Hitachi 31: guided diagnosis

Narrow it down

What you can check, and what needs a technician

Technician verification sequence

For the attending technician — these steps involve live equipment or sealed systems and are not homeowner checks.

1. Confirm the exact model and the manufacturer's own definition for the displayed code, including whether it describes memory, configuration, addressing, a link, an input/output condition or general control.
2. Establish which of those the code actually reports before anything is changed — the families of cause behind them do not overlap.
3. Review the history: power interruptions, servicing, and any board, controller or unit replacement, since each one changes which cause is likely.
4. Inspect the relevant harnesses, connectors and ribbon links on the control side for partial seating, corrosion and heat damage, following correct isolation practice.
5. Verify that the board and controller fitted are the correct pairing for this equipment, including generation and capacity.
6. Confirm addressing, unit numbering and configuration against the manufacturer's documentation for the exact model rather than against generic values.
7. Verify the supply and input conditions the board receives where the code or the evidence implicates them.
8. Use only the manufacturer's supported service or configuration procedure where one is required, obtained for that model.
9. Establish whether the fault persists once the history, configuration, addressing, connections and supply have been excluded.
10. Only then treat the board or its memory as the failed hardware, and confirm the system configures and runs correctly afterwards.

About resetting

Switching the system off at its wall switch or isolator once, waiting ten minutes and switching it back on is reasonable, and what happens next is genuinely useful evidence: a control code that clears and stays away behaves differently from one that returns immediately.

When not to reset: Do not keep resetting it. A control code that returns is reporting something real, repeated resets remove the pattern a technician would use, and on a memory or configuration fault they can leave the control repeatedly interrupted mid-write rather than resolving anything.

What this code can point to

Configuration or parameter state that does not match the equipment Most likely

The control holds configuration for the equipment it is fitted to. Values missing, incomplete, out of range or belonging to a different model produce a control fault that is corrected by configuration rather than by parts — and it is a frequent finding after a board, controller or unit change.

Addressing, unit-number or system setup mismatch Likely

On systems where several units and controllers share a network, addresses and unit counts must agree. A duplicate address, a unit never set, or a count that no longer matches after a change produces a control fault with every part healthy.

Harness, connector or ribbon connection on the control side Likely

Control signals and board links run through connectors and ribbons that can be partly seated, corroded or heat-affected — often disturbed during servicing. A connection is a likely area to inspect before a board is suspected, because it is common, cheap to correct and frequently the actual cause.

Recent servicing, board or controller change Likely

A capacity mismatch very often appears after a unit or board change, because the setting has to match the combination actually installed.

Stored data or memory state on the control board Possible

The memory holding the unit's stored settings could not be read, written or verified. That may indicate the memory device, but a disturbed write, an incomplete configuration and a link to that device produce the same report — which is why this is a finding rather than a verdict.

Model-specific diagnosis required Possible

Control codes carry model-specific configuration, addressing methods and documented service procedures, and some brands reuse the same characters across generations with different meanings. Where that applies, documentation for the exact model decides what the code means and what may safely be changed.

Temporary control state that may clear Possible

A control can latch a fault state and hold it until it is properly restarted. Where the code follows a supply event and clears cleanly on one controlled restart, that is a possible cause worth establishing first because it costs nothing to rule in or out.

A recent power interruption or supply event Possible

An interruption, a generator changeover or a voltage dip while the control was writing or reading its stored data will produce a memory or control-state fault with no part having failed. This is a likely area to consider whenever the code appeared around a power event.

Control data integrity or state fault Possible

The control rejected data it holds or exchanges internally. That establishes a control-state fault and identifies the area, not the component: the data may have been disturbed, left incomplete, or be failing to hold.

Board or controller link fault Possible

Where a code names a link between boards, or between a board and its own controller, the path has two ends and a harness in between. A link fault may indicate either end, and the path is proven before either is condemned.

Supply arriving at the control board Less common

A board reporting its own supply is reporting something that starts outside itself: the feed to it, its connector, a protective device or a supply event. It is a possible cause that belongs before any conclusion about the board.

Input or output condition at the control board Less common

The board either did not receive what it expected on an input or did not achieve what it commanded on an output. Both ends and the wiring between them are part of the picture, and what is connected to that circuit is inspected alongside the board.

Which control-system condition this code refers to

Addressing, unit-number or setup mismatch. In the addressing and unit-numbering setup shared between the units and controllers on the system.

The code concerns how the units and controllers are addressed and numbered rather than whether a part works. Duplicated addresses, a unit never set, a count that no longer matches after a change and an addressing method that differs between generations all produce this — and it appears far more often after installation or service work than on its own.