Trace a rail from its source to its load, from both ends

A net (rail/signal) is an electrical path, and every net has two kinds of ends:

• Source: Where the net is driven/generated. For a power rail, the source is the regulator/PMIC output that produces it (e.g., the source of PP_VDD_MAIN is a converter). For a signal, the source is the output pin of the chip that drives it.
• Load: Where the net is consumed. For a power rail, the load is the chips it feeds. For a signal, the load is the input pin of the chip that receives it.

Rail tracing = following a net from one end to the other: from source to load, or from load to source. Why is this so important?

Today's question: If a rail is 'missing' at a chip, is the fault in that chip, in the rail's source, or in the path between? The answer can't be given without tracing the rail from both ends.

Because a 'not arriving' fault has three possible locations: (1) the source isn't producing, (2) the path between source and load is broken, (3) the load itself is faulty (rarest). Blaming the load (the chip) directly, without tracing the rail, is the most common and most expensive mistake. Today we learn to trace a rail correctly — by name and from both ends.

Recall from yesterday: an Apple schematic is multi-page, and nets are mostly connected by name, not by wire (same name = connected). This has a direct consequence for tracing: you can't keep chasing a wire with your eyes, because the wire enters an off-page connector at the page edge and 'disappears' — but the net continues under its name on another page.

The correct method is to follow the name:

• In boardview/schematic software: You click a net, and the software highlights every point (across all pages) carrying that name. With one click you see the net's entire world.
• In PDF/by hand: You search for the net name and jump page to page by following the off-page connector + cross-reference page number at the edge.

Apple example: While tracing the PP_VDD_MAIN rail, you see it on one page, find the 'goes to sheet 40' arrow at the edge, jump to that page, and find the net continuing under the same name. Even though the wire looks cut off, the net continues.

Reading tip — the golden rule: Don't think 'the rail ended on this page.' Note the net name, follow the off-page arrows and cross-reference page numbers. Follow the name, not the wire. This is the only reliable way to trace without getting lost in a multi-page schematic.

A single net (same name) is a single electrical node. But a real power/signal path often passes through series elements, and each side of a series element is a different net (usually with a different name). Series elements commonly found along a path:

• 0Ω jumper resistor: a bridge that closes/opens/splits the path.
• Ferrite bead / inductor (L): a series element that filters noise.
• Sense resistor: a small resistor that samples current/voltage.
• Load switch / converter: a switch that passes the rail under a separate name beyond a point.
• Connector (J): a transition between boards/flexes.

Tracing a path end to end = jumping net to net across these series elements. The net name changes often:

• Apple example 1: PP_VDD_MAIN → a divider resistor (R5901 1.3M / R5902 100K) → PP_VDD_MAIN_VMON (monitor line). The name changed because it passed through a series element (the divider).
• Apple example 2: PP_VBUS1_E75_RVP → series 0Ω (R5903) → PP_HYDRA_ACC1. Same power, new name.

Why it's critical: These series elements are where the path can break (go open circuit) — and they're exactly the points where you'll measure. If a ferrite bead goes open, the source produces the rail but it never reaches the load.

Reading tip: When tracing a path, mark the series elements (0Ω/ferrite/sense/connector). Don't let the name change stop you — it's the continuation of the same path. These points answer both 'where does it break' and 'where do I measure.'

A chip isn't being powered (its rail is missing at the pin). Verify the rail from both ends:

Step 1 — Find the net name: Read the net name at the chip's unpowered pin (e.g., PP_xxx). From now on you'll follow this name, not a 'wire.'

Step 2 — Find the source: Trace the net backward by name (click in boardview / follow off-page in PDF). Who produces this rail? Which regulator/PMIC output, on which page? Identify the source.

Step 3 — Map the path: What series elements are on the path from source to load (the chip) — 0Ω/ferrite/sense/connector? List them by following the name changes.

Step 4 — Measure from both ends: Is the rail present at the source? Is it present at the load (the chip pin)?

• Missing at the source → the fault is on the source/regulator side (not producing).
• Present at the source, missing at the load → the fault is in the path between; probe the series elements (ferrite/0Ω/connector) one by one and find the open one.
• Present at both, but the chip still doesn't work → now the load (the chip) can be a candidate (last).

Step 5 — Conclusion: Locate the fault as source / path / load. Blame the chip last; most 'not powered' cases are the source or the path.

Today's essence — rail tracing:

• Every net has two ends: source (driven/generated) and load (consumed). Rail tracing = following from one end to the other.
• Follow the NAME, not the wire: in a multi-page schematic the rail 'ends' at the edge but continues under its name; click-highlight in boardview, follow off-page + cross-reference in PDF.
• The path passes through series elements: 0Ω/ferrite/sense/connector; each has a different net on its two sides (the name changes). These points are both 'where it breaks' and 'where I measure.'
• Verify from both ends: present at source / present at load? No source→regulator; source present, load missing→path (series element); both present→chip (last).

Now you can work out where a rail comes from, where it goes, and where it breaks. Tomorrow we scale this skill: not a single rail, but reading an entire power tree — battery → PMIC → buck/LDO → rails → loads. Finding a rail's source begins with knowing where it sits in that tree.