Coverage & methodology
What XrefBase covers, and how a claim gets here.
A small graph of relationships, each traceable to where it came from, and a plain count of everything not yet assessed. These figures are computed from the same public records the lookup reads.
Figures computed . Recomputed at most hourly.
The catalogue in figures
- Parts in catalogue
- 3,601
- Manufacturer part numbers the lookup can resolve, each filed under one manufacturer.
- Manufacturers
- 496
- Manufacturer records those parts are filed under. Some imported names still await an identity check, so this is not yet a count of distinct companies.
- Parts with any recorded relationship
- 1022.8% of parts
- Parts whose lookup returns anything at all, in any of its six groups — from a published replacement to a mating counterpart. Most are context, not substitution claims.
- Parts with a published replacement
- 100.3% of parts
- Parts whose lookup returns something under “Published replacements” — the only answer that asserts one part can stand in for another.
- Parts with a functional equivalent
- 00% of parts
- Similar function, not a confirmed drop-in. Engineering review is required before substitution.
- Aliases recorded
- 142
- Other numbers for the same physical part. An alias is identity, never an alternative.
- Obsolete parts recorded
- 10
- Parts whose lifecycle is recorded as obsolete. For 3,591 others no lifecycle is recorded — unknown, not active.
- Open Orphans
- 0
- Parts somebody needs an alternative to, still seeking one on the board — open, with a candidate declared, or under review.
3,499 parts (97%) have no recorded relationship. That means not assessed — not “no alternative”.
No manufacturer notice, part-number grammar or reviewed declaration has placed them in the graph yet. A lookup for one of them says exactly that, and so does this page.
Relationships by type
One row per kind of relationship, worded exactly as the lookup words it. Counts are published pairs: a relationship between two parts counts once, whichever end it is looked up from.
| Relationship | Claim | Published pairs | How it is created | Confidence |
|---|---|---|---|---|
| Drop-inPublished replacements | Asserts substitution | 0 | A manufacturer statement or two independent public sources, then independent review. The database refuses to publish one without high confidence and a source. | None published |
| SupersessionPublished replacementsLooked up from the newer part, the same pair is listed under “Superseded predecessors” — history, not an alternative. | Asserts substitution | 10 | A manufacturer change notice (PCN/PDN), checked by an independent reviewer and given a revalidation date. 10 from manufacturer change notices · 10 of 10 cite a public document |
|
| FunctionalFunctional equivalents | Not a substitute | 0 | A reviewed declaration or a specification match. Published labelled and scored; engineering review is still required before use. | None published |
| Same seriesSame series | Not a substitute | 118 | Part-number grammar, only where a manufacturer’s numbering is self-documenting. The part numbers are the evidence, so it caps at medium confidence. 118 from part-number grammar · 0 of 118 cite a public document |
|
| VariantVariants | Not a substitute | 50 | Part-number grammar: the same configuration under a suffix whose meaning is unconfirmed. Caps at medium confidence. 50 from part-number grammar · 0 of 50 cite a public document |
|
| Mates withMating counterparts | Not a substitute | 6 | Part-number grammar, kept in a separate table so it can never be returned as an alternative. Caps at medium confidence. 6 from part-number grammar · 0 of 6 cite a public document |
|
Where the coverage is
Each bar is a group’s parts in the catalogue, all on one scale. The fills say how many of those parts have a published replacement, how many have some other recorded relationship, and how many have none — which means not assessed.
- Published replacement
- Relationship recorded, but no published replacement
- Nothing recorded — not assessed
By manufacturer — the 15 largest, then everyone else
Show these figures as a table
| Manufacturer | Parts | Any recorded relationship | Published replacement | Not assessed |
|---|---|---|---|---|
| TE Connectivity | 487 | 20 | 10 | 467 |
| Wago | 153 | 0 | 0 | 153 |
| Deutsch | 150 | 82 | 0 | 68 |
| Harting | 140 | 0 | 0 | 140 |
| Amphenol | 128 | 0 | 0 | 128 |
| Aptiv | 111 | 0 | 0 | 111 |
| Lapp | 106 | 0 | 0 | 106 |
| Phoenix Contact | 95 | 0 | 0 | 95 |
| Lear | 87 | 0 | 0 | 87 |
| Schlemmer | 87 | 0 | 0 | 87 |
| Carling Technologies | 77 | 0 | 0 | 77 |
| Huber+Suhner | 76 | 0 | 0 | 76 |
| Molex | 66 | 0 | 0 | 66 |
| Weidmüller | 63 | 0 | 0 | 63 |
| HellermannTyton | 62 | 0 | 0 | 62 |
| All other manufacturers (481 manufacturers) | 1,713 | 0 | 0 | 1,713 |
By top-level category
Show these figures as a table
| Category | Parts | Any recorded relationship | Published replacement | Not assessed |
|---|---|---|---|---|
| Connectors | 1,055 | 81 | 0 | 974 |
| Terminals & Contacts | 692 | 20 | 10 | 672 |
| Wire & Cable | 393 | 0 | 0 | 393 |
| Other | 291 | 0 | 0 | 291 |
| Cable Protection | 250 | 1 | 0 | 249 |
| Fuses & Fuse Holders | 176 | 0 | 0 | 176 |
| Seals & Grommets | 176 | 0 | 0 | 176 |
| Wire Management | 138 | 0 | 0 | 138 |
| Switches & Indicators | 128 | 0 | 0 | 128 |
| Terminal Blocks | 103 | 0 | 0 | 103 |
| Labels & Marking | 93 | 0 | 0 | 93 |
| Backshells & Accessories | 69 | 0 | 0 | 69 |
| Relays | 37 | 0 | 0 | 37 |
Review freshness
- Reviewed relationships
- 10
- Reviewed between
- 5 Sept 2026 – 18 Sept 2026
- Next revalidation due
- 5 Sept 2027
- Overdue for revalidation
- None
Every reviewed relationship is due a second look on its revalidation date: a manufacturer can withdraw or revise a notice, and a claim nobody re-checks only gets older. The other 168 published relationships carry no review date, and none of them makes a claim: the lookup presents them as context, never as a substitute.
Methodology
How a relationship gets created — cheapest and most certain first
- 1 · Part-number grammarWhere a manufacturer’s numbering is self-documenting — Deutsch DT and HD are — a parser derives same-series, variant and mating links. It is deterministic, but a grammar is not a manufacturer statement, so these cap at medium confidence and never assert substitution. Where a grammar cannot be verified, no parser runs: TE’s was evaluated and rejected because it would have invented structure.
- 2 · Manufacturer change noticesA PCN or PDN in which the manufacturer says one part replaces another. Authoritative and public, it is the only source that earns high confidence with a citation. Each is checked by an independent reviewer before publishing and carries a revalidation date.
- 3 · Alternate designationsOther numbers already embedded in imported part numbers become aliases — the same part under another number, never an alternative to it. Conservative patterns only: a bare slash is usually part of the number.
- 4 · Specification matchingDatasheet attributes compared within a category to propose functional candidates. Not in use yet: it is the expensive tier, and it waits until the mechanical ones are exhausted.
- 5 · Reviewed declarationsAn engineer, producer or distributor declares a candidate — against an Orphan or directly — with public documents. It stays a declaration until a reviewer turns it into a candidate relationship, and a separate decision publishes it.
Three rules that do not bend
- A drop-in is published only when verified.High confidence and a public source, enforced by the database rather than by good intentions. A wrong drop-in claim stops a production line.
- A mating part is never an alternative.Counterparts live in their own table so they cannot be returned as one. DT04-3P mates with DT06-3S — the one part guaranteed not to substitute for it.
- Every published claim cites a public source.A private file may point at what to check; only the publicly re-confirmed fact is published, citing the public source. Family, variant and mating links cite nothing further: the part numbers are the evidence.
What “not assessed” means
A part with no recorded relationship — or a part that is not in the catalogue at all — has not been looked at yet. That is not a finding that no alternative exists, and XrefBase never presents it as one. The same holds when a lookup fails: “we could not check” is not “none found”, which is why a section of this page that cannot be read says so instead of showing a zero.
How to help
The parts that get review time next are the ones people are blocked on. If you need an alternative, open an Orphan — your reason, customer and volumes stay private. If you make or sell a credible substitute, declare it with its public documents; it never becomes a published relationship without independent review.