IMDTouchKnowledge BaseShop main store

Selection guide

Choosing a Panel PC With PRP Network Redundancy Support

Panel PCs and displays Published 7 min read

Short answer

A Panel PC joins a PRP network only through a documented PRP interface, normally an add-on card in a verified expansion slot, or by sitting behind a RedBox. Two onboard Ethernet ports prove nothing. Confirm the exact computer, slot, card, driver and operating-system combination in writing before ordering, or route the panel through a RedBox.

On this page
  1. What PRP demands from an endpoint
  2. Two Ethernet ports prove nothing
  3. Three ways to put the HMI on a PRP network
  4. What to verify before ordering
  5. Worked example: one external slot on an x8 daughter card
  6. Commissioning and the failover test
  7. Questions asked

A Panel PC joins a PRP network only through a documented PRP interface — in practice an add-on card in a verified expansion slot — or by sitting behind a RedBox. Two onboard Ethernet ports prove nothing: PRP is a duplicate-and-discard function in the network interface and its driver, not a connector count. IEC 62439-3, with PRP in Clause 4, requires every doubly attached node to sit on two separate LANs at once and delivers zero-time recovery when either LAN fails. The selection decision is short: if the exact panel PC variant has a slot the PRP card is supported in, buy that combination with written confirmation; if it does not, pair a panel display with a separate industrial PC that carries the card, or attach the panel to a RedBox. The sections below give the checks, the comparison, and a worked slot-and-port budget.

What PRP demands from an endpoint #

PRP is defined in Clause 4 of IEC 62439-3; the current edition is dated 2021. In plain terms, every doubly attached node holds two Ethernet ports, one on each of two independent LANs — LAN A and LAN B — that never exchange frames except through doubly attached nodes. The sender transmits a copy of every frame on both LANs; the receiver keeps the first copy that arrives and discards the duplicate. Either LAN can fail with no lost frames and no reconfiguration, which is why PRP suits controls that cannot tolerate a switchover hiccup. The standard also lets singly attached nodes reach the redundant network through a RedBox.

PRP works at the Ethernet layer, so the protocols above it — EtherNet/IP, Modbus TCP, OPC UA — run unchanged over the redundant pair. The same standard defines HSR, the ring counterpart in which nodes forward frames around a ring in both directions, and its Annex C carries a Precision Time Protocol profile that holds synchronization across PRP to 1 μs, relevant when the HMI time-stamps events. One boundary to keep: PRP raises network availability; it is not a safety function, and machine safety stays on its own rated devices whatever the network does.

Two Ethernet ports prove nothing #

A standard NIC pair forwards each frame once and has no idea a second LAN exists. The duplicate-and-discard function has to live in the interface hardware and its driver, and on PC-class hardware it comes as an add-on card, most often a riser card, rather than as a built-in feature of a panel computer.

A card that physically fits can still lack a driver for your operating-system version, and a slot present on one hardware generation of a product family can be absent on the next. The buying test is documentary, not visual: written vendor support naming the exact computer model and configuration, the expansion slot, the card, the riser if one is used, the driver, and the operating system. A datasheet that lists ports and slots is not evidence of PRP; a line that names PRP support for that card in that computer is.

Three ways to put the HMI on a PRP network #

Path Where PRP lives Verify in writing before ordering Choose it when
PRP card inside the Panel PC Add-on or riser card in the panel's expansion slot; the panel itself is the doubly attached node Slot type, lane width and mechanical space on the exact variant; card supported in that slot; driver for your operating system You want one panel-mounted box and the slot is confirmed on the configuration you can order
Panel display plus a separate industrial PC The card sits in the industrial PC; the panel becomes a display Card, slot and driver support on the industrial PC; display cable type and length; panel mounting The touchscreen computer cannot take a supported card but an industrial PC can
Panel PC behind a RedBox A switch acting as a RedBox attaches a singly attached node to both LANs and performs the duplication RedBox model and topology against its documentation; the single cable from panel to RedBox No supported card path exists, or several non-PRP devices share one drop

The RedBox row is the easiest to procure and the weakest at the endpoint: the one cable from RedBox to panel is a single point of failure, so the panel can lose the network while both LANs are healthy. A directly attached panel has no such link — either LAN can die unnoticed. If this HMI is the machine's only operating window, that difference alone rules out the third row. If the panel is also replacing an older operator terminal in the same project, the rest of that swap is covered in planning an HMI terminal replacement from an older operator panel.

What to verify before ordering #

  1. Fix the network role first: must the panel be a doubly attached node, or is a RedBox acceptable? One answer here halves the hardware search.
  2. Pull the expansion documentation for the exact variant being ordered, not the family brochure. Panel PC lines have carried PCI or PCIe expansion on earlier variants and changed it later; a slot listed for another generation is not a slot you can order.
  3. Confirm how the slot is reached. On some units an external slot is fed by a daughter card plugged to a board connector; the card you add must be supported on that arrangement, within the card height and length that fit behind the panel.
  4. Check the card vendor's support list for the card, the slot type, and a driver for your exact operating-system version.
  5. Get the whole combination — computer, slot, riser, card, driver, operating system — confirmed in writing before the purchase order.

The PRP interface is one axis of the selection; the same unit still has to meet the sealing and mounting conditions of its enclosure, the ground covered in selecting an IP65 industrial touchscreen for a dusty HMI.

Worked example: one external slot on an x8 daughter card #

Inputs:

  • The network: two LANs, A and B, kept fully separate, per the IEC 62439-3 design.
  • The panel PC under evaluation, of the class used for PC-based HMI runtimes: its board carries a PCIe x16 connector, occupied by an x8 daughter card that presents the single external expansion slot — the arrangement observed on production units of this type.
  • The candidate PRP interface: one riser card that occupies one slot and carries the two Ethernet ports.

The rule in words: a doubly attached PRP node needs one Ethernet port per LAN, two ports in total; PC PRP cards deliver both ports on one card; one card consumes one expansion slot.

The arithmetic:

  • Ports required: 2 LANs × 1 port per LAN = 2 ports.
  • Ports supplied: 1 card × 2 ports per card = 2 ports. Spare PRP ports: 2 - 2 = 0, and the onboard NICs are not part of the attachment.
  • Slots: 1 external slot available - 1 slot per card × 1 card = 0 slots left for anything else.
  • Board connector: the x16 connector is already occupied by the daughter card, so the only place the PRP card can live is the external slot on the x8 link.

The result: the direct-PRP path fits this unit exactly — one supported card, no spare slot, both PRP ports on the card rather than the onboard NICs. The decision: one verified slot plus one supported dual-port PRP card is the minimum for a direct PRP panel; with both confirmed in writing for card, slot, riser, driver and operating system, order that configuration. A variant with zero external slots closes this path — choose a separate industrial PC carrying the card, or a RedBox.

Commissioning and the failover test #

Install and configure strictly from the documentation of the exact card and RedBox; settings carried over from unrelated hardware are how PRP nodes end up singly attached in practice.

  1. Install the card and riser, then confirm the operating system lists the network hardware and the driver named for your version.
  2. Land one card port on LAN A and the other on LAN B; check link and communication on each path alone before connecting both.
  3. With the HMI running, disconnect LAN A. The application must continue without a missed update — PRP recovery is zero-time — then reconnect and confirm both paths again.
  4. Repeat the pull on LAN B, then record the hardware, driver versions and results for maintenance.

Two traps to respect: never cable LAN A to LAN B, because the LANs must stay separate and a bridge between them defeats the discard logic and can flood duplicates; and never read a lit link LED as proof of redundancy — it shows one link, not dual attachment. If the HMI stumbles on a pull, suspect the landing, both ports on one LAN, or the driver before the hardware. Above PRP, the controller connection is configured exactly as on a single network, the pairing walked through for connecting an HMI panel to an S7-1500 PLC.

Questions asked

Can any Panel PC with two Ethernet ports join a PRP network?

No. Sending every frame twice and keeping only the first copy that arrives is a function of the network interface and its driver, and ordinary dual-port hardware does not implement it. The panel becomes a doubly attached node only through a documented PRP interface, normally an add-on card supported in a verified slot. Have the computer, slot, card, driver and operating-system combination confirmed in writing before you order.

When is a RedBox the better choice for the HMI?

Use one when the panel PC has no supported PRP card path, or when several non-PRP devices can share the same drop. The RedBox attaches singly attached devices to both LANs and performs the duplication and discard for them. The tradeoff is the single cable from RedBox to panel: that one link can fail while both LANs stay healthy, something a directly attached panel never experiences.

How do I prove PRP failover at commissioning?

Unplug each LAN in turn while the HMI runs and watch the application, not the link lights. PRP recovery is zero-time, so a missed update means the node is not truly doubly attached: check that one port lands on each LAN, that nothing bridges the two LANs, and that the loaded driver matches the card documentation before replacing hardware.

Does PRP double the traffic on the network?

Every frame is transmitted on both LANs, so each LAN carries the full application traffic and the pair together carries twice that. HMI-to-controller polling is light next to link capacity, so bandwidth is not what decides this selection; slot and driver support are.

Can the same card also join an HSR ring?

Some PRP cards also support HSR, the ring-topology protocol defined alongside PRP in IEC 62439-3, and a card documented for both can be reconfigured between them. The cabling differs, two parallel LANs for PRP against a ring for HSR, so treat the mode as part of the supported combination you confirm in writing.

Keep reading

  • Selection guide

    Planning an HMI Terminal Replacement From an Older Operator Panel

    Decide an operator-terminal replacement on installed cost — the terminal, cutout adaptation, application software, engineering, commissioning, downtime and a spare — never on terminal price alone. Measure the opening against the replacement's documented cutout range and count displays before ordering: a slightly smaller terminal needs a bezel adapter, and a large converted application can force the higher-performance model. Convert in the same brand when the source project opens cleanly; rebuild cross-brand when you accept the extra engineering hours for source control.

    HMI and touchscreens

  • Selection guide

    Selecting an IP65 industrial touchscreen for a dusty HMI

    Yes, a third-party touchscreen works with a CX5340 and TE2000 build once four checks pass: the IP65 claim covers the surfaces your installation actually exposes, the HMI client's video output matches the monitor input, the touch interface and its driver match the client's operating system, and the panel cutout leaves a sound gasket land. Buy on those four lines, not on the brand.

    Panel PCs and displays

  • How-to

    Connecting an HMI Panel to an S7-1500 PLC

    Confirm the HMI's documented driver covers the S7-1500 for your exact panel model and software revision, put the panel and CPU on the same subnet, permit the connection in the CPU protection settings, then validate one known PLC tag before building screens. A working Ethernet link alone proves nothing about tag access.

    HMI and touchscreens