Dreadnaught

The second video is up: Credit to Ben Adam Smith's twitter feed for the link.

Its the hardware and software within the APs (Wifi access points) that matters. The only effective solution is to use APs that support WiFi roaming and seamless handoff, 802.11r or a proprietary equivalent. The AP requires a special controller functionality within and the vast majority of home WiFi routers do not have it. Without WiFi roaming, using the same SSID for multiple WIFI sources is generally not a good idea as, as you have found, it doesn't assist with roaming and it can make debuging more difficult. Which APs are you using? Are they true APs or just home routers connected by ethernet cabling and set to bridge mode, each creating its own sub-net and with multiple NAT. Or even worse WiFi extenders connected wirelessly, again with multiple sub-nets and NAT, and likely a halving of data throughput?

Just picking this thread up again after more than a year. Pondering a green roof for my coming build. @vivienz did you get anywhere with your idea? Has anyone currently got a green roof? Anything good or bad to say about them? Cost? Maintenance? Do they play well with roof lights? Complete beginner here.

Thank you all. Excellent points. Plenty of new thoughts to mull. Notably… splays decrement delay of vacuum panels may be low but actual heat flow will likely be tiny the risk of puncturing vacuum panels alternatives to vacuum panels, e.g. aerogel a green roof giving evaporative cooling and being an additional thermal buffer. (Neighbour would like the look of it too) I am now minded to avoid vacuum panels and be armed with all these new ideas instead. Thank you.

The issue is less about an initial connection but more about how long a device will hold on to a dwindling network as you walk away. With multiple access points ("AP"), it is far better to have an AP with the technology to switch between APs as you roam. Most domestic (i.e. non-commercial) APs don't have this. It makes all the difference. Ubiquiti have their own technology for WiFi roaming but also support 802.11r. The same applies if you manually give the same SSID to two APs. Without the ability to roam, it will still hold on to the first network for too long, until the signal strength dwindles. In the reality of a home environment it often means that it never switches automatically at all. And having always to manually switch networks is an annoyance, a distant second best in my experience.

As for Ubiquiti kit, you should have plenty of change from your budget. UniFi AC Lites cost about £80 each and you can pick them up on eBay second hand from time to time for about £50 or so. Suggest to buy 2x UniFi AC Lite's and nothing else. (Plus a laptop to do the setup remotely using the free Unifi software.) Regarding "2x or more", it depends on coverage areas and signal attenuation by thick walls. Suggest trial-and-error. Start with 2x and buy more if you need them. In a modern house, two did the whole house. In a big older house with brick internal walls, 3x were needed. If you need an weather-tight outside AP, they are available too from the "mesh" range (e.g. UAP-AC-M). Optional extra: a PoE ethernet switch (made by Ubiquitu or a third party, doesn't generally matter) so that PoE injectors, which are included in the box, are not needed at each access point. Initially you don't need a UniFi Cloud Key, which is for remote monitoring of your network and more commonly deployed for commercial installations.

Thanks @SteamyTea. Would it be more meaningful to say vacuum panels have a (near) infinite decrement delay factor, not zero as in the linked information? And, even more relevantly, in your view should I not be concerned about decrement delay in a thin roof with vacuum panels?

From @JSHarris and others I have learnt about the importance of decrement delay as a characteristic of the fabric of a house in providing internal comfort. Background. My build may well be a single-floored, flat-roofed building of contemporary design. As a rear-garden plot surrounded by other dwellings, it has precious few sight lines and instead will have a profusion of roof lights to let in light. As a (near) Passive House, the roof will be thick and there is a concern that the roof lights will give the impression from inside of a house deep underground. One suggestion to address this is to thin the roof, to use high-prefromance insulation such as vacuum panels in the roof to improve the aesthetic look instead of cellulose filled I-beams. So far so good but I worry that a roof with vacuum panels will have a low decrement-delay factor. So to the subject decrement delay. As I have learnt, a cellulose-filled roof would have a welcome decrement factor. But what about vacuum panels? I have done a little google-ing and came across the following summarised from here: Thus to my question: how can Vacuum Insulated Panels have a decrement delay of 0 hr? I understand that the decrement-delay factor is product of λ (lambda), which is very low for vacuum panels. But is also related to specific heat capacity and the density of the material concerned and I do not understand how to consider these two for a vacuum insulated panel. I wonder, can anyone enlighten me? (For those interested in learning more about decrement delay factor, I found this explanation a help: http://www.greenspec.co.uk/building-design/decrement-delay/)

A tweet today: What significance does this have for us self builders, if any?

+1 for UniFi. Done two installs of UnIfi AC Lite's recently and very impressed with the outcome in each case. Rock solid. EDIT: By the way, if you do go down this route, feel free to ask for advice. As a product often used for commercial installations, the plethora of options can be a little dauting. A basic install for home use is much simpler than it may at first seem.

On the nature of PCM34, as Andrew Bessell of SunAmp told us, PCM88 uses Strontium Bromide Hexahydrate. Source reposted by @Alphonsox from the previous forum here: In the letter from Andrew Bessell on the current issues re posted with permission by @le-cerveau in the thread below it says only that SU34 is a bromide with a different metal ion, i.e. not Strontium.

Out of interest, why?

Just to add to your thoughts, I found the whole chapter on moisture control in the Passive House Handbook very helpful in understanding these issues. In that chapter, and for a sandwich layers of materials in, for example, a wall and from inside to outside the layers ideally need to have a progressively lower vapour permeability. It states that vapour permeability of the material on the outer surface should be 5x that of the internal surface as a rule of thumb. For example, wood fibre board is very vapour permeable and so well suited to being the outer layer. It also helpfully states "Moisture within the fabric of a building is the most common cause of building failure".

No, its not too late. But that would be to forego a plot in a location that arrives perhaps once a decade. Its location is that good. For me, its worth navigating the admittedly daunting set of complexities for that reason. Forgive me for not saying more about the location now. I will be happy to do so once the ink dries on the paperwork.

Yes, of highly contemporary design. Made Ground (0 - 2m or so) over river-terrace deposits (then 0.2 – 1.5m) to Gault clay (at 2 - 3 m total depth). Recommendation is for screw piles to a suspended slab over a sub-floor gap. I have no idea at present how I am going to have the passive-house levels of insulation in the floor. (Oh, and the plot is between flood zones 2 and 3. The full card deck of issues!)

There are big trees adjacent to the site (a Horse Chestnut Tree, a Plane Tree, two Beech Trees, all mature) and the neighbours kicked-up a fuss about each of them. Its a conservation zone in the middle of the city so all those trees effectively have TPOs. And more than half the site is a no-dig root-protection zone. I did say the site has all the constraints imaginable

Yes, would love one but planning again say no. Tree roots galore! I have the world's most difficult site. It has the full card deck of constraints.

Thanks @lizzie. That's really helpful to visualise. I suspect the vendor thinks of his 50cm as a concession and we will have to move our property back 70 cm too to get a gap of about 120 cm. I have to weigh-up whether that is a fair deal in the wider context. Thinking…

Good challenge. Hmm, some more thought is needed. Thank you all. That has been very helpful!

I worry more about the construction phase. Can I build a timber frame house (single floor) so close to a boundary, with just a 50 cm gap? Also, does anyone know of any planning principles about how close a new build can be to a boundary? I have heard of a 1.2m figure mentioned, but I am not sure if that's just a guideline or whether its a requirement.

Sadly, planning wont allow it. No flue. Could be rainwater pipes.

Not sure yet. Architect has not done their work. No windows on that side, just roof windows (which hopefully would be > 1m from the boundary as the crow flies). Could do. But swings and roundabouts. Its part of a wider deal.

The fence will be the responsibility of the neighbour. Regarding maintaining my wall, I imagine it would mean squeezing in the 50cm gap. Is that too little? Its a single floor (bungalow) but good point about ladders for cleaning the roof windows on that side. It could be, but I have a tiny plot so keen to use it all.

This will be a new build bungalow. I am negotiating boundaries on my plot with the vendor. He is proposing to provide a little extra space to erect a fence beyond the outside wall of my future house (the boundary fence was missed off the plot as advertised). The construction method of my house will be timber frame, wooden I-beams (passive house) and rendered. He has suggested 0.5m (50cm) of gap between my outside wall and the fence (a fence that will be in place throughout the build). How much space does a contractor need to erect such a wall? Is 50 cm enough? What about scaffolding?

I'm amidst buying the plot.