1. Problem Definition
On a SIMOREG DC MASTER 6RA70 drive controlling an intermediate DC motor used for paper support in a printing machine, two coupled symptoms appear in EMF (back-EMF) control mode with an analog speed/torque reference:
- The drive enters alarm state
A031(speed controller monitoring) during operation, especially after extended dwell periods in which the displayed torque symbol shows "–" (no active quadrant). - The diagnostic value
r021(actual torque) intermittently reads approximately-100 %, indicating the internal torque setpoint is being clamped or driven into the negative (motoring II / regenerating I) quadrant even when the process physically requires motoring torque only.
The operator initially set P165 = 1 to block negative torque. After that change, the drive stayed in the "–" quadrant much longer than before, current collapsed to near zero, and A031 reappeared more frequently. The problem is therefore not just an alarm-clearing exercise; it is a sign that the controller setpoint chain is no longer matched to the mechanical load characteristic of the paper-support section.
Axxx on the PMU and toggles the FAULT LED). It does not latch the power section off by itself, but repeated A031 events will degrade web quality long before they trip a hard fault. Always treat A031 as a symptom of controller mismatch, not a fault to suppress.2. SIMOREG 6RA70 Hardware and Firmware Context
The 6RA70 series is a fully digital three-phase, fully controlled B6 bridge DC converter rated for armature outputs up to 1200 A (chassis) and field outputs up to 30 A in the standard product. Firmware releases from V1.x through V3.x are in service; parameter numbering and alarm codes are stable across these releases, but a small number of monitoring parameters (notably in the P38x speed-controller block) were refined in V2.1 and later.
The reference parameter list for this troubleshooting case includes:
- P165 – Torque direction enable (limits torque setpoint to one polarity)
- P388 – Speed deviation threshold for n-controller monitoring
- P390 – Monitoring time before A031 is raised
- P150 – Torque setpoint smoothing time constant
- P151 – Speed setpoint filter time constant
- P155 / P156 – Positive / negative torque limit 1
- P157 / P158 – Positive / negative torque limit 2
- P160 – Torque limit source selection
- P082 – Analog input 2 function assignment (typically EMF feedback or auxiliary reference)
- r021 – Actual torque display (signed %, +100 % = P100/P101 scaling)
- r026 – Actual speed setpoint at controller input
- r027 – Actual speed (encoder or EMF-derived)
- r039 – Armature voltage actual value
Alarm A031 is described in Section 10 of the 6RA70 Operating Instructions; refer to the official Siemens diagnostic FAQ for the canonical alarm list and remedy chain: FAQ for 6RA70 DC MASTER and SIMOREG CM – Alarm Messages (PDF).
3. Drive and Motor Specifications
The intermediate paper-support section uses a separately excited DC motor. The nameplate values quoted in the service report are:
| Parameter | Value | Notes |
|---|---|---|
| Armature current | 56 A | Continuous, with appropriate overload margin |
| Armature voltage | 400 V DC | At base speed, field fully energized |
| Field current | 3.6 A | Separately excited, stabilized supply |
| Base speed | 3000 RPM | Field-weakening above this point if required |
| Drive control mode | EMF control | No tachogenerator; speed loop closed on back-EMF |
| Speed/torque reference | Analog ±10 V | Master tension controller in printing machine |
At 400 V armature and 56 A, the converter apparent armature rating is approximately 22.4 kVA DC, comfortably inside a 30 A-class 6RA70 (typical selection: 6RA7018-6DS22 or 6RA7025-6DS22 depending on overload). The 3.6 A field current fits inside the standard 5 A field module on the same unit. For the actual converter selection use:
kVA_DC = V_A × I_A / 1000 = 400 × 56 / 1000 = 22.4 kVA
Confirm that the chosen 6RA70 has P100 (armature current) and P101 (armature voltage) entered at the commissioning stage to match these nameplate values; otherwise the internal torque scaling (used to compute r021) will saturate before the motor reaches rated torque.
4. EMF Control Mode and Quadrant Behaviour
EMF (back-EMF) regulation closes the speed loop on the induced voltage computed from armature voltage minus IR drop, without a tachogenerator. It is a four-quadrant-capable mode by default: positive reference drives motoring (I-quadrant), negative reference drives motoring (III-quadrant), and the firing-angle logic automatically regenerates when the load overhauls the reference.
The torque symbol shown on the PMU is one of four states:
- I – positive torque, positive speed (motoring forward)
- II – positive torque, negative speed (regenerating reverse)
- III – negative torque, negative speed (motoring reverse)
- IV – negative torque, positive speed (regenerating forward)
- – – no active torque command (setpoint near zero, drive in coast/idle)
In a tension-controlled nip roll, only quadrants I and IV (forward pull with light hold-back) are physically meaningful when the paper is moving in the design direction. Quadrant II would imply the section is trying to push paper backwards, and quadrant III implies the section is dragging paper. Both produce slack-then-snap web behaviour, exactly the symptom reported before P165 was changed.
5. A031 Alarm Mechanism
A031 fires when the absolute difference between the speed setpoint input to the n-controller and the actual speed (EMF-derived or tachogenerator, depending on configuration) exceeds P388 for longer than P390 milliseconds. The two parameters are factory-defaulted to conservative values for unloaded bench testing; they almost always need site-specific adjustment.
| Parameter | Function | Default | Recommended for tension section |
|---|---|---|---|
| P388 | Speed deviation threshold (% of P101) | ~6–8 % | 12–20 % |
| P390 | Monitoring time (ms) | 200–500 ms | 1000–2000 ms |
Important: enlarging P388 or P390 only hides the symptom. A031 is the controller telling you the setpoint chain cannot be physically realized. The correct fix is to make the setpoint chain realizable, then bring P388/P390 back to their tight production limits.
6. Root Cause Analysis
Three independent factors combine to produce the observed behaviour:
- Master tension loop swings through zero. A PID master outputting ±10 V for tension trim will pass through 0 V during steady-state correction. In EMF control that maps to 0 RPM setpoint and 0 torque. The drive sits in the "–" symbol and armature current collapses.
- P165 = 1 blocks the negative half of the response. Once negative torque is forbidden, the controller cannot regenerate to hold back the section when the line momentarily tries to pull the section faster than the section's own torque would allow. The drive output becomes one-sided (positive only) but the load is two-sided; the integral term of the speed controller accumulates large wind-up, and on the next correction it slams into the positive torque limit.
-
Wind-up drives r021 to -100 % transiently. When the integrator is finally allowed to discharge, the controller walks the torque setpoint all the way through the negative rail before the limit clamps it.
r021reads -100 % during that discharge, even though the final applied value is +0 %. - Setpoint vs. realized EMF diverges. With the integrator wound up and the reference swinging around zero, the EMF-derived actual speed lags the setpoint by more than P388 for longer than P390, triggering A031.
The combination of factors 2 and 3 is why P165 = 1 made A031 worse, not better. It treated the symptom (negative torque) rather than the cause (improper tension-loop scaling and integrator behaviour).
7. Diagnostic Procedure
Execute the following steps in order. Record values to a commissioning sheet before changing any parameter.
-
Capture the alarm history. Use the PMU to scroll
r047(last fault code) andr048(alarm word). Confirm that A031 is the only active warning; if other alarms (A034 overcurrent, A035 overvoltage, A041 field loss) appear simultaneously, fix those first. -
Record the analog chain. With the machine at standstill, command a small positive reference and read
r026(speed setpoint),r027(actual speed),r039(armature voltage), andr021(torque). They should move together within P388 limits. -
Inject a step reference. Use the PMU parameter
P085or the analog test source to apply +5 %, then -5 %, then +5 % setpoint steps while watching the response on DriveMonitor or the PMU. The torque polarity should follow the polarity of the reference; if r021 shows -100 % on a positive step, the P155/P156 torque limit signs are inverted. -
Verify the EMF feedback sign. Disconnect the analog reference. Run the motor briefly with a small positive torque limit (
P155 = 5 %,P156 = 5 %) usingP160in test mode. If the motor runs the wrong way, swapP082(analog input 2) sign or invert the EMF sign bit (P706depending on firmware). - Capture the tension loop output range. On the printing machine, run a known web speed and read the tension PID output to the 6RA70 with a DMM. Most masters are scaled to ±10 V = ±10 % trim around the line speed. If the swing is larger than ±10 %, the tension controller is over-tuned and is itself driving the swings.
- Read P155/P156/P157/P158. Default values often leave the negative torque limit at -100 %. This is what allows r021 to read -100 %.
8. Parameter Reference – Corrective Values
The following table summarizes the recommended changes for a tension-only paper-support section in EMF control. All values assume commissioning against the 56 A / 400 V / 3.6 A motor in Section 3.
| Parameter | Description | Default | Recommended | Reason |
|---|---|---|---|---|
| P082 | Analog input 2 function | EMF feedback | Verify sign | EMF sign wrong → r021 inverts on positive reference |
| P150 | Torque setpoint smoothing | 0 ms | 50–150 ms | Filter master tension loop ripple |
| P151 | Speed setpoint filter | 0 ms | 30–80 ms | Reduce n-controller bandwidth so P388 is not violated on every line disturbance |
| P155 | Positive torque limit 1 | 100 % | 100 % | Keep; allows full motoring capacity |
| P156 | Negative torque limit 1 | -100 % | 0 % | Negative-torque clamp at zero – replaces P165 = 1 with proper limit, leaves n-controller functional |
| P160 | Torque limit source selection | Fixed | Fixed (P155/P156) | Do not switch to analog torque limit unless the master explicitly demands it |
| P165 | Torque direction enable | 0 | 0 | Restore to 0; clamping is now done by P156 = 0, not by P165 |
| P388 | n-controller deviation threshold | ~6 % | 12 % (start), 8 % (production) | Tolerate EMF ripple during tension trim |
| P390 | A031 monitoring time | 200 ms | 1500 ms (start), 500 ms (production) | Allow momentary deviation without spurious alarm |
| P401 / P402 | Speed controller Kp / Tn | Site default | Reduce Kp by 30 %, increase Tn by 50 % | Lower bandwidth prevents integrator wind-up when reference dithers around zero |
The key change is P156 = 0 % instead of P165 = 1. The P156 = 0 approach:
- Allows the n-controller integrator to function normally (no direction lockout).
- Clamps the applied torque setpoint at zero on the negative side, achieving the same physical result as P165 = 1.
- Prevents r021 from walking to -100 % during integrator discharge because the limit is applied at the output stage, not the input stage.
9. Step-by-Step Corrective Procedure
- Place the drive in commissioning / parameter mode via the PMU (S1 switch on CUD board to position 1 if DriveMonitor is not available).
- Navigate to
P156and enter0.0 %. Confirm the parameter is accepted (PMU returns no error). - Navigate to
P165and enter0(restore two-quadrant torque direction). - Adjust
P150to0.100 sandP151to0.050 s. - Adjust
P388to12.0 %andP390to1.500 s. - Slowly reduce
P401(Kp) by 30 % from current value; increaseP402(Tn) by 50 %. - Save parameters (
P052 = 1to save RAM to EEPROM if not done automatically). - Run the paper at nominal line speed. Monitor
r021continuously on DriveMonitor; torque should remain in the 0 % to +25 % range with short excursions, never negative. - After 30 minutes of stable running, tighten
P388to8 %andP390to0.500 sfor production sensitivity. - Re-tighten the n-controller Kp/Tn if mechanical response feels sluggish, in 10 % steps, alternating between Kp and Tn.
10. Verification Tests
Run the following four tests after the parameter changes. The drive must pass all four before being declared production-ready.
| Test | Procedure | Pass criterion |
|---|---|---|
| T1 – Static | Apply 0 V reference for 5 minutes. | r021 stays at 0.0 % ± 0.5 %; no A031. |
| T2 – Positive step | Step reference from 0 % to +8 %. | Torque ramps to +8 %, settles within 1 s, no overshoot below 0 %. |
| T3 – Disturbance | Simulate a 10 % web tension step from upstream master. | r021 stays > 0 %; transient response < 2 s; no A031. |
| T4 – Sustained line run | Run full production speed for 1 hour with random tension trim. | Zero A031 events; armature current (r038 or equivalent) > 5 % of rated at all times when paper is moving. |
If T4 still raises A031, the master tension PID is too aggressive. Adjust the master's proportional band upward or its integral time upward; do not loosen P388/P390 again.
11. Engineering Notes and Edge Cases
-
Field weakening above 3000 RPM: This motor is rated at 3000 RPM base speed. If the section ever runs above that, ensure field-weakening entry (
P103) and field current ramp (P257) are correct. Above base speed, armature voltage is clamped at P101 and torque falls off; the EMF control loop becomes more sensitive and A031 can reappear for an unrelated reason. -
Three-phase supply quality: A031 can also be raised by heavy line-side distortion. If multiple 6RA70 units share a transformer with non-linear loads, check the DC link ripple with a scope on
r039; values above 8 % of P101 indicate the armature reaction loop is fighting line noise. - Cooling and derating: At 56 A continuous with no overload, the 6RA70 runs well below its thermal limit. Confirm fan operation (if fitted) and that the cabinet ambient stays under 40 °C; otherwise the I²t foldback can momentarily reduce available current and produce apparent A031 events during long dwells.
-
Encoder fallback: If A031 persists after the corrective procedure, the EMF-derived speed signal is too noisy for the application. Switch the drive to tachogenerator mode (
P083assignment, depending on firmware) or fit a pulse encoder on the motor and assign it via the encoder option board. - Firmware version check: The P388/P390 behaviour was tightened in firmware V2.1 and again in V3.0. Before assuming a parameter mismatch, upgrade the CUD firmware to the latest service pack listed on the Siemens 6RA70 product page. Re-commission after every firmware change.
r040 or equivalent DC-link voltage display) before opening the drive cabinet. The 6RA70 contains line-side capacitors that retain charge for several minutes after the line is disconnected. Refer to the Operating Instructions, Section 1, for the prescribed discharge time.12. Summary of the Corrective Strategy
Replace P165 = 1 (which disables the n-controller) with P156 = 0 % (which clamps only the torque output). Restore two-quadrant torque direction. Add modest setpoint filtering via P150 and P151. Set P388 and P390 to site-realistic values for a tension section, and tune the speed controller Kp/Tn to a lower bandwidth than factory default. Verify with the four-test sequence above. This combination eliminates the root cause of A031, prevents r021 from reading -100 %, and keeps the paper-support drive in continuous positive torque as required by the process.
For canonical alarm definitions and a wider fault-diagnostics reference, consult the official Siemens document: FAQ for 6RA70 DC MASTER and SIMOREG CM – Alarm Messages.
What does alarm A031 mean on a Siemens 6RA70 drive?
A031 is the speed-controller monitoring alarm. It triggers when the difference between the speed setpoint at the n-controller input and the actual measured speed (EMF-derived or tachogenerator) exceeds parameter P388 for longer than the time set in P390. It indicates a controller mismatch, not a hardware fault, and should be resolved by tuning the setpoint chain rather than widening P388/P390.
Why does r021 show -100 % torque even though P165 is set to 1?
P165 = 1 disables the torque direction at the controller input, but r021 displays the internal torque setpoint before the direction lockout is fully applied during integrator discharge. The negative reading is the controller unwinding its integral term. Replace P165 = 1 with P156 = 0 % (negative torque limit clamped at zero) to prevent the negative excursion from appearing at the output stage and on r021.
Which parameters control the A031 monitoring thresholds on the 6RA70?
P388 sets the speed-deviation threshold (as a percentage of armature voltage P101) and P390 sets the monitoring time in milliseconds. For a tension-controlled paper section, start with P388 = 12 % and P390 = 1500 ms, then tighten to P388 = 8 % and P390 = 500 ms once stable operation is confirmed.
Can I use EMF control for a paper tension drive without a tachogenerator?
Yes. EMF control closes the speed loop on the computed back-EMF (armature voltage minus IR drop). It is adequate for steady tension sections where web speed changes slowly. For applications with rapid line-speed transients or where A031 recurs despite correct P388/P390 settings, fit a pulse encoder or tachogenerator and assign it via the encoder option board.
What is the correct way to make a 6RA70 torque-positive-only for a tension section?
Set P155 (positive torque limit 1) to 100 %, set P156 (negative torque limit 1) to 0 %, and leave P165 = 0. This clamps the torque setpoint at zero on the negative side while keeping the speed controller fully functional, which is the physically correct approach for a paper-support section that must produce only motoring torque.