Siemens LOGO!8 Input Stuck at 0: 24VDC Ground Reference Fix

Problem Summary

A 12/24 VDC Siemens LOGO! 8 base module - for example 6ED1052-1MD08-0BA1 (LOGO! 8.3 BM 12/24 RCE) - reads digital input I5 as logic 0 in the LOGO! Soft Comfort online monitor even when a sensor clearly drives the terminal toward 24 V. Communication with the LOGO! is healthy, all other program functions behave normally, and forcing the input from the simulation tool works correctly. The input LED on the front of the module is also off. This is the canonical "two isolated 24 V supplies" failure mode of the LOGO! 8 digital input stage and is one of the most frequently reported support topics for first-time LOGO! installers.

The condition is not a hardware defect, not a programming error, and not a wiring fault in the I5 wire itself. It is a missing common reference between the LOGO! 8 base module and the sensor's 24 VDC power supply. The digital input stage inside the base module measures voltage between the input terminal (I5) and the module's M terminal (also labeled 0V). If the sensor's 0 V return is not at the same potential as the LOGO!'s M terminal, the differential voltage seen by the input comparator can be undefined, below the guaranteed "1" threshold of +8.5 V, or even slightly negative - all of which register as logic 0.

Affected Hardware and Identification

The failure described in the source case applies to every member of the LOGO! 8 (and LOGO! 8.1, 8.2, 8.3, 8.4) family that is built on a 12/24 VDC base module or a 24 VDC digital expansion module. Identify the affected part by the Siemens article number printed on the side label of the module.

The "RCE" suffix indicates integrated Ethernet, "CE" the variant without Ethernet. The trailing "-0BA1", "-0BA2", or "-0BA3" identifies the firmware generation (8.1, 8.2, 8.3, 8.4) and matters only for certain analog-input-on-digital features; the grounding rules in this article apply identically to every variant listed.

Digital Input Electrical Specification

Every Siemens LOGO! 8 12/24 VDC variant uses the same input stage. The published numbers from the LOGO! 8 system manual are summarized below and are the values the field engineer must respect when selecting a sensor and supply.

Root Cause: Why Two Independent 24 V Supplies Break an Input

The 12/24 VDC variants of LOGO! 8 do not contain an isolated input stage. The input comparator inside the module is referenced to the LOGO!'s own M/0V terminal. From the LOGO!'s point of view, the input state is computed as:

V_diff = V(I5) - V(M)

The result is then compared against the +5 V (logic 0 ceiling) and +8.5 V (logic 1 floor) thresholds shown in the previous table. For the input to register as logic 1, this differential must rise above +8.5 V. The complication arises when the sensor is powered from a second, independent 24 VDC supply:

• If the sensor supply 0V (return) and the LOGO! 8 M terminal are NOT bonded, V(M) is left "floating" with respect to the sensor's 0 V. The only thing tying them together is the internal pull-down inside the LOGO! input stage (approximately 5.6 kOhm to M). This is enough to register a clean logic 0 but not enough to define a stable reference for the sensor's switching transistor.
• If the two 0 V returns are bonded through protective earth only (PE), there may be several volts of difference due to ground-loop drop. This is still inside the indeterminate band 5 V to 8.5 V, and the LOGO! may flicker between 0 and 1, or stay at 0.
• If the two 0 V returns are bonded directly (a single wire between the two PSU 0 V terminals), the references are equal, V_diff equals the sensor output, and the input works normally.

This is why the field community's three-option answer is the canonical fix:

1. Power the sensor from the same 24 VDC source that powers the LOGO! 8 (best for simple installations).
2. Use two independent supplies and bond their 0 V/return terminals with a dedicated wire.
3. Use two independent supplies WITHOUT bonding, and place a relay or optocoupler between the sensor output and the LOGO! input. The optocoupler provides galvanic isolation and re-references the signal to the LOGO!'s M terminal.

Three Wiring Topologies

The three valid wiring strategies are shown schematically below. The highlighted path represents the 0 V/common reference path that must exist for any non-isolated topology.

Topology 1 - Single Common 24 VDC Supply

This is the simplest and recommended topology for most bench and panel builds. The same 24 VDC source feeds both the LOGO! 8 base module and the sensor. There is no possibility of a ground reference mismatch because there is only one reference.

Topology 2 - Dual 24 VDC Supplies with Bonded 0 V

Two independent supplies are kept, but a dedicated equipotential bond wire is added between their 0 V terminals. The bond wire carries no significant load current (the sensors only sink what they need) and is sized like a protective conductor (typically 2.5 mm² Cu minimum). This topology is common in panels where the sensor supply is a dedicated, fused, and current-limited rail separate from the LOGO!'s logic supply.

The dashed red line in the diagram is the equipotential bond. It is what makes Topology 2 work. Without it, the sensor and the LOGO! are isolated and I5 cannot register a defined state.

Topology 3 - Dual 24 VDC Supplies with Optocoupler Isolation

When the sensor must remain galvanically isolated from the LOGO! (long cable runs, sensitive instrumentation, or different grounding systems), use an optocoupler or a relay between the sensor output and the LOGO! input. The optocoupler is the standard solution because it draws negligible current from the sensor and provides several kV of galvanic isolation. Common choices on a 24 VDC logic level are the Phoenix Contact PLC-OSC-24DC/24DC/2 (10 mm wide DIN-rail module), the Wago 857-400, or a discrete optocoupler such as the PC817 with a 2.2 kOhm input resistor.

The optocoupler's input side is referenced to the sensor PSU 0V. The optocoupler's output side is referenced to the LOGO! M terminal. The two references are now isolated by the optical barrier, and the LOGO! input sees a clean 0V or 24V signal regardless of the sensor PSU's floating potential.

Step-by-Step Diagnostic Procedure

Before changing any wiring, run the following checks in the order shown. Each step rules out a class of fault before moving to the next.

1. Open the LOGO! Soft Comfort online monitor (Tools -> Online -> Start Online Test). Note whether the LOGO! shows "Online" status and whether the input LEDs on the front of the module respond at all when you short I5 to +24V by hand with a piece of wire.
2. Force the input to 1 in the simulation tab of LOGO! Soft Comfort. If the program logic reacts correctly but the real input does not, the program is fine and the issue is electrical.
3. With the sensor disconnected, bridge I5 directly to a known +24V terminal on the LOGO! base module with a short piece of wire. If the input now reads 1 in the monitor and the green LED turns on, the LOGO! input stage is healthy. The fault is in the sensor circuit.
4. Measure V(I5) to V(M) on the LOGO! with a digital multimeter set to DC volts. With the sensor un-driven: expect 0 V to 1 V typical (pull-down internal to LOGO!). With the sensor driven (PNP switching on): must read at least +10 V DC. If you measure 24 V from I5 to PE instead, you have confirmed the floating reference issue.
5. With the sensor driven, measure V between sensor 0V terminal and LOGO! M terminal. If the reading is more than 0.5 V, the two 0V references are not bonded and you have identified the root cause.
6. Bond the two 0V terminals with a 2.5 mm² wire and retest. The input should respond within one scan cycle (typically 20-50 ms in LOGO! 8).
7. If bonding the 0V is not acceptable for the application (e.g. medical or hazardous-area installation), install an optocoupler between the sensor output and the LOGO! input as shown in Topology 3 and re-test.

PNP vs NPN Sensor Wiring on LOGO! 8

The source case uses an optical sensor with a PNP (sourcing) output. PNP is the preferred output type for LOGO! 8 12/24 VDC inputs because the LOGO! 8 expects current to flow INTO the input terminal at logic 1. The wiring is straightforward for PNP:

For an NPN sensor on a LOGO! 8 digital input, you must add an external pull-up resistor (typically 4.7 kOhm, 0.25 W) between the LOGO!'s +24V terminal and the input terminal. The LOGO! 8 internal pull-down is approximately 5.6 kOhm to M, which is too weak to overcome an active NPN pull-down - the input will otherwise stay at 0 even when the NPN transistor is fully off.

FE vs PE: Functional Earth and Protective Earth

The LOGO! 8 base module exposes distinct earth/ground terminals whose meanings are not interchangeable:

FE and PE are connected at exactly one point (the panel star ground) so that shield currents have a defined return path but fault currents do not flow through the FE conductor. Connecting FE to PE at multiple points creates a parallel path and may inject noise into the input stage.

The source case mentions that FE is left unconnected. This is acceptable for bench testing with short, unshielded wires but should be bonded in a real panel for EMC reasons. The FE terminal is NOT the place to bond the two 0V supplies; use a dedicated M-to-M wire as shown in Topology 2.

Expansion Modules: Repeating the Same Mistake

When you add a DM8 12/24V or DM16 24V digital expansion module, each module has its own M terminal that is internally connected to the base module's M terminal via the bus connector on the right side of the modules. As long as the base module's M terminal is properly bonded to the sensor supply 0V, expansion module inputs work without any extra wiring.

The trap appears if you power the expansion module from a SECOND 24VDC source (some configurations allow this when isolating field and logic). In that case, the expansion module's M terminal is no longer the same potential as the base module's M terminal, and the high-speed digital bus may malfunction or the inputs may be misread. Always verify that every expansion module's M terminal is at the same potential as the base module's M terminal. If you need to power an expansion from a separate source, use a relay or optocoupler on the outputs of that expansion to re-reference signals back to the base module's M.

The DM8 230RCE and DM16 230RCE (AC inputs) do not suffer from this issue because AC inputs are referenced to L and N, not to a common DC return. Be careful not to mix AC and DC modules on the same base if you assume the rules are identical.

LOGO! Soft Comfort Configuration of Digital Inputs

While the electrical fix is the primary cause of the failure described here, a software-side check should not be skipped. In LOGO! Soft Comfort V8.x and later, open the project and double-click the digital input block (or use the Inputs/Outputs menu) and verify the following:

• Input type: Confirm the input is configured as "Digital" and not "Analog" or "Counter". On LOGO! 8.3 and 8.4, inputs I1, I2, I7, I8 on the BM 12/24 can be re-purposed as analog (0 to 10 V) inputs. If the analog option is selected, a digital 24 V signal will be misread.
• Debounce filter: The default debounce is typically 10 ms. For high-speed counters on I3-I6, set the debounce to 0 or to the minimum required by the application (e.g. 0.1 ms).
• Invert function: Confirm the input is NOT inverted in the program. An inverted input will appear to be "stuck at 1" when the actual signal is 0, which can confuse diagnosis.
• Block connections: Verify that I5 is actually wired to a useful block in the FBD/LAD program. An unused input may show "online" but the program's reaction will be invisible.

Use the LOGO! access tool 6ED1052-1MD08-0BA1 SD card or the Web-Based Management interface (Tools -> Online) on the LOGO! 8 RCE variants to monitor live input states directly on the module's built-in web page. This avoids any ambiguity about what the program is seeing.

Cable Selection and EMI Best Practices

Sensor wiring on a LOGO! 8 input is not trivial when the cable runs longer than 3 meters or runs alongside VFD output cables. Follow these rules:

• Wire gauge: Use 0.5 mm² to 1.5 mm² stranded copper for sensor runs. Avoid solid-core wire in vibrating panels.
• Cable type: Use shielded twisted pair (e.g. LiYCY or similar) when the run exceeds 5 m, runs through a panel with VFDs, or runs near welding equipment. Ground the shield at the LOGO! end only to the FE terminal.
• Maximum length: For a standard 24 V PNP sensor with 100 mA of output capability, runs up to 50 m are reliable. Beyond 50 m, add a 10 kOhm pull-up at the LOGO! end to maintain noise margin.
• Separation from power: Maintain at least 200 mm separation between sensor signal cables and AC power cables, and 300 mm from VFD motor cables. Cross at 90 degrees if a crossing is unavoidable.
• Common-mode choke: For long runs in noisy environments, install a ferrite sleeve on the signal wire at the LOGO! entry point. This suppresses common-mode RF without affecting DC levels.

Verification Procedure After the Fix

1. Power-cycle the LOGO! 8 and the sensor supply. Allow 5 seconds for the LOGO! to complete its startup diagnostic.
2. Open LOGO! Soft Comfort, go to Tools -> Online -> Start Online Test. Confirm that the input monitor for I5 (and any other affected inputs) transitions correctly when you trigger and de-trigger the sensor.
3. On the LOGO! 8 front panel, confirm that the green LED for I5 turns on and off in sync with the sensor state. The LED is a direct hardware indicator of the input comparator output and will illuminate whenever V(I5)-V(M) > ~8.5 V.
4. Measure V(I5)-V(M) with a multimeter at the LOGO! terminal block. PNP on: should read +22 to +26 V DC. PNP off: should read 0 to +2 V DC.
5. If the application uses a high-speed counter (I3, I4, I5, or I6), run a 1 kHz pulse train through I5 and confirm that the LOGO! counts correctly. If counts are off by 50% or more, the input is borderline (between 5 V and 8.5 V) and the wiring is still marginal.
6. Document the bonding point between the two 0V terminals on the panel schematic. Future maintenance personnel will need this to avoid accidentally lifting the bond.

Related Symptoms and Field Notes

• Input works for a few hours then sticks: classic symptom of a missing equipotential bond combined with thermal drift in one of the PSUs. The bond fixes the issue permanently.
• Input reads 1 for a brief instant at LOGO! startup: the floating reference is pulled high by the LOGO!'s internal pull-up at boot. Adding a 10 kOhm external pull-down to M suppresses this false trigger.
• Analog inputs (AI1-AI4 on LOGO! 8.3/8.4) read near zero but not exactly zero: same root cause - missing 0V reference - but more obvious because the analog value quantizes the floating offset.
• Inputs read correctly when the panel door is open but incorrectly when closed: RF interference picked up by the floating input wiring. Bonding M to M and adding a 100 nF X2 capacitor across the input to M filters the noise. The capacitor must be rated for the working voltage and must be X2 or Y2 safety class if the wiring leaves the panel.
• Sensor LED on the sensor body lights up but LOGO! input stays 0: sensor is powered correctly, but the output transistor is unable to source enough current. Add a 1 kOhm pull-up to the LOGO!'s +24V terminal. This is rare but documented for very small photo-electric sensors.
• Input works on the bench but not in the panel: usually a grounding or shielding issue; check the panel star-ground integrity and verify the FE bond is in place.

Frequently Asked Questions