Floating 24 VDC: Why Both Conductors May Need Protection

In a floating or ungrounded 24 VDC control circuit, neither pole is intentionally bonded to protective earth. That changes the fault path: a first earth fault may not operate an overcurrent device, while a second fault on the opposite polarity can create a severe current path through bonded metalwork or PE conductors.
24 VDC
floating DC
earth fault
two-pole protection
control circuits
First question
Is either pole bonded to PE?
Main risk
Second earth fault
Core ruleDo not decide from the labels “+24 V” and “0 V” alone. First establish the earthing arrangement. A conductor called 0 V may still be electrically floating with respect to protective earth, and a later field connection can unintentionally change that condition.
Floating 24 VDC — normal condition
+24 V poleNo intentional PE bond
Control loadPLC · relay · sensor
0 V / return poleNo intentional PE bond
PE / bonded enclosure: normally outside the intended load-current path. Connected equipment may still create capacitive, resistive or diagnostic references to PE.
Direct answer

Should both positive and negative be fused in a floating 24 VDC circuit?

No — not automatically. First establish whether either conductor is intentionally grounded. In many IEC-oriented ungrounded control circuits, overcurrent protection is applied to both ungrounded conductors; a source that inherently limits fault current below the conductor and equipment rating can change the overcurrent-protection requirement. Overcurrent protection and simultaneous two-pole isolation are separate functions and should not be treated as interchangeable.

What “Floating 24 VDC” Actually Means

A floating 24 VDC system is one in which neither output conductor is intentionally connected to protective earth at the source. The load still sees approximately 24 V between the two output poles, but the voltage from either pole to PE is not fixed by a deliberate low-impedance bond.

This is different from a grounded control circuit, where one pole — commonly the return conductor — is intentionally bonded to PE or to the machine protective-bonding system at a defined point. Once that bond exists, the other pole is the ungrounded conductor and an earth fault on it has a clear return path.

“Floating” does not mean electrically invisible to earth. EMC capacitors, surge-protection components, insulation-monitoring circuits, analogue interfaces and connected field equipment can create high-impedance references. Those references can make pole-to-PE meter readings look stable even though there is no intentional solid bond.

Terminology
In this article, earth fault means an unintended connection between one DC conductor and PE or bonded conductive parts. Grounded means an intentional low-impedance reference exists. Ungrounded/floating means neither pole is intentionally solidly referenced to PE.

Grounded and Floating Control Circuits Behave Differently

The same 24 V line-to-line voltage can produce a very different fault current because the return path is different.
Grounded control circuit
+24 VUngrounded conductor
LoadNormal current path
0 VIntentional bond to PE
Earth fault on +24 V: PE can complete a low-impedance return path toward the bonded return point, so overcurrent protection can see substantial fault current.
Floating / ungrounded control circuit
+24 VNo intentional PE bond
LoadNormal current path
0 VNo intentional PE bond
First earth fault: one pole can become referenced to PE without automatically creating the high-current loop expected in a grounded circuit.

First Earth Fault: Why a Fuse May Stay Intact

Assume a genuinely floating 24 VDC source and a single insulation fault from +24 V to the bonded enclosure. The fault fixes that pole near PE potential, but there is still no intentional low-impedance connection from the opposite pole back to PE. The current is therefore limited by leakage paths, EMC components, measuring networks and insulation impedances rather than by a normal metallic short-circuit loop.

That current can be far below the operating threshold of a fuse or circuit breaker. The circuit may continue working, which is one reason ungrounded systems are used where continuity after a first fault is valuable. The disadvantage is that the system has lost its original insulation condition and now has less tolerance for another fault.

Fault 1 — one pole touches PE
+24 V branchAccidental contact to enclosure
PE / metalworkNew reference point
No solid return yetOnly leakage / impedance paths
Typical consequence: the first fault may not create enough current to operate an overcurrent device. It still needs to be found because the system is no longer in its original floating state.

Second Earth Fault: The Condition That Changes the Risk

A second fault on the opposite polarity can turn bonded metalwork or PE conductors into part of the fault-current path.
Fault 2 — opposite polarities fault to PE at different points
+24 VFault A to PE
PE / bonded structureConductive fault path
0 VFault B to PE
Result: the two faults can bridge the DC poles through the protective-bonding network. Schneider documentation for ungrounded DC explicitly identifies double-fault conditions in which both positive and negative polarities are involved.
1 · System starts floatingNeither pole has an intentional solid connection to PE.
2 · First insulation faultOne pole becomes referenced to PE; the system may continue operating.
3 · Second fault appearsThe opposite pole contacts PE elsewhere in the cabinet or machine.
4 · Fault loop is completedCurrent can flow through the PE/bonding path and protective devices must clear the fault according to the actual design.

Why Both Conductors May Need Overcurrent Protection

In a grounded control circuit, the intentionally grounded conductor is not treated in the same way as the ungrounded conductor. In an ungrounded control circuit, both DC conductors are ungrounded relative to PE, so either conductor can participate in a fault that must be cleared.

A Siemens control-panel reference based on IEC 60204-1 reflects this distinction: overcurrent protection is applied to all ungrounded control-circuit conductors, summarised as grounded control circuit → one-pole protection and ungrounded control circuit → two-pole protection. It also identifies an exception where the current source limits current below the current-carrying capacity of both the conductors and connected equipment.

Standards context: IEC 60204-1:2016+A1:2021 Clause 7.2.4 addresses overcurrent protection of control circuits. Clause 9.4.3.1 addresses insulation faults that can cause control-circuit malfunction. These are related but separate design questions: conductor overcurrent protection should not be confused with the strategy used to detect, tolerate or clear an insulation fault.

That does not mean “install two separate fuses in every 24 VDC circuit”. The correct device arrangement depends on the source, branch topology, conductor ratings, applicable standard, required isolation, fault-clearing performance and equipment instructions.

Current-limited 24 VDC sources

Many DIN-rail switch-mode power supplies use electronic output-current limiting. Depending on the model, overload behaviour can be constant-current, foldback, hiccup/cyclic or shutdown. If the source cannot deliver the current needed for a conventional MCB’s instantaneous trip region, a downstream short circuit can depress the 24 V rail before that MCB clears selectively. Where selective branch shutdown is required, an electronic 24 VDC protection or selectivity module can be better matched to the power supply. See 24 VDC Electronic Circuit Breakers for that separate protection topic.

Electronic protection is not earth-fault protection

Electronic branch protectors such as Phoenix Contact CAPAROC or Lütze LOCC-Box are designed primarily for overload and short-circuit protection, with functions such as current limitation, channel status and diagnostics. Those functions can improve selectivity on a current-limited 24 VDC bus, but they do not automatically provide insulation monitoring, earth-fault detection or the simultaneous two-pole isolation required by a particular circuit design. Treat branch overcurrent protection, earth-fault monitoring and isolation as separate functions unless the selected equipment explicitly combines them.

NEC Class 2 is a separate design case

NEC Class 2 is a limited-power circuit classification, not an earthing arrangement. Some 24 VDC supplies and electronic protection systems can provide Class 2-compliant outputs or channels, which can change the permitted wiring and protection approach. That classification alone does not prove that a floating circuit needs only one protected pole, nor does it replace the need to verify grounding topology, conductor protection, equipment listing and any required simultaneous disconnection.

Do not simplify this to a slogan
Floating does not automatically mean two independent fuses. The engineering question is which conductors are ungrounded, which faults must be cleared, whether the source can supply damaging current, and whether the required protective device also has to disconnect both conductors together.

Fuse, Two-Pole Circuit Breaker and Disconnect Are Not the Same Function

Device / function
Primary purpose
What it can do
Important limitation
Device / function: Single fuse
Primary purpose: Overcurrent protection
What it can do: Open one protected conductor when its time-current conditions are met.
Important limitation: Does not guarantee simultaneous disconnection of the opposite conductor and is not automatically a suitable isolation device.
Device / function: Two fuses
Primary purpose: Overcurrent protection in both conductors
What it can do: Protect each ungrounded conductor where the design requires it.
Important limitation: The two fuse elements are independent; one may open while the other remains intact.
Device / function: Common-trip 2-pole breaker
Primary purpose: Overcurrent protection plus linked interruption
What it can do: Open both poles together when correctly rated and applied.
Important limitation: Its DC voltage, polarity, pole arrangement and breaking capability must be suitable for the circuit.
Device / function: 2-pole disconnect
Primary purpose: Isolation / switching
What it can do: Disconnect the required live conductors together when the device is designed for that duty.
Important limitation: A disconnect does not necessarily provide overcurrent protection unless specifically designed and rated to do so.
Device selection
For ungrounded DC, the number of poles alone is not enough. Check the device’s DC rating, permitted pole configuration, common-trip or linked operation, breaking capacity, source fault-current capability and manufacturer wiring instructions.

How to Establish Whether the 24 VDC Output Is Really Floating

Start with the schematic and the power-supply documentation. Look for an intentional 0 V-to-PE link, a dedicated earth-reference terminal, a grounding module, an insulation-monitoring device or a field device that deliberately references one pole to earth.

Then consider the installed system rather than the power supply alone. Surge suppressors, EMC capacitors, analogue signal conditioners, shield connections, communication equipment, test equipment and other power supplies can introduce a high-impedance or solid reference. A system that was floating on the drawing can cease to be floating after a later modification.

Electrical measurements must be interpreted with the circuit topology in mind. A high-impedance digital meter can show apparently meaningful voltage from both poles to PE because tiny capacitive or resistive currents are enough to charge the meter input. That reading alone does not prove a solid earth bond.

Schematic evidenceLook for an intentional connection symbol between either DC pole and PE or the protective-bonding network.
Equipment evidenceCheck source, surge protection, monitoring, interface and field-device manuals for intentional references or leakage paths.
Measurement evidenceUse continuity, resistance and live-voltage measurements only under suitable safe procedures and interpret them with connected equipment in mind.

Interpreting Common Measurements

These patterns are clues, not universal pass/fail limits. Connected electronics can alter pole-to-PE readings.
Observed condition
Possible interpretation
What to check next
Do not assume
Observed condition: +24 to 0 V ≈ 24 V; 0 V to PE ≈ 0 V
Possible interpretation: The return may be intentionally or accidentally bonded to PE.
What to check next: Drawing, continuity path, grounding link and connected equipment.
Do not assume: That the bond is correct merely because the voltage looks familiar.
Observed condition: +24 to PE and 0 V to PE both show non-zero voltage
Possible interpretation: The source may be floating or referenced through high impedance.
What to check next: Insulation-monitoring network, EMC components and connected field devices.
Do not assume: That the two readings should add perfectly to 24 V under every condition.
Observed condition: One pole suddenly moves close to PE after a fault
Possible interpretation: A first earth fault may have established a new reference.
What to check next: Ground-fault indication and insulation resistance of branches.
Do not assume: That continued machine operation means the fault is harmless.
Observed condition: Unexpected continuity between 0 V and PE
Possible interpretation: A later modification or connected device may have created a bond.
What to check next: With the circuit de-energised, safely isolated and stored energy discharged, disconnect branches methodically and locate the introduced path.
Do not assume: That the original design was grounded.
Blown-fuse LEDs in floating circuits

Do not treat a blown-fuse indicator as proof that the downstream branch is electrically dead. Some fuse terminals use an LED indication circuit connected in parallel with the fuse, so a small current intentionally flows through the indicator after the fuse opens. That current is normally very small, but it can matter when interpreting high-impedance measurements on a floating 24 VDC system. Depending on the indicator design, an open fuse may not produce a useful LED indication when there is no downstream return path or load current available for the indicator circuit. If fuse indicators are fitted in both conductors and both fuses are open, the indication state can become misleading because the expected current path through either indicator may no longer exist. Use the indicator to identify fuse status, then confirm the actual circuit state with the schematic, manufacturer data and appropriate measurements. A blown-fuse LED is not an insulation monitor and does not prove the absence of an earth fault.

Why Ground-Fault Detection Matters

An ungrounded system can preserve operation after the first earth fault, but that benefit only remains controlled if the first fault is detected and removed before a second one appears. Schneider’s ungrounded-DC guidance explicitly recommends ground-fault detection to avoid a double-fault condition.

In low-voltage control systems this function may be implemented with an insulation-monitoring device, ground-fault detector or another method suited to the source and installation. The important engineering objective is to make a hidden first fault visible before the protective-bonding network becomes part of a second-fault current path.

Continuity versus protection
A floating system can be chosen so one earth fault does not immediately stop the process. That operating continuity increases the importance of fault indication, maintenance response and documentation of the earthing arrangement.

A Practical Decision Sequence

1. Is either DC pole intentionally bonded to PE?

Yes: treat the circuit as grounded and identify the ungrounded conductor(s). No: continue as a floating/ungrounded candidate.

2. Can the source deliver damaging overcurrent?

Compare source current limiting, branch conductor capacity and connected-equipment ratings. An inherently current-limited source may change the overcurrent-protection requirement.

3. Which conductors require protection?

Apply the governing standard and equipment instructions to every ungrounded conductor. Do not copy a grounded 24 VDC arrangement into an ungrounded circuit without checking the fault path.

4. Must both conductors be disconnected together?

If isolation or common disconnection is required, use a device and pole arrangement suitable for linked DC switching. Two independent fuses do not by themselves provide common isolation.

5. How is the first earth fault detected?

If the design intentionally remains energised after one fault, provide a practical method to identify that loss of insulation before a second fault develops.

6. Has a later modification changed the reference?

Check new field devices, surge protection, shields, test connections and secondary supplies for an unintended 0 V-to-PE or +24 V-to-PE path.

Failure Patterns Worth Investigating

First-fault alarm, no tripThis can be normal behaviour for an ungrounded system if the first earth fault current is too small to operate branch overcurrent protection. Locate the insulation fault rather than treating the absence of a trip as proof of health.
One fuse opens after a second faultThe fault path may now involve PE and the opposite polarity. Check both branches and the bonding path before simply replacing the fuse.
System was floating, now one pole sits at PEA connected device or field modification may have introduced an intentional or accidental bond. The original protection assumptions may no longer be valid.

Related Reading

Keep this page focused on earthing topology and fault paths; use the adjacent references for source and branch-protection behaviour.

Common Questions

Does every floating 24 VDC circuit need two fuses?

No. The required arrangement depends on the earthing system, applicable standard, source characteristics, conductor and equipment ratings, and the protective or disconnecting device used. In many IEC-oriented ungrounded control circuits, both ungrounded conductors are protected, but current-limited sources can alter the overcurrent-protection requirement.

Why might a first earth fault not trip a fuse in a floating 24 VDC circuit?

A genuinely floating source has no intentional low-impedance connection from either pole to protective earth. One earth fault can therefore establish a reference to earth without creating the high fault current needed to operate an overcurrent device.

Why is a second earth fault more serious?

If a second earth fault occurs on the opposite polarity elsewhere in the system, protective earth or bonded metalwork can become part of a low-impedance fault path. The resulting current can involve both polarities and may be much higher than the first-fault current.

Is a fuse the same as two-pole isolation?

No. A fuse provides overcurrent protection and opens only when its operating conditions are met. Isolation requires a device and arrangement suitable for disconnecting the required live conductors, often with linked or common operation where simultaneous disconnection is required.

How can I tell whether a 24 VDC supply is actually floating?

Do not rely on wire colours or labels alone. Review the schematic and equipment documentation, then verify whether either pole has an intentional or unintended connection to protective earth. Connected filters, surge devices, measurement circuits and field equipment can create high-impedance or accidental references.

Why is ground-fault detection useful on an ungrounded DC system?

An ungrounded system may continue operating after the first earth fault, so the fault can remain hidden until a second fault creates a more severe path. Ground-fault or insulation monitoring helps locate the first fault before that condition develops.

Safety and standards
Control cabinets can contain hazardous voltages even when the control circuit is only 24 VDC. Earthing, overcurrent protection and isolation requirements depend on the applicable standard, jurisdiction and equipment instructions; inspection or modification should be performed by a competent person using suitable safe-working procedures.