HART and Foundation Fieldbus (FF) are the two protocols most commonly stamped on modern smart valve positioner nameplates, but they solve different problems: HART is a digital overlay on a 4-20 mA analog loop used for intelligent device management, whereas FF is a fully digital, two-wire bus that carries both IDM and real-time closed-loop control end to end [S1][S5].
The choice is driven by four engineering variables: brownfield wiring vs new install, control architecture (centralized DCS vs field-decentralized), device count per cable run, and the maintenance crew's training level. Getting any of these wrong tends to lock the plant into a protocol for the next 15-20 years, which is why the decision is made at the P&ID stage, not at procurement [S1][S4].
Physical layer: 4-20 mA+FSK vs IEC 61158 two-wire digital
HART superimposes a 1200 bps / 2200 Hz FSK Bell 202 digital signal on top of a conventional 4-20 mA loop, so the same two wires still carry one analog process variable for control and an additional digital channel for configuration, calibration, diagnostics, and viewing of internal variables [S1][S5].
FF replaces the analog current with a 31.25 kbit/s Manchester-coded, intrinsically safe two-wire segment governed by IEC 61158; multiple devices (typical segment design 12-16, max 32) share the same pair, the bus itself powers the field device, and the same wire pair carries both the process value and the control loop [S1][S5]. The two physical layers are not interoperable: an FF positioner cannot ride on a 4-20 mA loop, and a HART-only positioner cannot publish its PV as a bus variable for a peer-to-peer FF control block.
What each protocol actually does on a positioner
On a HART positioner, the 4-20 mA current is still the only signal the DCS or PLC uses to command valve position; HART only runs in parallel for IDM tasks such as push-button calibration, partial-stroke test trending, and pneumatic actuator leakage diagnostics [S3][S5]. Control is therefore centralized in the controller, and the positioner behaves like a 4-20 mA device with a maintenance modem bolted on.
On an FF positioner, control moves into the field: standard function blocks (AO, PID, AI, DI, DO, IS, AR, etc.) execute inside the device, two or more blocks in different devices can be linked peer-to-peer over the same segment, and a single PID block in the positioner can close a loop against a pressure or temperature transmitter on the same wire pair without round-tripping through the DCS [S2][S5]. This is the structural reason FF positioners are used for mission-critical loops (boiler drum level, compressor surge, ESD partial-stroke) where the engineer wants the control loop to survive DCS scan-time jitter and where the segment is built to Foundation Fieldbus HSE/ISE integration rather than to a protocol gateway patched onto a HART loop.
Decision matrix: HART vs FF for a new positioner spec
Engineers picking between the two for a valve positioner spec line should weigh five criteria side by side. On installation cost, HART reuses existing 4-20 mA wiring and I/O cards, so brownfield projects typically come in at a fraction of an FF install that needs new segment cards, fieldbus gateway / power conditioners, and dedicated cable [S1][S2].
On diagnostics, both protocols expose partial-stroke test, friction/breakout trend, and seat-load data, but FF can do it on a time-scheduled basis linked across multiple valves on the same segment, while HART relies on the DCS polling each device sequentially [S2][S5]. On control architecture, FF allows in-device PID and peer-to-peer linking that HART physically cannot; on training burden, HART is closer to the technician's existing 4-20 mA mental model and can be troubleshot with a handheld communicator, whereas FF segment design, LAS redundancy, and function-block configuration typically require a week or more of formal training [S1][S3]. On device count per cable, FF carries many devices per segment while HART is normally point-to-point (multi-drop HART exists but is rarely used on control loops because the analog value is lost), so FF is the cheaper option only when the engineering buys into the new wiring [S1][S5].
Where HART still wins in 2026
For brownfield DCS migrations, tank-farm level control, on/off valve islands, and any loop where the 4-20 mA I/O already exists, HART remains the lowest-risk choice: the existing controller and I/O stay put, the smart camera-class of handheld configurators, PLC analog cards, and DCS HART pass-through all keep working, and the maintenance team does not need to learn segment engineering [S1][S3][S4].
For skid-mounted packages, OEM machine builders, and end-users running Emerson DeltaV or Honeywell Experion with native FF I/O, FF is the more efficient fit because the segment design, function-block programming, and asset-management tools are already in the controller's DNA [S2][S5]. For projects where the control philosophy is still being defined, a phased rollout (FF on the new unit, HART retained on the rest, bridged by a protocol gateway) is the most common real-world compromise documented by integrators [S1][S4].
Failure modes and constraints that decide the spec
Three constraints routinely overturn a default HART or FF choice. First, hazardous-area certification: a HART positioner on a 4-20 mA loop can ride an entity-concept intrinsic-safety barrier; an FF segment on the same device will need a FF-certified barrier and a FISCO power supply, which limits cable length and segment device count, and these physical limits are not interchangeable with HART's 4-20 mA distances [S1][S4]. Second, control-architecture consistency: mixing HART and FF on the same valve train forces a gateway, adds latency, and creates two parallel asset databases, so most EPCs pick one protocol per unit and stick to it [S4][S5].
Third, latency: HART's 4-20 mA update rate is dominated by the DCS scan (typically 0.25-1.0 s), while FF H1 segments execute scheduled macrocycles in the 100-500 ms range with deterministic bus arbitration, which is why fast loops (compressor anti-surge, turbine trip) almost always end up on FF or hardwired 4-20 mA without HART overhead [S2][S5]. Engineers should treat any vendor claim of "WirelessHART replaces FF" with caution: WirelessHART is a 2.4 GHz mesh intended for monitoring and slow IDM, not for the deterministic control loops where FF is specified.
Sourcing and standards to put on the spec sheet
Reference these directly on the datasheet and the I/O list: IEC 61158 for the FF H1 physical and data-link layer, the FF-846/FF-942 Function Block specifications for AO, PID, AI, and AR blocks inside the positioner, the HART 7/HCF specifications for universal/common-practice commands, and IEC 60079-11 (Ex i) / IEC 60079-15 (Ex n) for hazardous-area segment design [S1][S4].
For the project, also pin the segment design rules: maximum trunk length, max spur length, number of devices per segment, LAS redundancy, and termination, all of which are FF-specific and have no equivalent in a HART point-to-point loop. Document whether the positioner is being procured as a HART device with optional FF firmware variant or vice versa, and confirm the fieldbus gateway mapping for any loop that has to bridge to a non-native controller. With those spec lines locked, the choice between HART and FF stops being a debate and becomes a controlled engineering decision [S1][S4][S5].
Trackable next signals to watch: any EPC tender issued in the next quarter that names FF H1 as the default positioner protocol on a new grassroots unit (a leading indicator of greenfield FF momentum), and any controller-vendor release note that adds or deprecates native HART pass-through on the next-generation I/O card. Cross-reference both with the smart valve positioner selection matrix before reissuing a master I/O list.
Background reading: Two-Part Epoxy Pot Life: Definition, Working-Time Buckets, and 2026 Process Levers.