Controls engineers, panel builders, machine OEMs, and system integrators designing 24VDC rails inside industrial control cabinets. Especially useful if you need monitoring (DC OK / alarms), or you’re building UL 508A-ready panels for the U.S. market.
On paper, “two 24V power supplies” sounds redundant. In real control panels, redundancy fails for four repeatable reasons: (1) backfeed (one supply energizes the other through a shared bus), (2) shared fault paths (a single branch short collapses both supplies), (3) lost voltage headroom (ORing drop + wiring drop pushes the PLC below its tolerance), and (4) no monitoring—so one supply can be dead for months while the panel “seems fine,” until the remaining supply fails.
The fix is not a single magic part. It’s a wiring-first design: decide where isolation happens (diode ORing, MOSFET ORing, or full bus separation), then add explicit monitoring signals (DC OK contacts, alarm relays, or PLC inputs) so operators know when redundancy is gone. If you only do “two supplies in parallel” without output isolation, you don’t actually have a reliable redundant system.
Most industrial control cabinet projects land in one of these three patterns. The “best” choice depends on current level, voltage headroom, how critical the loads are, and whether you need alarms/telemetry for maintenance.
| Approach | Where isolation happens | Typical pros | Typical risks / watch-outs |
|---|---|---|---|
| ORing diode | At the ORing diode junction | Lowest complexity; easy retrofit | Voltage drop + heat; limited monitoring unless added |
| Redundancy module | Inside module (diode or MOSFET/ideal diode) | Lower drop; better diagnostics; some add current balancing | Higher cost; still need correct downstream protection |
| A/B split-bus | Isolation by design (separate buses) | Best fault containment; simplest “proof” per bus | Some loads can’t tolerate split supply; needs architecture discipline |
ORing means you connect multiple sources so the load is automatically fed by the source with the higher voltage, while preventing reverse current into the other source. The most straightforward implementation is two diodes feeding a shared 24V bus. This is popular in small PLC cabinets because it’s fast to wire and easy to explain during commissioning.
ORing diodes can work well when your load current is modest and your 24V rail has voltage headroom (for example, if your devices tolerate 20–30V or you trim the supply slightly higher where allowed). They also fit retrofits: you can add diode ORing between existing supplies and a legacy bus without redesigning the entire cabinet.
The tradeoff is predictable: diode forward drop reduces the bus voltage, and that drop turns into heat. At higher current, this can become a thermal and reliability constraint—especially inside compact enclosures. Treat it as a first-principles check:
Practical takeaway: if your PLC/safety rail is already tight on voltage, ORing diodes are the first place you’ll lose margin.
A redundancy module is essentially a purpose-built ORing/decoupling device designed for industrial panels. Many families offer two internal technology options: economical diode decoupling, or MOSFET-based “ideal diode” decoupling with much lower voltage drop. The wiring is typically cleaner than discrete diodes, and the more important advantage is operational visibility: alarms, status outputs, and (in some designs) current balancing.
The safest wiring pattern is: PSU A → module input A, PSU B → module input B, then a single module output → 24V bus, followed by branch protection and labeled terminal blocks. Do not parallel PSU outputs directly unless the PSU specifically supports current sharing and the architecture is designed for it.
If you can separate loads, an A/B split-bus design often provides the best fault containment: one supply feeds a critical bus (PLC CPU, safety controller, core comms), and a second supply feeds a field bus (valves, contactor coils, sensors, peripheral I/O). This eliminates the shared fault path where a single short drags down everything.
Put PLC/safety on Bus A, field wiring on Bus B. If a field device shorts, Bus A stays up and the controller can still report the fault.
Some devices accept two 24V inputs (A and B) and handle switchover internally. Use this where you need uninterrupted power at the load.
If a single load must remain powered through a supply failure and it cannot accept dual inputs, split-bus alone won’t solve it. In that case, you typically OR the supplies (preferably via a redundancy module for headroom and monitoring) to power that critical load group.
The most expensive redundancy system is the one you think you have—but don’t. Redundancy can be lost silently if one supply fails, if a feed wire loosens, or if the load quietly grows until one supply can’t carry it alone. Your wiring should therefore include explicit redundancy-health signals to the PLC/HMI or maintenance system.
If you’re building U.S.-market panels, keep the monitoring wiring clean and documented: labeled terminals, consistent wire numbering, and an I/O map that explains the alarm logic. This reduces commissioning time and speeds future troubleshooting.
Redundancy is not complete until you can verify it on the bench (or at least during factory acceptance). Below is a practical checklist that panel shops use to avoid field failures. It focuses on measurable outcomes: bus voltage under load, thermal rise, and behavior during forced faults.
Even when your article focus is wiring, the U.S. market expects you to speak the “panel language”: UL 508A is the common industrial control panel standard context for component selection and documentation, while machine electrical safety standards commonly reference SELV/PELV concepts for low-voltage control supplies. Separately, many 24V control circuits are treated as power-limited (for example, Class 2 sources in NEC context), which can affect how you size supplies and route wiring.
Here’s a common “high availability” PLC cabinet pattern for U.S. industrial automation: keep the controller rail alive through a single PSU failure, while preventing field wiring faults from resetting the PLC. This example uses a redundancy module for the critical rail and a separate supply for field loads.
Send your load list (amps), cabinet ambient, and whether you need alarms/monitoring. TPS can recommend DIN-rail supply options and help you design a cabinet-ready wiring and documentation package.
For most standard DIN-rail supplies, simple paralleling without output isolation risks backfeed and unpredictable sharing. A redundancy/decoupling method (diode ORing, MOSFET ORing, or a dedicated redundancy module) makes behavior repeatable and safer.
It depends on diode type and current, but forward drop often becomes the dominant “hidden” loss at higher current. Treat the drop as lost headroom at the bus and translate it into heat with P ≈ I × Vdrop.
When you can separate critical controls from noisy or failure-prone field wiring, A/B split-bus avoids shared fault paths. It’s especially effective when a field short should never reset the PLC/safety controller.
At minimum: PSU A DC OK, PSU B DC OK, and a “no longer redundant” alarm (from the redundancy module or derived logic). Log transitions so maintenance can replace a failed supply before the second failure.
Use a DMM, clamp meter, and a basic thermal camera if available. Perform three fault cases: PSU A off, PSU B off, and a single branch short. Confirm the critical bus stays within device tolerance and the alarm/logging works.
Start with TPS DIN-rail power supply collection and request an RFQ if you want a fast fit check. If you need a cabinet-ready build with wiring and documentation, use the Integration Solutions page or Contact Us.
