Fibre Optic Network LInk

Posted on by
Reply

Introduction

As mentioned in an earlier post about a Fibre Connection Between the House and the Outbuildings there’s an 85m length of 8-core single-mode fibre running between the two buildings. While it’s possible to buy custom-made pre-terminated fibre, the relatively bulky end connectors make this much more difficult to pull through small ducts than unterminated fibre (which is only about 6mm diameter) – so while the fibre is already installed it’s not usable until it has been terminated. That wasn’t initially a problem since the fibre is really just future-proofing for the time when a network connection faster than 1Gb/s is required (or multiple physically-separate connections are required; the single CAT5E Ethernet cable already carries multiple logically-separate networks using 802.1Q VLAN tagging). Despite being close to the nominal length limit of 100m, this CAT5E cable was happily synched at 1Gb/s.

These days, single-mode (OS2) fibre is recommended over multi-mode (OMn) and rather than attempting to fit a connector to a 9 micron fibre on-site, the preferred approach is to fusion-splice a length of fibre which has already been terminated into a connector at one end, known as a ‘pigtail’. While automated fusion splice machines are expensive, they’re not uncommon and they’re designed to form high-quality splices even when used in challenging conditions (e.g. outside). For example, they’re carried by all the OpenReach technicians who install FTTP Broadband.

The decision to install single-mode fibre (rather than multi-mode) was straightforward, since single-mode is definitely “better” and is no longer significantly more expensive. However, there are several further design decisions required for other aspects of the fibre installation.

Simplex or Duplex?

Each of the 8 fibre cores can either be used as one half of a duplex link, where one core is used for Transmit and another for Receive (at a wavelength of 1310nm for single-mode fibre) or one core can be used for a bi-directional link (typically using 1310nm for one direction and 1550nm for the other). I took the view that 4 separate duplex links over the 8 cores was plenty of capacity – and dedicating a fibre to each direction seems less of a technical challenge than the bi-directional approach (and so less prone to issues, maybe?). Having said that, FTTP Broadband uses one single mode fibre for bi-directional communications and that works fine.1

Since the TX fibre at one end of a duplex link becomes the RX fibre at the other, it’s necessary to arrange for the fibres to be ‘swapped’ at some point along the route. Best practice is to swap them in the fixed section of the link so that it doesn’t matter whether the patch leads are straight-through or also ‘crossed’. (Duplex patch leads are in fact ‘crossed’ too, as standard, but provided there’s an odd number of crosses in an end-to-end link that’s fine.) This convention is part of the TIA 568 fibre cabling standard.

SC or LC Connectors?

There’s a huge range of fibre connector options but in practice the SC and LC types are the most common, with the older, metal bayonet-style ST connectors also fairly widely available . For duplex links two connectors are physically paired together. The smaller LC connectors are universal on duplex SFP transceivers and are increasingly common but the larger SC connectors seem slightly more robust for use in patch panels, so I settled on terminating the ‘fixed’ fibre links with Duplex SC connectors at each end.

Fibre patch leads with SC connectors at one end and LC connectors at the other are widely available – if anything more so than SC-SC patch leads.

DIY or Professional Fusion Splicing?

Automated fusion splice machines are expensive to purchase but can be hired for a few hundred pounds per week – if you know how to use them (and the related ‘cleaver’ tools). Then test equipment is required to ensure the spliced link performs to specification. I elected to commission local fibre optic specialist Fibrecomm Solutions Ltd to fusion-splice the pigtails onto the 8-core fibre (16 splices in total) and then test the links to validate the results.

Media Converters or SFP Modules?

Modern network equipment such as switches or routers is typically made compatible with fibre cabling via the inclusion of SFP or SFP+ sockets. It’s then the SFP(+) adaptor which takes account of whether the fibre is single- or multi-mode, simplex or duplex etc.2

Older equipment tends to only offer RJ45 sockets for Copper cabling – so then a separate Media Converter is required to convert between Copper and Fibre cabling. There’s quite a range of these available from different manufacturers. Some have SFP sockets which take optical modules but others have their own SC or LC sockets and tend to be cheaper. A few Media Converters accept a power feed via PoE.

I opted to use two of the TP-Link Omada MC210CS 1000Base-T to 1000Base-LX/LH Singlemode Fibre Media Converter, SC Connector (15km) units which incorporate pairs of SC sockets for duplex single-mode links. These take a 9V DC feed from a separate power adaptor but can be adapted for supply by PoE with a separate 9V-capable PoE Splitter such as the TP-Link Omada TL-POE10R, enabling them to be powered by a remote UPS.

  1. Out of interest, FTTP Broadband uses single mode fibre in a ‘simplex’ configuration (i.e. different frequencies along a single fibre for Transmit and Receive). The end connectors are coloured Green, indicating they are APC-style (rather than Blue for UPC-style). Another difference is that to avoid the need for active equipment ‘in the field’ there are passive optical splitters / combiners that form a Gigabit Passive Optical Network (GPON). ↩︎
  2. The SFP(+) adaptor receives power from the network equipment, which means that if the network equipment is powered via a UPS then the fibre link will stay up even if the grid power gets interrupted. ↩︎

Driving Rain Penetration through East-Facing Garage Wall

Posted on by
Reply

One minor disappointment with the Outbuildings is that the East-facing gable wall of the garage block has been getting damp coming through to the inside on a few occasions when there’s been persistent driving rain. This is a ‘solid’ wall (i.e. without a cavity) about 210mm thick, formed of brickwork on the outside and blockwork on the inside. Since it’s in the unheated-and-uninsulated half of the building this wall didn’t need a cavity for insulation purposes but on reflection it would have benefitted from an uninsulated cavity to stop penetrating damp coming all the way through.

Inside of the East-facing wall of the Garage block, in mid-November 2025
Outside view of the same wall

Based on the internal photo, the damp appears to be worse at the bottom than the top (and the left (i.e. North) rather than the right) but later in the winter the whole wall appeared to be more evenly damp. Since the bricks used throughout the Outbuildings are very tough – almost Engineering Brick grade – the builders’ conclusion was that the water probably isn’t coming through the bricks themselves but is more likely to be mostly coming through the mortar. It’s possible that the additives included in ready-mixed mortar, to help that remain workable for several days, might make it less waterproof than site-mixed mortar. Over the winter, the ambient humidity tends to remain high and even with the garage door left open the wall didn’t dry quickly.

It’s not clear how much consideration was given to this risk at the design stage. The fact that the wall faces due East might have been expected to mean it would be relatively sheltered from the prevailing wind – which is generally assumed to come from the South-West. Research into prevailing wind directions (such as Seasonal variation of the prevailing wind direction in Britain) confirms that South-West is the dominant wind direction (based on data from a location about 90 miles away). With East Midlands Airport being only a few miles away, their weather data is generally directly applicable and the Met Office publish weather data for UK airports – including the ‘Windrose’ shown below, which indicates the wind is indeed generally from the West and South-West. In particular, the strongest winds are mostly from the South-West. This expectation probably excuses driving rain on an East-facing wall being considered a significant risk at design stage – but clearly there’s still enough wind-blown rain coming from the East and North-East directions for there to be a problem, on this relatively exposed and unsheltered gable wall.

‘Windrose’ for East Midlands Airport, from https://www.metoffice.gov.uk/services/transport/aviation/regulated/airfield-climate-data

The Architect for the project, acting in their capacity as Project Administrator, concluded that the builders had faithfully built what was shown on the drawings and so were not liable for this being considered a ‘construction defect’. In the end we all agreed that the best course of action was to try applying some sort of water repellent treatment to the outside of the wall – the expectation being that even a modest waterproofing effect would probably be sufficient.

Stormdry Masonry Protection Cream claims a 25-year lifespan and has BBA Approval. As a goodwill gesture the builders agreed to pay £125 for 5 litres of this. There was no point trying to apply it while the weather was cold and the wall was still damp but a warm weekend in early July seemed like the ideal time, with no rain having fallen for a day or two and none being forecast for more than 48 hours.

In the tub, the Stormdry cream is bright white and roughly the consistency of mayonnaise. It cannot be poured in any sort of controlled fashion so it was necessary to ‘spoon’ it into a roller tray. It went on well with a brush around the edges and a long-pile 9″ roller (intended for masonry paint) made it easy to apply over a large area. The white cream turns into a clear liquid once applied which then soaks into the masonry.

A 5 litre tub with a nominal coverage of 25m2 actually covered about 32m2 – no doubt helped by the fact that the bricks are not very porous. The last few bits from the tub were brushed into some of the mortar joints that appeared to be drying quicker than the rest of the wall, suggesting they needed more treatment.

Slightly darker brickwork on the right of the expansion joint, about an hour after application

The success (or not) of this treatment won’t be evident until the Winter. It’s probably best never to paint the inside of this wall (or any of the ‘solid’ walls) to allow any moisture that does get through to evaporate. If the problems persist a more dramatic solution would be to build an extra wall on the inside, to form a (ventilated) cavity – so that at least the inner wall remains dry.