[Device]: OptiXaccess S0316 #785

Closed
opened 2026-11-08 06:20:41 +00:00 by jaylouisw · 2 comments
Owner

Vendor

Huawei

Exact model name

OptiXaccess S0316

Category

Other / not sure

Ports and interfaces

  • Active distribution unit (ADU) for power-over-fiber (PoF)
  • 2x upstream optical ports, 16x downstream PoF ports

PoE budget (watts)

None (no PoE output)

Datasheet or product page URL

https://e.huawei.com/en/products/optical-access/optixaccess-s0316

What do you want to model with it?

PoF scenarios: power + optical signal delivery to ONUs over PoF cables.

Capabilities

Power-over-fiber distribution; optical power split

Before submitting

  • I searched the catalog and existing issues, and this model is not there yet.
  • The port list above comes from a datasheet, not from memory or another wiki.
### Vendor Huawei ### Exact model name OptiXaccess S0316 ### Category Other / not sure ### Ports and interfaces - Active distribution unit (ADU) for power-over-fiber (PoF) - 2x upstream optical ports, 16x downstream PoF ports ### PoE budget (watts) None (no PoE output) ### Datasheet or product page URL https://e.huawei.com/en/products/optical-access/optixaccess-s0316 ### What do you want to model with it? PoF scenarios: power + optical signal delivery to ONUs over PoF cables. ### Capabilities Power-over-fiber distribution; optical power split ### Before submitting - [x] I searched the catalog and existing issues, and this model is not there yet. - [x] The port list above comes from a datasheet, not from memory or another wiki.
Author
Owner

Modelled - devices/huawei/optixaccess-s0316.json, commit 6ff7be0 on fix/sfp-plus-connectors (the .html URL again - the extensionless page renders only navigation).

The spec table is unusually complete for a box this size:

  • NNI: 2 x SC/UPC - two upstream optical ports;
  • UNI: 16 x Hybrid SC(PoF) - sixteen downstream ports carrying both the optical signal and power;
  • PoF output power: max 30W per port (PoE/PoE+), total 400W.

Three things this one gets right that are usually guessed:

  1. The PoE budget is quoted, not borrowed. Here the 400 W really is the PoF output figure the datasheet names, so poeBudget: 400 with poeOut: 30 on each UNI port is a published claim - unlike #527, where "400 W" was a supply rating and only volts and amps were published, so no budget was declared;
  2. UPC, not APC. This is the one PON-family device whose ferrule polish the vendor states - SC/UPC on the NNI - so the ports are sc-upc instead of inheriting the sc-apc convention every OLT in this catalogue uses. The sixteen hybrid SC UNI ports match the hybrid SC cords already in catalogue/optics/hybrid-patch-cords.json, whose sc-upc end is what makes a PoF run legal;
  3. No split ratio is invented. The page says the splitter is integrated but never states the ratio; 2 NNI to 16 UNI could be two 1:8s or one 1:16 with a protected feed. Declaring either would put a wrong loss figure in front of someone budgeting a PoF run, so splitRatio is absent.

One assumption, flagged in notes: the datasheet publishes no optical rate. The ports carry GPON to ONUs, so they carry the GPON ladder [1244, 2488] used by every OLT and ONU here - a family convention, not a figure from this page.

Modelled - `devices/huawei/optixaccess-s0316.json`, commit `6ff7be0` on `fix/sfp-plus-connectors` (the `.html` URL again - the extensionless page renders only navigation). The spec table is unusually complete for a box this size: * **`NNI: 2 x SC/UPC`** - two upstream optical ports; * **`UNI: 16 x Hybrid SC(PoF)`** - sixteen downstream ports carrying both the optical signal and power; * **`PoF output power: max 30W per port (PoE/PoE+), total 400W`**. Three things this one gets right that are usually guessed: 1. **The PoE budget is quoted, not borrowed.** Here the 400 W really is the *PoF output* figure the datasheet names, so `poeBudget: 400` with `poeOut: 30` on each UNI port is a published claim - unlike #527, where "400 W" was a supply rating and only volts and amps were published, so no budget was declared; 2. **UPC, not APC.** This is the one PON-family device whose ferrule polish the vendor states - **SC/UPC** on the NNI - so the ports are `sc-upc` instead of inheriting the `sc-apc` convention every OLT in this catalogue uses. The sixteen hybrid SC UNI ports match the hybrid SC cords already in `catalogue/optics/hybrid-patch-cords.json`, whose `sc-upc` end is what makes a PoF run legal; 3. **No split ratio is invented.** The page says the splitter is integrated but never states the ratio; 2 NNI to 16 UNI could be two 1:8s or one 1:16 with a protected feed. Declaring either would put a wrong loss figure in front of someone budgeting a PoF run, so `splitRatio` is absent. **One assumption, flagged in `notes`:** the datasheet publishes no optical rate. The ports carry GPON to ONUs, so they carry the GPON ladder `[1244, 2488]` used by every OLT and ONU here - a family convention, not a figure from this page.
Author
Owner

Correction to this spec - two errors, both caught this iteration. Commit ea303fc removes both.

  1. The PoF watts are no longer machine-readable. I had put poeOut: 30 on the sixteen UNI ports and poeBudget: 400 on the device. test/poe-inference.test.js caught it with its catalogue-wide rule: no port on a non-copper cage may source power. That rule is right - this engine's PoE model is a copper model, computing PSE class, per-port output, a shared budget and the PD draw across Ethernet pairs, and sc-upc is an optical cage. Power over Fibre is real and this box really does 30 W a port / 400 W total, but a machine-readable claim would have the engine evaluate a power budget over a medium it does not simulate. The figures now live in notes, and docs/DEVICE_REGISTRY.md §"What the catalogue cannot represent" says why: making PoF first-class needs a PoF media id, PoF cords and power on the optical rail - a feature, not a spec file.
  2. The hybrid-cord claim was wrong. I wrote that these sixteen hybrid SC ports "match the hybrid SC cords in catalogue/optics/hybrid-patch-cords.json". They do not: hybrid in those cords means ferrule polish (SC/UPC one end, SC/APC the other, one factory assembly - the issue #18 case), not power. There is no PoF cord in this catalogue at all, which is the same gap seen from the other side.

What stands unchanged and verified: 2x SC/UPC NNI, 16 hybrid SC UNI, sc-upc because the vendor states UPC on this box, no splitRatio (2 NNI to 16 UNI is genuinely ambiguous), and the GPON ladder flagged as a family convention rather than a figure from the page.

**Correction to this spec - two errors, both caught this iteration.** Commit `ea303fc` removes both. 1. **The PoF watts are no longer machine-readable.** I had put `poeOut: 30` on the sixteen UNI ports and `poeBudget: 400` on the device. `test/poe-inference.test.js` caught it with its catalogue-wide rule: *no port on a non-copper cage may source power*. That rule is right - this engine's PoE model is a **copper** model, computing PSE class, per-port output, a shared budget and the PD draw across Ethernet pairs, and `sc-upc` is an optical cage. Power over Fibre is real and this box really does 30 W a port / 400 W total, but a machine-readable claim would have the engine evaluate a power budget over a medium it does not simulate. The figures now live in `notes`, and `docs/DEVICE_REGISTRY.md` §"What the catalogue cannot represent" says why: making PoF first-class needs a PoF media id, PoF cords and power on the optical rail - a feature, not a spec file. 2. **The hybrid-cord claim was wrong.** I wrote that these sixteen hybrid SC ports "match the hybrid SC cords in `catalogue/optics/hybrid-patch-cords.json`". They do not: *hybrid* in those cords means **ferrule polish** (SC/UPC one end, SC/APC the other, one factory assembly - the issue #18 case), not power. There is no PoF cord in this catalogue at all, which is the same gap seen from the other side. What stands unchanged and verified: 2x SC/UPC NNI, 16 hybrid SC UNI, `sc-upc` because the vendor states UPC on this box, no `splitRatio` (2 NNI to 16 UNI is genuinely ambiguous), and the GPON ladder flagged as a family convention rather than a figure from the page.
Sign in to join this conversation.
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#785
No description provided.