Jeremy Harris

The outdoor stuff has a much stiffer and thicker outer sheath.

The main issue with cheap Cat 5 cable is that it kinks very, very easily when pulled. I acquired a load of purple Cat 6 cable (a new building for 900 workstations was flood wired with the wrong type of cable, many km of it, and it was all pulled out and scrapped. . . ) so used that. It was much easier to pull through than the cheap Cat 5e stuff I had, as it's a lot stiffer, because of the internal core that keeps the pairs separated. The stuff I used also had an outer covering that was more like outdoor network cable. I have a feeling it may have been some sort of low smoke and flame cable.

Just to be clear, there is no law that states this. The guidance is that any adverse impact on neighbours must be taken into account during the consultation, but there is nothing that says there should be zero perceived adverse impact. Some perceived adverse impact, like the loss of a view, the partial reduction of natural light levels or the appearance of a neighbouring structure, may be deemed acceptable. It all comes down to the particular circumstances, together with the planning policies that are in force at the time, and how they are interpreted by the planning officer. Windows and overlooking usually have some form of defined planning rule that limits overlooking.

If it's Alternative A PoE I don't think it matters, as that voltage is super imposed on the data pairs, as a common mode voltage. If it's either Alternative B PoE or the sort of bodged PoE down the spare pairs that I'm using to power some stuff, then it won't work. It also won't work for Gigabit connections, as they use all 4 pairs in the cable, whereas 10-100 only uses 2 of the 4 pairs. Not sure that there are any real applications for Gigabit Ethernet in a domestic scenario, though, given that it's massively faster than any likely external internet connection.

Just to close this off, I swapped out the old Netgear router this afternoon for a better (but still really an affordable option) Asus RT-AC66U B1, that cost a bit under £100 delivered (as much as I wanted to pay, really). This seems to have completely fixed our Wi-Fi problems, with my laptop showing a solid 5 bars of signal everywhere in the house, and the car showing 4 bars out on the drive. Best of all, for some completely bizarre reason, our broadband speed has improved by about 50%. I can't get my head around this, as we're still using the same VDSL modem, so all the router is doing is providing DHCP etc. I can't for the life of me see how that can change the broadband speed, but it definitely has (I've done four tests now, and all were much the same). Best of all is that SWMBOs iPad has now connected at 5 GHz (not sure why it never did this when I had 5 GHz enabled on the old router) and she's reporting that it's running a lot faster. Pretty impressed with the Asus, for what is still really a budget router. Dead easy to set up, although the set up doesn't make it that clear as to how you can set the DHCP IP range (needed to change it as we have a few things that only work properly with a fixed IP). My only other criticism is that the router admin page is hopeless at identifying all the connected clients, which led to me having to run IPscan just to check that everything in the house was connected. The router also seems to have some pretty bizarre DHCP rules, as it has set IPs all over the place in the sub-net. Never had one do that before, they usually seem to allocate dynamic LAN IPs from the bottom up, starting with the first device to come alive.

Yes, using the spare pairs, but I've not tried it.

It may have been like the Tesla granny charger. That can charge at up to 32 A, and has a 5 pin waterproof connector at the mains in side, that takes a variety of different short adapter leads. The car comes as standard with a 13 A plug lead and a 16 A commando lead, and sets the charge current to 10 A when the 13 A lead is used and 16 A when the commando is used. Tesla sell an optional 32 A commando adapter lead for ~£40 and when this is fitted the granny lead will charge at 32 A.

I put in 8 Ethernet cables and really should have doubled every run, as there have been a couple of times I could have done with two connection instead of one. I'm inclined to agree that you can't have too many Ethernet cable runs, especially as the cable is pretty cheap. BTW, I think @Onoffs post about minimum cable clip spacing may have gone "whoosh" overhead. . . (hint: Those cables need to be tidied up and clipped properly, in accordance with Appendix D of the OSG and also, perhaps, Section 522.10 of BS7671:2018)

The max current is typically 32 A in the UK. The standard allows for more, but I don't know of any AC charge points that can be set to deliver more than 32 A here in the UK. The maximum current that's available from any AC charge point is advertised to the car by using a pulse width modulated 1 kHz signal on a control pilot wire within the charge cable. This is a low voltage signalling wire that allows bidirectional analogue communication between the car and the charge point (usually - Tesla do things differently. . . ). For AC charge control,the relationship[1] between the PWM duty cycle and the advertised maximum current that's available is that the current (in A) is equal to 0.6 x the duty cycle %. So a 10% duty cycle = 6 A, a 50% duty cycle = 30 A and a 80% duty cycle = 48 A[2]. The practical limit is set by both the supply capacity and the car onboard charger rating. I don't know of any UK EVs that have a maximum single phase rating of higher than 32 A, and generally 32 A is about as much as can be realistically added to a UK single phase supply, anyway, given that this current will be drawn continuously for several hours. For charging from a 13 A outlet, then the charge current limit is determined by the continuous rating of a BS1363 plug. Because the fuse inside a 13 A plug generates a fair bit of heat, the maximum continuous rating of a plug with a 13 A fuse is 10 A, rather than 13 A. For this reason, the portable charge points that are often supplied with EVs (often called a granny charger, because it's so slow) is limited to 10 A maximum. Because of the high continuous current, it's normal practice to over size the supply cable for a fixed charge point, primarily to limit the temperature rise at the terminations and to reduce the voltage drop and associated power loss. For runs up to maybe 20m or so 6mm² cable is used, longer runs may need 10mm² cable. Only 2 core SWA cable is needed normally, as the regs require that a charge point be protected such that any exposed metalwork cannot exceed a touch voltage of 70 V above the local ground reference under fault conditions (such as a lifted PEN conductor). The safest way to ensure this is to just connect the charge point installation as a TT installation, with an earth electrode and suitable DC tolerant RCD/RCBO (either a Type B or a Type EV). There are one or two manufacturers who claim to have ways around the need to install an earth electrode, but every one I've looked at fails to comply with the provisions in section 722 of BS7671:2018. How they can make such claims is beyond me. [1] This relationship only holds for an advertised charge current from 6 A to 51 A. From 51 A to 80 A the relationship is current (A) = (duty cycle% - 24) x 2.5 [2] 48 A/80% is higher than any UK EV or single phase AC charge point can deliver.

Any EV, nothing to do with the make. The charge point control protocol that they all use when AC charging only allows the charge current to be advertised from 6 A minimum to (normally, it can going higher in some countries) 32 A maximum. The problem is when a smart charge point starts to vary the advertised current available in order to make charging better fit a varying tariff. Not all EVs play properly with charge points that do this, as it was never contemplated when the standard was written, and with 6 A as the lowest current it means that the charge point needs to turn off if the tariff, or something like PV generation, changes to indicate that less than 6 A would be useful. This then requires the whole handshaking cycle to be gone through when charging recommences, and some EVs just don't respond to that.

It wouldn't make much, if any, difference, as the number of EVs with 3 phase chargers is limited and those that do only charge at 16 A per phase. My car is one of those that does have a three phase charger onboard, but it only charges at 11 kW from 3 phase, versus 7 kW from single phase (because it can charge at up to 32 A on single phase). This isn't an unusual arrangement; there aren't many EVs around that can charge at the max that AC charging can deliver, which is about 22 kW at the moment (32 A on three phase). Car charging is a real pigs ear when it comes to some of the ToU tariffs, as the lowest charge rate possible is 6 A, single phase (so about 1.4 kW) and smart charge points just don't seem to work very well with some cars, and not at all in the case of the Tesla Model 3. It will not charge at all under normal smart charge point control, much to the annoyance of all those that have had to have a smart charge point installed to get the OLEV grant. Other makes have a few issues with smart charging, too, and it seems that there's a fair way to go before all the problems around smart charge points and the different ways that EV manufacturers have implemented their charging systems are resolved. The only half-way reliable system seems to be to use the car's built in charge scheduling, and that really restricts charging to tariff like E7.

This illustrates well how an individual usage pattern can make a massive difference to the best choice. For us, Agile would have cost £68, Go would have cost £76 and our existing E7 cost £57. The problem for us with Go, and to a lesser extent Agile, is that we tend to need the cheap rate for the full 7 hours of the E7 period. Car charging, in particular, doesn't fit well with either Agile or Go. 4 hours at the maximum charge rate of 7 kW is only just over 1/3rd of a charge for the car, and I can often need a bit more than that after a day out somewhere. The UFH also needs more than 4 hours to get enough heat into the slab. If I increase the flow temperature to reduce this time the room temperature tends to overshoot a lot the next day.

Although UFH is always less efficient than a heating system like radiators or warm air, because it increases the temperature differential across the floor insulation, and hence increases the heat loss, it is worth putting some real numbers in to a typical case to see the impact of that. Increasing the floor surface temperature to 23°C with a room temperature of 21°C, and a ground temperature of 8°C, delivers about 18.8 W/m² of heat to the house, more than enough to keep a reasonably well-insulated and airtight house warm in cold weather. Assuming a floor U value of 0.11 W/m².K, then the additional heat lost to the ground by having the floor surface at 23°C rather than close to room temperature, 21°C, is ~ 0.22 W/m². The total heat loss through the floor will be about 1.65 W/m² with UFH, or about 1.43 W/m² without UFH (for the case above). If heating the house with a heat pump, with a COP of 3 and an electricity cost of £0.16/kWh, then the additional heat loss from losing an extra 0.22 W/m² in cold weather (the loss will be a lot less in mild weather) would be about £0.0026/m²/day. So for a 100m² heated ground floor area the additional cost would be about £0.26/day (in cold weather). Averaged out over the heating season the additional cost is likely to be around 1/3rd of this, maybe a fair bit less, (based on our experience of the ratio between the average and peak heating requirement in winter). The question is then whether that is a price worth paying or not. For our house, with ~75m² of ground floor UFH and an annual heating requirement of about 1,550 kWh, the additional cost of using UFH over the warm air system we have (but don't use) is about £0.85, in terms of paying for the additional heat loss the UFH creates. One advantage the UFH gives us is that it allows us to heat the floor slab using E7 electricity overnight, so it acts as a storage heater through the day. That does mean that, allowing for our heat pump COP of about 3 (it's usually better than that) we pay around £0.027/kWh for the heat delivered to the house. If using another form of heating it might not be possible to take advantage of cheap rate electricity, which would increase the running cost by ~80% or so.

Shouldn't need any sort of driver, as it has a standard UART serial interface, running at 9600 baud. Might need level shifting, though, as it's a 5V device. Here are the data sheets I have for it: 6890478_T6004_Application%20Note.pdf UART_SPI_6004_X04_Protocol_02.pdf Looking at my code, it just spits out the following 8 bytes at 9600 baud to the Telaire module to ask it to send data: $FF $FF $FE $02 $02 $03 $76 $05 and then receives back 6 bytes of data (also at 9600 baud) from the module. I've just ignored the first 4 bytes and concatenated the last two bytes into a 16 bit word that is the CO2 ppm value. I went on and converted this to five ASCII bytes, as both the data storage and display uses ASCII. I'll have a dig around and see if I can find one.

@joth I may have a spare Telaire 6004 somewhere. They have a serial interface running at 9,600 baud IIRC. I could try and dig one out if I can remember which box I put them in when we moved. . .

I've been using some surplus Telaire 6004 NDIR modules that I bought a while ago for some time now. They need the CO2 concentration to drop to 400ppm once every couple of weeks to do their "intelligent calibration" thing, and realistically the house never gets below about 420ppm, so there is always a small calibration error, but the results are good enough as an indicator of air quality. I can cross check the fixed sensors with a portable unit I built, that uses the same sensor. By leaving the portable unit outside when there is a decent breeze I suspect the CO2 concentration drops to as low as it's likely to get. So far I've never seen more than about a 10ppm difference between the sensors. Right now the sensor in the hall here is reading 580ppm, which seems about right.

Some do and some seem to need an external pump. Not sure why some ASHPs don't have an internal circulation pump, but then perhaps that's no different to boilers, some of them have internal circulation pumps and some don't. Because the UFH pump only pushes water around the UFH loops, it doesn't normally push water around the primary pipe loop. The primary loop can be shut down or restricted and the UFH pump can continue to push water around the loops. I use this feature to equalise heat around the floor, by running the pump when the ASHP is off. It works well to even out the floor temperature, by drawing heat from areas of the floor than might get warm from solar gain and moving it to areas of the floor that are cooler.

The manifold pump only pumps around the loops though. You'll need a pump to push water through the heaters as well.

Pretty sure it's a Landini, and the reg is probably right, as Landini were owned by Massey-Ferguson. Looks like it has grass tyres, so may well have been used for looking after sports fields, maybe a golf course (or a very large lawn). Not sure of the Landini model, though. Might be a Landini 7500, but with the grill changed and the lights moved outside the grill. Could well have been a mod done to give better light coverage for a specialist grounds keeping type job, as the grill looks to be non-standard.

I quite like the soldered heat shrink ones. Last ones I used were supplied with the cable jointing kit for our water pump. Easy to use, just make a staggered joint (as you've shown, @Onoff) but you just push the wires into the soldered shrink butt splices and heat them up with the heat gun set to a high temperature, so that they shrink and make a soldered joint at the same time. I also prefer the adhesive lined heat shrink over the other stuff, as well as sealing things up it also adds a fair bit of mechanical strength, in case the cable gets pulled.

Crimp new wires in a staggered joint and use heat shrink sleeving to provide the double insulation that's required. This has to be a maintenance free connection, so screw terminals (as used in that connector box) aren't allowed. I have a feeling that @Onoff posted some photos of a proper crimped and sleeved inaccessible cable joint some time ago, but I can't find them from a quick search. Not hard to do, and by staggering the joints you can keep things from getting too bulky under the final sleeve. If you have room, then Wago connectors fitted inside a Wagobox are classed as maintenance free, I believe. There's also the Hager/Ashley maintenance free connectors (same product, different name) like this: https://www.tlc-direct.co.uk/Products/ASJ803.html

We had to get a BT OpenReach line moved for exactly the same reason, it was too low and at risk of being damaged by one of the diggers. It was a complete nightmare, and took several months to get resolved. In the end I managed to get hold of the local OpenReach engineer, who was very helpful. He dropped off cable and ducting and agreed we could lay it underground around the plot, so we didn't have any cables oversailing the plot at all. We did this, and left the cables coiled up ready for OpenReach to connect, but they took months to come and take down the overhead wires and re-connect the new cables. In the end we were advised that the only way to get the job done in a reasonable time frame was to have an "accident", where one of the diggers "accidentally" snagged one of the overhead wires. We could then call the emergency helpline. When the chaps turned up about an hour after this "accident", they took one look at the new underground cables coiled up and realised that all they had to do was connect them and remove the old overhead wires completely. My advice would be to try and start negotiations with OpenReach about a year before you need the cables to be moved . . .

The annoying thing is that these short dropouts in the 2.4 GHz signal weren't noticeable just from looking at the unit, or looking at the log files it keeps. It looks very much as if the fault may be in the transmitter module within the router, causing it to shut down for maybe a second every now and then. My guess is that the buffering in connected devices can sometimes cope with this and sometimes can't, and that this may depend, to some extent, on range (just because range has a fairly hefty impact on data throughput). The worst problems we were having were with the internet radios. They seem very fussy about getting a continuous clean signal, and had the nasty habit of just dropping out and doing the "re-connecting" thing when their buffers ran out. My laptop would periodically fail to connect, but I'd not twigged that this was a router problem, as the laptop has pretty rubbish Wi-Fi anyway (consequence of the thing having an all metal case. . . ).

I've spent a happy few hours logging data from a laptop, using another wireless sniffer, Vistumbler. I've also tried switching over the ancient D-Link 802.11g router (so only 54 Mbps) in place of the much faster Netgear one that we've been using for two or three years. What I've found is that the Netgear router regularly just loses transmit power for maybe a second, anything between once a minute to once every two or three seconds. Seems to be heat related, as when I swapped over to the old D-link (which got rid of all the wireless problems we've been having) and then switched back to the Netgear, the Netgear ran perfectly from cold for maybe ten minutes or so, then started just dropping the signal again. I tried turning off the 5 GHz option, as the Netgear uses the same module for both 2.4 GHz and 5 GHZ, and that slowed the rate of the dropouts at 2.4 GHz a fair bit, but they are still there. So it looks as if 99% of our Wi-Fi problems may well have been from this weird fault with the router. I've switched back to the ancient D-Link one and that seems to be rock-solid everywhere in the house, and even out the front where my car's parked. TBH, the fact that it's only 802.11g isn't at all noticeable, as that's still a lot faster than our "broadband" (accepting that I doubt it ever actually delivers 54 Mbps). Clearly the first thing to do is just to get a new router, as it seems clear the Netgear has a fault. Hopefully replacing it will fix the problems we've been having, without having to faff around with additional APs.

Our experience is that CO2 doesn't track RH at all well in this house. It did in our old house, but that was less well ventilated and often had much higher CO2 levels, in particular. External RH variation seems to dominate the internal RH here, whereas at our old house I think that the apparent correlation between RH and CO2 concentration may have been because both were largely driven by occupant expired air.