poe(teltonika): the OTD140's PoE-out jack is 802.3af at 15 W, not PoE+ #2452

Open
jaylouisw wants to merge 1 commit from rev/1121-1642 into main
Owner

One question: what does the OTD140's PoE-out jack actually source?

Evidence

Verdict: 15 W, 802.3af (Alternative B, negotiated - not passive). One PSE socket, so the per-port ceiling is the whole budget.

What changed (data + test, no engine code)

  • devices/teltonika/otd140.json: eth2 poeOut: true -> 15, poeOutClass: "PoE+" -> "PoE"; new poe block budgetW: 15, standard: "802.3af", cited note; notes rewritten with the rows and the arithmetic.
  • test/otd140-poe-out.test.js (new, 26 assertions): the spec rows, the built device's budget/source, poeDraw() = 15 (not 1 for a boolean), a 12.95 W af AP powered and printed (Available:15.0(w) Used:13.0(w) Remaining:2.0(w)), and a 25.5 W at AP refused (PSE budget exhausted: 25.5 W requested, 15 W left, Oper power-deny).

Before: Device.poeBudget = 0 (model name spells no wattage), computePoe() skipped the box, show power inline printed budget not modelled. After: the jack powers a 12.95 W camera and refuses a 25.5 W one.

Gates: node scripts/syntax-check.js 542 files OK; audit-poe-budget silent PSEs 12 -> 11; hardware-invariants 26/26; npm test 462 suites, 15,630 assertions, 0 failed.

Unresolved: the OTD144 sibling still declares PoE+ with no budget (its datasheet was not read this run); and the engine budgets each PSE independently, so the box fed by 802.3at PoE-in can still be modelled as sourcing its full 15 W while drawing 12.95 W - the coupling of input and output is not expressed.

One question: what does the OTD140's PoE-out jack actually source? **Evidence** - Teltonika OTD140 datasheet **v1.20, 2026-07-24**, p.2 (text layer, read 2026-11-21): under **POE OUT** - `PoE ports 1 x PoE Out`, `PoE standards 802.3af Alternative B`, `PoE Max Power per Port (at PSE) 15 W Max (power supply unit dependent)`; p.7 `Power consumption Idle: < 2.5 W / Max: < 6 W / PoE Max < 21 W`. https://download.kamami.pl/p1207980-Datasheet_OTD140_1.20.pdf (HTTP 200, 2,264,361 B, 11 pp) - Same revision, second host, p.3: the same three cells. https://www.sicetelecom.it/wp-content/MU/datasheet/Teltonika-Router-TELTONIKA-Teltonika-outdoor-4G-ROU-OTD140-SICE-Distributore-Italiano.pdf (HTTP 200, 748,708 B, 10 pp) - Rejected: the earlier v1.6 sheet and flyer v1.4 say `24 W` and `802.3af/at`. 802.3af's class ceiling is 15.4 W, and the printed `PoE Max < 21 W` is 6 W of router plus 15 W out - a 24 W output has no input budget. Both are pre-correction revisions. **Verdict**: 15 W, 802.3af (Alternative B, negotiated - not passive). One PSE socket, so the per-port ceiling is the whole budget. **What changed (data + test, no engine code)** - `devices/teltonika/otd140.json`: `eth2` `poeOut: true` -> `15`, `poeOutClass: "PoE+"` -> `"PoE"`; new `poe` block `budgetW: 15`, `standard: "802.3af"`, cited note; `notes` rewritten with the rows and the arithmetic. - `test/otd140-poe-out.test.js` (new, 26 assertions): the spec rows, the built device's budget/source, `poeDraw()` = 15 (not 1 for a boolean), a 12.95 W af AP powered and printed (`Available:15.0(w) Used:13.0(w) Remaining:2.0(w)`), and a 25.5 W at AP refused (`PSE budget exhausted: 25.5 W requested, 15 W left`, Oper `power-deny`). Before: `Device.poeBudget` = 0 (model name spells no wattage), `computePoe()` skipped the box, `show power inline` printed `budget not modelled`. After: the jack powers a 12.95 W camera and refuses a 25.5 W one. **Gates**: `node scripts/syntax-check.js` 542 files OK; `audit-poe-budget` silent PSEs 12 -> 11; `hardware-invariants` 26/26; `npm test` 462 suites, 15,630 assertions, 0 failed. **Unresolved**: the OTD144 sibling still declares `PoE+` with no budget (its datasheet was not read this run); and the engine budgets each PSE independently, so the box fed by 802.3at PoE-in can still be modelled as sourcing its full 15 W while drawing 12.95 W - the coupling of input and output is not expressed.
poe(teltonika): the OTD140's PoE-out jack is 802.3af at 15 W, not PoE+
All checks were successful
CI / Server smoke and Docker build (pull_request) Successful in 34s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 6s
CI / Test (Node 22) (pull_request) Successful in 1m6s
CI / Test (Node 18) (pull_request) Successful in 1m8s
CI / Test (Node 20) (pull_request) Successful in 1m9s
CI / Arm auto-merge (pull_request) Successful in 2s
407e01f89e
The row said poeOut: true with class PoE+ and carried no poe block, so Device fell through to the model-name parser, got 0 W, and computePoe() skipped the box: show power inline printed 'budget not modelled' and a camera on the PoE-out jack stayed dark. Datasheet v1.20 (2026-07-24) p.2: POE OUT '1 x PoE Out', '802.3af Alternative B', '15 W Max (power supply unit dependent)', and p.7 'PoE Max < 21 W' = 6 W router + 15 W out. The jack now declares 15 W / class PoE and the one-port PSE gets a 15 W budget; a 25.5 W 802.3at device is refused with the arithmetic.
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-21 16:53:47 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-21 17:00:30 +00:00
Author
Owner

Blocked: superseded-by-collision with #2453 — needs Jay's decision, not mine.

Survey found #2452 (rev loop, 407e01f8) and #2453 (catalogue loop, c94d5ab2) open at the same time and both editing devices/teltonika/otd140.json. Each is individually green on its head SHA and individually clean into eab52764, but they are mutually exclusive:

git merge-tree --write-tree tmpcheck2453 tmpcheck2452
CONFLICT (content): Merge conflict in devices/teltonika/otd140.json

They agree on the headline (eth2: poeOut: 15, poeOutClass: "PoE", poe.budgetW: 15, 802.3af, 15 W not PoE+) and diverge on two things:

  1. eth1 poeIn — #2453 changes 12.95 → 21 (flyer v1.8 "PoE In: 802.3af/at … max 21W"); #2452 leaves 12.95 and says in its notes that the PoE-in class is deliberately NOT re-modelled.
  2. Sources — #2452 cites datasheet v1.20 (2026-07-24, kamami mirror); #2453 cites flyer v1.8 (2024) plus the Teltonika wiki pinout. Both quote the same "15 W / 802.3af Alternative B" and the same "< 21 W PoE Max" row.

#2453 also carries devices/teltonika/otd144.json; #2452 carries a new test/otd140-poe-out.test.js (130 lines) that #2453 does not.

Question for Jay, because it is about behaviour and not about conflict text: is poeIn the negotiated class draw (12.95 W, class 0) or the inlet ceiling (21 W)? If 21 W is the inlet total, it already contains the 6 W of router plus the 15 W the outlet passes through, and the engine budgets each PSE independently — so 2453's eth1/eth2 pair may double-count unless the engine couples them.

Not merged, not rebased, nothing closed from this side. Whichever of the two lands first, the other must be rebased and re-read by its own loop.

**Blocked: superseded-by-collision with #2453 — needs Jay's decision, not mine.** Survey found #2452 (rev loop, `407e01f8`) and #2453 (catalogue loop, `c94d5ab2`) open at the same time and both editing `devices/teltonika/otd140.json`. Each is individually green on its head SHA and individually clean into `eab52764`, but they are mutually exclusive: ``` git merge-tree --write-tree tmpcheck2453 tmpcheck2452 CONFLICT (content): Merge conflict in devices/teltonika/otd140.json ``` They agree on the headline (eth2: `poeOut: 15`, `poeOutClass: "PoE"`, `poe.budgetW: 15`, 802.3af, 15 W not PoE+) and diverge on two things: 1. **eth1 `poeIn`** — #2453 changes `12.95` → `21` (flyer v1.8 "PoE In: 802.3af/at … max 21W"); #2452 leaves `12.95` and says in its notes that the PoE-in class is deliberately NOT re-modelled. 2. **Sources** — #2452 cites datasheet v1.20 (2026-07-24, kamami mirror); #2453 cites flyer v1.8 (2024) plus the Teltonika wiki pinout. Both quote the same "15 W / 802.3af Alternative B" and the same "< 21 W PoE Max" row. #2453 also carries `devices/teltonika/otd144.json`; #2452 carries a new `test/otd140-poe-out.test.js` (130 lines) that #2453 does not. Question for Jay, because it is about behaviour and not about conflict text: **is `poeIn` the negotiated class draw (12.95 W, class 0) or the inlet ceiling (21 W)?** If 21 W is the inlet total, it already contains the 6 W of router plus the 15 W the outlet passes through, and the engine budgets each PSE independently — so 2453's eth1/eth2 pair may double-count unless the engine couples them. Not merged, not rebased, nothing closed from this side. Whichever of the two lands first, the other must be rebased and re-read by its own loop.
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-21 18:08:37 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-21 18:09:20 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-21 19:02:40 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-21 19:30:26 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-21 21:31:14 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-21 22:01:29 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-21 22:01:45 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-21 22:30:53 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-21 23:01:20 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-22 01:02:14 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-22 02:01:30 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-22 02:30:48 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-22 03:31:04 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-22 04:01:41 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-22 04:01:43 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-22 04:30:52 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-22 09:01:09 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-22 09:30:30 +00:00
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-22 12:01:56 +00:00
forgejo-actions canceled auto merging this pull request when all checks succeed 2026-11-22 12:30:25 +00:00
All checks were successful
CI / Server smoke and Docker build (pull_request) Successful in 34s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 6s
CI / Test (Node 22) (pull_request) Successful in 1m6s
CI / Test (Node 18) (pull_request) Successful in 1m8s
CI / Test (Node 20) (pull_request) Successful in 1m9s
CI / Arm auto-merge (pull_request) Successful in 2s
This pull request can be merged automatically.
This branch is out-of-date with the base branch
You are not authorized to merge this pull request.
View command line instructions

Checkout

From your project repository, check out a new branch and test the changes.
git fetch -u origin rev/1121-1642:rev/1121-1642
git switch rev/1121-1642
Sign in to join this conversation.
No reviewers
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
jaylouisw/packetforge!2452
No description provided.