not casually interpret these indicators as ordinary decimal separators.

Resetting can also destroy evidence. The manual says SET can reset the P and C display values to zero; §4.5.3 also uses SET to open the selected data display, so the button function depends on the current screen. This article gives no button sequence. Establish the screen state before reading and do not reset original records for convenient comparison. If a reset occurred during recording, identify the counter and time. Counts before and after it cannot be treated as a continuous series with one origin.

5. Form a defensible maintenance conclusion

  • P and C have different proportions: verify PA21, the active numerator and PA07 before comparing increments over the same interval. A ratio difference alone does not justify encoder replacement.
  • P or C suddenly becomes smaller: distinguish lower-five-digit display, sign indicators and recorded reset events. If uncertain, report “higher digits or counter origin unknown,” not a confirmed reversal.
  • P changes but C appears unchanged: first account for different observation times and invisible higher digits, then investigate using existing E, operating-state and alarm records. One count cannot identify an amplifier, controller or mechanical component as faulty.
  • E remains abnormal: preserve raw values, signs, operating phase and complete alarms. Do not alter gearing, clear counts or reset equipment to remove differences from the evidence.

A useful handover includes full amplifier and motor models, software version, control mode, effective PA21/PA06/PA07 and relevant PC32–PC34 values, CM1/CM2 states when numerator switching is used, time-stamped raw P/C/E/n records, counter-reset events, known display-boundary crossings and operating phases. If the ratio or origin is missing, state that quantitative comparison is unavailable. These counts cannot establish table-position accuracy either: semi-closed-loop motor feedback is not an independent load-side measurement.

Safety limits: use existing records and information that can be read safely under the approved site procedure. No trial motion, live wiring, parameter writing or protection bypass is instructed. Zero P, C or E does not establish an isolated power circuit or released mechanical energy. Qualified personnel must follow the machine procedure and manual safety instructions for maintenance isolation.

Source: MR-J4-_A_(-RJ)/MR-J4-03A6(-RJ) Servo Amplifier Instruction Manual, SH(NA)030107ENG-V (November 2025, English original): §4.5.3, 4-23 and 4-27; §5.2.1, PA06/PA07, 5-16; §3.6.1(4), 3-43; §5.2.3, PC32–PC34, 5-50; Appendix 7.3.1, App.-29; Safety Instructions A-2. References are printed page numbers. Examples and diagnostic decisions are editorial analysis based on these definitions, not manufacturer repair cases. Do not directly extend the conclusions to MR-J4-B, MR-J5, linear motors or fully closed-loop load-side counts.

Related technical articles

MR-J4-A analog traces: tenfold scaling differences and the 5 V zero pointMR-J4-A: why INP can turn on while the axis is still movingMR-J4-A load monitoring: effective, peak and regenerative ratios are not interchangeable