Dirac ConsultingMassive Dynamic commissioning platform
← All manuals·Logic
🔀 Cx logic diagrams

See the whole job as a diagram you did not have to draw.

Packages, dependencies and execution order drawn from the project's real wiring and schedule, with a layout you can edit per project.

Topology or execution orderDerived, scoped diagramsEditable layout
Start here ↓

What do you want to do?

Jump straight to the answer

🔓Open the Logic DiagramFrom the rail, or ask Dirac to summon one.→📦Package by equipment typeWhy the boxes are grouped the way they are.→👀Read the diagramColours, arrows and what a line means.→✏️Edit the layoutMove boxes, add custom ones, per project.→🧭Derived logic for one assetUpstream, downstream or both.→❓Common questionsFAQ.→

Overview

📦

Package-to-package, not asset-to-asset

One node per real Package Manager block. Individual member assets aren't drawn — that's the whole point of grouping similar equipment into a package in the first place: fewer boxes, a readable diagram of the whole project.

🔌

Real cross-package connections

One edge per (source package, destination package, layer) combination that has at least one real canvas connection crossing between them. A control-layer link and an operational-power link between the same two packages are two separate, separately-colored edges.

The diagram is generated automatically from your real Package Manager blocks and their real cross-package connections — nothing here is manually authored.

Opening the Logic Diagram

1
Open a project and click 🔀 Logic Diagrams in the left rail.
2
There's one diagram for the whole project — no group picker. Every real Package Manager block with at least one real asset member shows up as a node automatically.
3
If no package qualifies yet, the panel says so plainly: put real assets into a Package Manager block on the canvas and connect them to another package.

Why package by equipment type?

Grouping assets of the same equipment type into one package is what keeps this diagram small and readable. A project with hundreds of individual assets can still show as a handful of package nodes.
ScenarioWhat happens
Two packages have a real connection between any of their member assetsOne edge appears between the two package nodes, on whichever layer that connection is on.
Two connections between the same pair of packages are on different layersTwo separate edges — one amber (control), one blue (operational power) — each with its own checklist detail.
Two assets are connected but both belong to the SAME packageNo edge is drawn — that link is internal to the package and considered obvious.
An asset isn't in any package yetIt's out of scope for this diagram — add it to a Package Manager block to bring it in.

Reading the diagram

Every solid box is a real Package Manager block, labeled with its real asset count. Every dashed box is a custom annotation you or someone else added (see "Custom boxes" below) — it has no backing package.

Edge colorMeaning
AmberControl layer — the same amber the canvas's own connection lines use for control-layer wiring.
BlueOperational power layer — same blue the canvas uses.
Green, dashedA custom connection you drew yourself — no backing real data.
Click any edge to open a detail panel on the right: it lists every underlying real connection ("Asset A → Asset B") and, below that, every checklist item — all 4 stages, every item, not just what's outstanding — for the source-side asset(s), color-coded green/red/grey/pending by result. This panel is read-only: it's a second live view onto the exact same checklist data the Commissioning Tracker edits, not a copy, and not something you tick from here.

Editing the layout, and adding custom boxes (per-project)

The auto-generated grid layout is a readable default, not a fixed diagram. Everything below only appears for accounts with edit rights on the project — the same rule Management View and the 3D BIM Viewer already use; a viewer-tier account sees the diagram but can't change it.

1
Click ✎ Edit in the header to enter layout mode.
2
Drag any box — package or custom — to reposition it. It snaps to a quarter-grid-cell on release and saves automatically; there's no separate Save step.
3
Click + Box to add a custom annotation box (default label "New note"). Double-click any custom box to rename it — its text is fully editable, for annotating something that has no backing package, such as a hold point or a note for a third party.
4
Click 🔗 Connect, then click two boxes in turn (package, custom, or a mix) to draw a custom connection between them — shown as a dashed green line, distinct from the solid lines drawn from real canvas connections. Click the small circle at the midpoint of a custom connection, or the delete marker on a custom box, to remove it.
5
Click Reset Layout (appears once anything has been moved) to discard every custom position and go back to the auto grid — this does not delete custom boxes or custom connections, only their positions.
Scoped to this one project. Layout, custom boxes, and custom connections are stored once per project — editing one project's diagram never affects another project's.
A real edge (solid, colored by layer) can only be changed by editing the actual asset-to-asset connection on the canvas — it's a direct read of that data, not something this panel can override. Only dashed custom connections can be deleted here. Checklist items shown in an edge's detail panel are also read-only here — tick them from the Commissioning Tracker's Checklist modal, not from this panel.

Derived logic — the summoned Cx Logic Diagram

The third diagram kind, Derived logic, is not drawn by hand at all. It is a projection of the same canvas the Scheduler and Gates already read: every box is one asset × stage task from the real Scheduler task graph, every gate input is one real predecessor the Scheduler will wait for, and every schedule grouping (joint tasks, lockstep groups) is drawn as a bracket. Because it is derived from the live data, the picture can never disagree with the schedule — and it is read-only: to change it, change the canvas, the schedule groups, or the converter ticks.

Never the whole project. A derived diagram is always scoped: you name an anchor (an asset, a package, a room, a schedule group, or an equipment type such as "all transformers"), a direction (upstream = what gates the anchor, downstream = what the anchor gates, or both) and a depth in asset hops (default 2). Anything beyond the depth is counted as "cut at depth" in the header so you know the picture is partial.
Derived logic (and the Execution-order diagram, which walks the same real Scheduler task graph) computes in the background: the panel shows a loading skeleton while the first computation for a newly-opened or newly-scoped diagram runs. Changing Direction/Depth/Layer re-derives the same way; results for a given scope are cached in memory for the session, so re-opening or re-narrowing back to an already-computed scope is instant.

No packages? Draw from the wiring alone

Package Managers are optional. Topology and Execution-order diagrams are drawn between packages, so they need at least one; Derived logic does not. In + New Diagram choose Derived logic (the form opens on it by itself when the project has no package) and tick Whole project — every asset, from the wiring alone: every asset and every wire on every layer is drawn, in Project Management Only mode as well, with nothing else to set up. For a large project pick a room, a schedule group, an equipment type or a few assets instead, or keep the depth low; above about 150 assets the form warns you that the picture will be hard to read. Dirac does the same: logic diagram for the whole project.

Summoning one from Dirac

The quickest way is to ask Dirac. All of these create (or re-open) a derived diagram and open Logic Diagrams on it, and reply with the same logic in words:

  • logic diagram for XFMR P — both directions, 2 hops
  • upstream logic of PEMA-H / what gates PEMA-H — what must be done before it
  • downstream logic of HVS N — what it releases
  • control logic of LV Board / field logic for LV Board — one layer only (in PM mode: office / field / support; "scada" also means the control layer)
  • logic diagram for all transformers — every present asset of that equipment type as anchors
  • logic diagram for LV Room / logic diagram for XFMR pair — a room's contents or a schedule group
  • add depth 4 at the end to widen the walk

Summoning goes through the usual Confirm/Cancel gate because it creates a diagram entry in the project. Asking for the same scope twice re-opens the existing diagram instead of duplicating it. You can also create one by hand: + New Diagram → Derived logic, then tick anchors from the asset / package / room / group / equipment-type lists.

Reading the picture

1
Columns are the order of execution. A box's column is its longest chain of prerequisites inside the scope — column 1 has nothing before it, the rightmost column is last. Read left to right.
2
Boxes are asset × stage. "XFMR P — L2" is that asset's L2 task. Rounded boxes are shared (joint) tasks of a group. A dashed red box is a source that is wired in but not present on site (Available unticked) — nothing fed from it can be released until it is, or a temporary source stands in. Colour = live state: done, ready to start, waiting, not present, override (temporary source / executive decision). ★ marks the anchors, ⇄ a layer converter (PM mode: report → release), ⏱ a temporary source (PM mode: executive decision).
3
Gates sit in front of a box. & — every input drawn into it is required. ≥1 — any one of the inputs suffices (redundant feeds). Both can appear on one box. "+N outside scope" means the Scheduler has more inputs for that box than the current depth shows.
4
Edges are gate inputs, coloured by the layer the Scheduler took them from — the legend uses the project's own words (Operational power / Control / Mechanical, or Field / Office / Support in PM mode). Grey = the asset's own previous stage, purple = joint/lockstep sibling, red = safety gate (neighbours must be installed first). Solid = satisfied today, dashed = still open. Hover an edge for the reason in words.
5
Brackets are schedule groupings. Every asset group the Scheduler applies — joint (shared task) or lockstep — is drawn as a dashed purple bracket with its label. If the Scheduler treats it as a predecessor or gate, it is drawn here.
6
Click a box for the detail panel: state, stage, room, equipment type, its step in the order of execution, everything it needs (with the reason for each input) and everything it releases. In words toggles the same formula as text — the reply Dirac gives when you summon.
The controls above the picture (Direction / Depth / Layer) re-derive the diagram; they are the only thing you can change here. Edit mode, custom boxes and custom connections are not available on a derived diagram — it mirrors the schedule logic and nothing else.

FAQ

I put assets in a package but nothing shows up in Logic Diagrams — is that a bug?

No — check that the package has at least one real asset member, and that it has a real connection to another package if you're expecting an edge. A single isolated package still shows as a node with no edges, which is accurate if nothing's actually wired to it yet.

Why don't I see any lines inside one package's own assets?

By design — connections between two assets in the SAME package are considered obvious and aren't drawn. Only connections that cross from one package to another become edges.

Can I tick a checklist item from inside this panel?

No — the edge detail panel is read-only. Open the Commissioning Tracker's Checklist for that asset/stage to record a Pass/Fail/N/A result; it'll show updated here immediately since both views read the same data.

Will editing this diagram in one project affect another project?

No. Layout, custom boxes, and custom connections are stored on that project only.

Can a viewer-tier account drag boxes or add custom annotations?

No — "Edit", "+ Box", and "Connect" only appear for accounts with edit rights on the project, same convention as Management View and the 3D BIM Viewer.

I can't create a logic diagram — it asks for packages and I have none (or this is a Project Management Only project).

Choose Derived logic as the kind and tick Whole project (or pick the assets, rooms or types you want). It is drawn from the wiring alone and needs no Package Manager. Topology and Execution-order diagrams are the ones that need packages.

Why can't I drag boxes or add notes on a derived diagram?

Because it is a mirror of the live Scheduler logic. Change the canvas wiring, the converter ticks, the Available state or the schedule groups and the diagram follows; annotations belong on Topology / Execution diagrams. See Derived logic.

Dirac says it can't find the thing I named.

The name is matched against assets, then packages, rooms, schedule groups and equipment types (singular/plural tolerant, a unique partial match is enough). Use the exact canvas name, or say "all UPS" for an equipment type.

What happened to the earlier asset-to-asset version of this feature?

Replaced the same day it shipped. This version operates at the package level instead.