Troubleshooting Siemens LOGO! Analog Input Drift from 0-10V Current Transducers

When a Siemens LOGO! 8 (or LOGO! 6/7) base-module analog input shows a non-fixed value while the upstream 0-10 V current transducer delivers a stable voltage, the root cause is almost always one of four field-failure modes: AC primary current feeding a DC-output Hall-effect transducer, impedance mismatch between the transducer output stage and the LOGO! AI input, common-mode voltage from a floating or split power supply, or induced noise on a non-shielded cable run. This article walks an automation engineer through the diagnostic procedure, the wiring topology, the shielding rules, and the verification steps required to lock the LOGO! analog reading to a stable, scaled engineering value.

The discussion that prompted this reference described a 0-100 A panel-mount ammeter with a nameplate reading DC Current Transducer, IN DC 0-100 A, OUT DC 0-10 V. The technician had connected the transducer output to LOGO! base-module input I8 (AI2) and saw the LOGO! value wander even though a handheld DMM at the transducer terminals showed a steady voltage. The variations were real, not a software scaling bug, and the repair path is the same for almost every analogous case: verify the transducer family, verify the source impedance, verify the common-mode return, and verify the cable.

1. Siemens LOGO! Analog Input Hardware Reference

The Siemens LOGO! 8 family exposes analog functionality on two physical locations: the base module's integrated inputs I7/I8 and the optional AM2 / AM2 RTD analog expansion. Mixing these up is the first trap encountered in the field.

Critical: I7 and I8 are two independent channels, not a differential pair. Each is referenced to the same module 0 V (M). Wiring a transducer between I7 and I8 is a wiring error and will produce a reading that floats with module temperature and supply ripple. See the Siemens Industry Online Support LOGO! product tree for the latest manuals (article ID 109751654 for LOGO! 8 system manual).

The 0-10 V input is implemented as a resistive divider feeding an internal ADC. The published input resistance is on the order of 70 kΩ, which dictates the 1:10 source-impedance rule described later in this article. For the 0/4-20 mA current mode on the AM2 module, the loop load is ≈ 250 Ω, which is why a 0-10 V transducer cannot simply be tied across a 250 Ω shunt to "convert" to 4-20 mA — the resulting source impedance is the transducer's own output impedance plus 250 Ω, which can exceed the 1:10 budget.

2. Current Transducer Output Topologies

A "0-10 V ammeter" can be any of three fundamentally different devices, and the LOGO! wiring depends on which one is on the panel. Misidentification is the dominant cause of fluctuating readings in the field.

The panel device in the source case is a DC Current Transducer, IN DC 0-100 A, OUT DC 0-10 V. That nameplate is a single-rail DC-only Hall device with an output that does not produce a true representation of an AC primary current. When the technician measures the AC primary with this transducer, the Hall IC saturates or operates only on the positive half-cycle peak, and the 0-10 V output becomes a function of the AC waveform shape, frequency, and transducer bandwidth — not the RMS current. A handheld DMM on the output terminals may also read 0 V DC with a small AC component and misleadingly average it to a fixed number. That is precisely the symptom reported.

3. The 1:10 Source-Impedance Rule

The LOGO! AI input is ≈ 70 kΩ. To keep the absolute gain error introduced by the source-to-load divider below 10 % of full scale, the transducer source impedance must be at most one tenth of the AI input impedance, i.e. ≤ 7 kΩ, and preferably below 1 kΩ. This is the same rule of thumb that applies to any high-impedance PLC analog input — see the general background in the ScienceDirect analog-input overview.

Concretely, the voltage the LOGO! AI actually sees is:

VAI = Vtransducer × RAI / (Rsource + RAI)

For a 0-10 V output stage with R_source = 5 kΩ and R_AI = 70 kΩ, the divider gives V_AI = 0.933 × V_transducer, a 6.7 % loss that the LOGO! scaling block will treat as a real signal change. A higher-impedance transducer (e.g. a bare CT + burden at 10 kΩ) drops the reading to 0.875 × V_transducer, an 12.5 % loss that varies with the value of R_source if the transducer output is non-buffered. Any temperature drift in the input stage of the LOGO! AI becomes a reading fluctuation — and the engineer calls it "drift".

Active buffered Hall transducers from CR Magnetics, LEM, and Phoenix Contact typically quote < 100 Ω source impedance, which is invisible to the LOGO! 70 kΩ input. Passive rectifier-type CTs do not meet the 1:10 rule and require an external op-amp buffer or a dedicated 0-10 V transducer module.

4. Power Supply, Common-Mode, and Ground-Loop Diagnosis

A second class of fluctuating-readings complaints is created by ground loops and by floating supplies. The reference case showed the technician measuring between the transducer 0 V terminal and the LOGO! M terminal, which is exactly the right thing to do to discover a ground loop.

4.1 Single-supply vs split-supply

The Hall-effect DC transducer in the case requires a single 24 V DC supply. The LOGO! base module also requires 24 V DC. The correct topology is:

If the transducer and the LOGO! are powered from two different 24 V supplies, or from the same supply but with the transducer return bonded to a different ground point (e.g. panel PE), the LOGO! AI input sees a common-mode voltage that the single-ended ADC cannot reject. The result is a stable offset that walks up and down as the load changes and the panel PE current changes. The 70 kΩ input has no common-mode rejection specification because the AI is single-ended; any voltage present on the transducer 0 V terminal relative to LOGO! M is added directly to the 0-10 V signal.

4.2 Floating transducer supply

If the technician unplugs the 24 V supply from the transducer, the transducer output is no longer referenced to LOGO! M. The AI input then "sees" whatever the floating transducer output stage capacitively couples to ground — typically 1-3 V of garbage that wanders with hand proximity and with the position of the AC wiring in the panel. This is the most common single cause of an apparently "random" LOGO! AI reading.

5. Cable, Shield, and Noise Diagnosis

Even with the right transducer and a clean single-supply topology, a fluctuating reading can still be produced by EMI. The LOGO! AI bandwidth is intentionally low (first-order RC ≈ 22 Hz on the base-module AI), which is good for 50/60 Hz rejection but bad for any high-frequency noise that aliases through the 10-bit ADC. The rules are:

1. Use twisted shielded pair (Belden 8761, Lapp UNITRONIC LiYCY, or equivalent). The shield is the signal return, not a PE bond.
2. Connect the shield at the LOGO! end only, to the LOGO! M terminal, with a low-impedance pigtail < 50 mm. Do not bond the shield at the transducer end unless the transducer datasheet explicitly requires it.
3. Keep the analog cable in its own trough, ≥ 200 mm from any VFD output cable, contactor coil wiring, or relay-switched 24 V load. If it must cross, cross at 90°.
4. Add a 100 nF X7R ceramic + 10 µF electrolytic decoupling network across the transducer power terminals, as close to the transducer as possible, to suppress supply-borne ripple from the same 24 V bus that feeds the LOGO! outputs.
5. If the cable run exceeds 10 m, place a 10 kΩ resistor in series with the AI input at the LOGO! end and add a 100 nF capacitor to M at the LOGO! terminal. This forms a 6.4 kHz first-order filter and isolates the AI from cable capacitance.

6. Diagnostic Procedure (Step-by-Step)

Run these checks in order. Each step rules out one of the four root causes. Do not skip the first three — they catch 90 % of field cases.

1. Identify the transducer family. Read the nameplate. If the input range does not list "AC" or "true RMS", disconnect the AC source and substitute a known DC source. If the LOGO! AI is stable on DC, replace the transducer with an AC-rated unit.
2. Verify the transducer power supply. Measure the supply at the transducer terminals with the LOGO! disconnected. The supply must be 24 V DC ± 10 % and the 0 V terminal must be the same node as the LOGO! M terminal. A floating supply will read 0 V on a DMM in autoranging mode but will rise to half-supply as soon as a 10 MΩ DMM input impedance is connected.
3. Measure the open-circuit transducer output. Disconnect the LOGO! AI terminal. Measure the transducer output with a high-impedance DMM (≥ 10 MΩ). The value should match the expected DC voltage for the current being drawn.
4. Measure the loaded transducer output. Reconnect the LOGO! AI. Measure the voltage at the AI terminal, not at the transducer terminal. If the loaded value is lower than the open-circuit value, the source impedance is too high. Add a unity-gain buffer (e.g. LM358 with ±15 V or single-supply configured for 0-10 V) or replace the transducer.
5. Check for ground loops. With everything connected and the load running, measure the AC voltage between the transducer 0 V terminal and the LOGO! M terminal. Anything above 50 mV AC or 100 mV DC indicates a ground loop that must be resolved by bonding the 0 V at one point only.
6. Check for noise. Set the LOGO! AI to its 0-10 V range, disconnect the cable at the transducer end, and short the cable shield to the LOGO! M terminal. The AI should read 0 ± 2 counts (out of 1000). If it does not, the cable is picking up noise and must be re-routed or re-shielded.
7. Confirm the LOGO! software scaling. Open the LOGO! program in LOGO! Soft Comfort, place an Analog Amplifier block on the AI, and verify the input range is set to "0-10 V" and the gain is 1.0. A mis-set "4-20 mA" range on a 0-10 V transducer is a common source of "the reading wanders" reports because the LOGO! interprets 0 V as 4 mA and 10 V as 20 mA, producing a negative offset that saturates the display.

7. Wiring Topology Reference

Use the following SVG as a wiring reference for the corrected installation. The shield is bonded to LOGO! M at one end only, the 24 V supply is shared, and the transducer is the AC-rated true-RMS type.

8. Verification and Calibration

After the wiring is corrected, the LOGO! reading must be verified against a known current source. The procedure is:

1. Force a 0 A condition (open primary or zero supply) and read the LOGO! AI raw value. It must be 0 ± 2 counts (out of 1000).
2. Force 50 % of full scale (50 A through the transducer aperture with a calibrated DC or AC source) and read the LOGO! AI raw value. It must be 500 ± 5 counts.
3. Force 100 % of full scale (100 A) and read 1000 ± 5 counts.
4. In LOGO! Soft Comfort, apply a single-point gain correction if the 50 % point is off by more than 1 % of full scale. Use the Analog Amplifier block with gain 1.0 and offset 0 as the default; trim with the "Offset" parameter on the AI block only if a fixed bias remains after the wiring is verified.
5. Record the as-left values in the panel drawing and in the maintenance log. Re-verify annually or after any electrical work in the panel.

9. Troubleshooting Matrix

10. Reference Links and Manufacturer Documentation

For the underlying analog-input theory and ADC scaling used in compact PLC systems, see the National Instruments CompactRIO analog input modules reference (the resolution, input-impedance, and bandwidth figures in this article follow the same conventions used in NI documentation, and the scaling equations are platform-agnostic). The ScienceDirect analog input topic page provides a generic overview suitable for new engineers. The Arduino analogRead() reference is included as a quick refresher on the 0-5 V to 0-1023 mapping, which is the same pattern the LOGO! uses internally (0-10 V to 0-1000). Always cross-check the latest figures against the LOGO! 8 system manual on the Siemens Industry Online Support portal, article ID 109751654.

11. Frequently Asked Questions