Siemens S5-135U CPU 921 Resolving AS Interface Not Loaded CP RAM

1. Problem Overview

A Siemens SIMATIC S5-135U multi-CPU rack installed on a Leybold AG polyester rolling machine reports the following message on the programming device (PG) when an engineer attempts to open Program Status on CPU 921:

AS Interface not loaded / CP RAM not plugged in

The CPU continues to run, status variables can still be observed on the PG, and the parallel CPU 922 in the same rack displays Program Status of the ladder logic and status variables without error. The application must continue to control three DC drives that roll polyester yarn from one roller to the next, so the diagnostic effort must be performed on a live line without disturbing the running process.

This article decodes the S5-135U error string, isolates the failure to a missing or unrecognised memory submodule on the CPU 921, walks through the on-line / off-line Program Status procedure, and documents the field checks used to verify the S5-135U analog input and analog output cards that drive the DC motor references.

2. S5-135U System Architecture Reference

The S5-135U is a mid-size SIMATIC S5 PLC with a central rack (CR) and up to three expansion racks (ER) connected by IM 305/IM 306 interface modules. Up to four central processors can be operated in parallel inside the central rack; each CPU executes its own STEP 5 user program and exchanges data through the S5 backplane handshake flags or via coordinated DBs.

Memory is held on plug-in submodules of the 6ES5 375-x... family (RAM, EPROM, or EEPROM). The submodule is treated as part of the CPU: if the card is missing, defective, or holding an incompatible image, the CPU will not bring the user program into a state that the PG can display under Program Status. The 135U/155U family manual describes the diagnostic LED combinations and the PG status block used to identify this case.

Reference documentation:

• SIMATIC S5-135U/155U CPU 921/922/928 Manual (Siemens Support entry 1085938)
• S5-135U/155U System Manual — Memory Submodules 6ES5 375
• STEP 5 — Programming and Diagnostic Functions

3. Decoding the Error String

The PG message "AS Interface not loaded / CP RAM not plugged in" contains two phrases that the STEP 5 online interface uses to explain why the program image cannot be mapped into the editor:

The fact that the CPU stays in RUN and the parallel CPU 922 still shows Program Status proves that the rack, IM, and I/O are healthy. The fault is local to CPU 921 and its memory submodule.

4. Root Cause Analysis

From the field report, the most probable root causes for the message on a CPU 921 that does still run the process are, in priority order:

1. Memory submodule removed or partially ejected. The card is the only carrier of the user program image. If it has been lifted even a few millimetres (loose bezel screw, vibration, recent maintenance), the CPU will hold the process in RUN using the retain image of M/T/Z/P flags, but the PG cannot build the on-line Program Status image.
2. RAM battery exhausted / write-protection jumper mis-set. After a power-down the volatile RAM is cleared, the CPU no longer has a user-program image to map to the PG, and "CP RAM not plugged in" is reported even though the slot is physically populated.
3. Wrong submodule type installed. A read-only EPROM is acceptable for RUN, but the PG reports the missing "CP RAM" interface when the on-line edit buffer is unavailable. An EEPROM without a fresh write produces the same effect.
4. Submodule contacts oxidised. Common in 30+ year old S5 installations. Cleaning the edge connector restores continuity.
5. Corrupt program / version mismatch. If the user program on the card was last edited off-line and never re-loaded through the PG, the checksum does not match the on-line interface table that STEP 5 expects.
6. PG/STEP 5 project mismatch. The PG project does not correspond to the on-line program block. CPU 922 returns a clean image because its submodule was reloaded together with the PG project; CPU 921 was not.

5. Step-by-Step Diagnosis (No Production Stop)

1. Verify the rack status LEDs. On the CPU 921 front panel confirm RUN (green), STOP (off), and that the BAF (battery fault) LED is off. If BAF is lit, replace the lithium backup battery on the CPU before any further action.
2. Read the CPU diagnostic buffer from STEP 5. On the PG, PLC → Diagnostic Buffer (or ISTACK for older STEP 5 versions). The interrupt stack identifies whether the user program loaded successfully, the cause of the last warm restart, and any STL/OB error codes.
3. Compare the on-line and off-line program identifiers. In STEP 5, open File → Compare Block. If the on-line OB1/FB/DB list returns "no user program", the diagnosis is confirmed: the RAM submodule does not contain a valid STEP 5 image for the PG to map.
4. Physically inspect the memory submodule. With line power locked out and CPU 921 in STOP, remove the bezel screw, slide the card carrier, and verify:
   • The 32-pin card sits fully in its socket (no raised edge).
   • The part label matches the 6ES5 375-x...LC.. (RAM) or 6ES5 375-x...LA.. (EPROM) family.
   • Edge connector pads are clean, not verdigris, and the gold fingers are intact.
   • The write-protect jumper on the card is in the correct position (RAM = enabled for writing, EPROM = read-only).
5. Reload the program. Insert a freshly prepared EPROM (preferred for permanent production use) or a battery-backed RAM submodule containing the matching STEP 5 project. Use PLC → EPROM → Write to transfer, then perform a Cold Restart from the PG.
6. Re-test Program Status. With the project loaded, open a ladder block and press Status. The "AS Interface not loaded / CP RAM not plugged in" message must clear; symbols, setpoints, and current rung values must be visible.

6. CPU 921 vs CPU 922 — Why One Works and the Other Does Not

The S5-135U coordinator uses coordination flags to pass data between the CPUs; the user programs on each CPU are otherwise independent. Therefore, a single CPU reporting CP RAM not plugged in is fully consistent with the other CPU showing healthy Program Status. The differences that matter for the diagnosis are:

If the program on the CPU 922 was reloaded (for example, after a recent modification) and the CPU 921 submodule was not updated, the PG-side project is the one that runs on the line, but CPU 921 carries an older or empty image. The next time an engineer opens the PG project, the version check on CPU 921 fails and the "AS Interface not loaded / CP RAM not plugged in" string is displayed.

7. STEP 5 Online Program Status Procedure

Once the memory submodule is in good order, follow this sequence to enable on-line status on CPU 921:

1. Connect the PG to the CPU 921 programming port (front-panel 15-pin D-sub, AS 511 protocol, 9 600 bit/s by default).
2. Start STEP 5, open the matching .S5D project file.
3. Select PLC → Connect → Online. STEP 5 negotiates with the CPU. If a project mismatch is detected, the PG prompts to upload or cancel — upload is the safer option on a running line.
4. Open OB1 in the editor.
5. Press F4 = Status (or click the lightning icon). The status bar at the bottom of the editor shows the Status word: STATUS / AS-INTERFACE OK / CP-RAM OK.
6. Verify the rolling of values: I, Q, F, T, C symbols should update at scan frequency.

8. Memory Submodule Verification Procedure

A field engineer should perform the following bench test on the suspect memory card once it is removed from the line. A standalone PG 720 / PG 740 with the S5-EPROM utility, or any modern PG with a USB-to-TTY adapter and STEP 5 V7, is sufficient.

Bench verification steps:

1. Insert the card in the PG's memory submodule slot.
2. Read the card with File → Read EPROM → Auto-detect. If the read fails or returns a checksum error, the card is unusable and must be replaced.
3. Compare the read content with the off-line project: Block → Compare.
4. If the read succeeds and matches, erase (EPROM) or clear (RAM/EEPROM) and re-write from the canonical project file.
5. Re-install the card in the CPU 921, perform a cold restart, and re-test Program Status.

9. Program Status Troubleshooting Matrix

10. Diagnosing the Motor-Sequencing Fault

The field report also records that "when the start button is pressed the second motor/drive starts first instead of the first one, which does not start." The S5-135U ladder logic is the most probable cause because the I/O is shared with CPU 922, which controls the same machine, and the digital I/O card already tested as faulty has been removed. With Program Status restored on CPU 921, the engineer can confirm the following on-line:

1. Open OB1 in Status and trace the start-button input (e.g. I 32.0) through the rung.
2. Verify that the bit is set when the button is pressed and the contactor-coil output (e.g. Q 32.0 for motor 1, Q 32.1 for motor 2) follows.
3. Compare with the cross-reference list (Cross-Reference menu). If the symbol MOTOR_1_RUN is set, but the output that energises the first contactor is bound to Q 32.1, the engineer is looking at a wiring/program mismatch, not a hardware failure.
4. Check the interlock chain: each subsequent drive must be enabled only when the previous drive has reached speed feedback. In a Leybold polyester winder the interlock is typically Speed OK → Permit Tension → Permit Next Roller. A missing interlock in OB1 will cause the second drive to start without the first having been proven healthy.
5. Check the analog setpoint path: each DC drive is fed by an analog output from the S5-135U. If the wrong output slot/channel is wired, the second drive will receive the reference intended for the first.

11. Analog I/O Card Diagnostics

The S5-135U uses analog modules from the 6ES5 4xx family. The most common in a DC-drive rolling application are:

Reference: S5-135U/155U Analog I/O Modules manual (Siemens Support 1108302).

11.1 Verifying an Analog Output Card

1. Open Status on the data word feeding the analog output (e.g. DW 20 in DB 50).
2. Force a known value (e.g. 0 for 0 V, 2048 for +5 V, -2048 for -5 V if the module is bipolar, using the canonical S5 scaling: 0 → 0 V, +2048 → +10 V, -2048 → -10 V, ±32767 not used).
3. Measure the analog output terminal with a calibrated DMM at the field terminals of the module.
4. Repeat at the drive input terminals to rule out cable faults.
5. If the module output is correct at the terminals but the drive does not respond, the fault is downstream of the analog output (drive enable, wiring, drive electronics).
6. If the module output is stuck at zero or saturated, the module is suspect. Swap with a known-good spare and re-test.

11.2 Verifying an Analog Input Card

1. Disconnect the field device (or place it in manual) and inject a known signal at the card's terminal block using a calibrator (e.g. 4.000 mA, 12.000 mA).
2. Monitor the corresponding input word in Status (e.g. IW 64). With a 12-bit S5-135U AI module, 4 mA → 0, 20 mA → +2048. Tolerances ± 2 LSB are normal; outside that the card is suspect.
3. Reconnect the field device and verify the live reading against the calibrator.
4. If the reading drifts while the input is constant, the suspected cause is a noisy cable, a missing shield, or a damaged input amplifier on the card.

11.3 Quick Health Check Without Calibrator

If a calibrator is not available on a live line, perform the following indirect check using the drive's own healthy operation as the reference:

1. Verify that the AI reading on the tension/load cell tracks the actual draw (the operator can read this from the rolling stand).
   Source ambiguity: if the source value reported is the load cell range (e.g. 0–500 daN), compare the AI word against the engineering unit conversion in the STEP 5 code rather than against an absolute mA value.
2. Verify that the AO to each drive matches the setpoint requested by the program. If a drive is running at half speed despite a 100 % setpoint, the AO scaling is suspect.
3. Use the ISTACK and the diagnostic buffer entries for the AI/AO cards. An "analog module not responding" message identifies a backplane handshake problem; a "limit value exceeded" message identifies wiring or transducer scaling.

12. Verification Checklist Before Returning to Production

13. Frequently Asked Questions