joth

Yeah, I've given up harping on about this but if you search this forum for "polyphase" you'll see I've been trying to raise awareness of this for a number of years.. 😂

Looks like the ministers heard me! Sort of. https://www.ft.com/content/b0a9eb3e-38f8-4fc4-8a60-72a12142244f UK ministers consider electricity bill discounts for heat pump owners Households replacing gas boilers may not have to pay green levies on power costs I think it would be crazy to key the discount on presence of a heat pump and/or absence of a gas boiler, which needs surveying and monitoring, rather than simply "does this property have an active gas supply", which can be done with a simple grep on a central database. But hey, ministers must create 2 new bureaucratic hurdles for each one they remove

If you have a 3 phase supply anyway, and the price is not prohibitively higher, then yes it is. Splitting your generation over phases means less voltage lift per phase, so less risk of the inverter shutting off if the local supply voltage runs high. I believe this depends a lot on DNO and location. I heard a while back Western power were starting to insist on doing 3 phase install (even if only 1 phase brought out of the head) wherever they have capability for it.

Haven't seen this answered, but, even if you do have different phases to different areas, so long as the whole lot are being supplied via a DNO cut out and single mater, you can still put a 3 phase inverter onto the main 3 phase supply and then divide the phases out to sub distro boards as needed. To a degree this is always going to be the case - there's no such thing as a 3 phase domestic "ring main" socket so some areas of the house will be on different phases. And you'll probably want all 3 phases in the garage anyway for the car lift or, whatever. If you're worried about balancing PV generation to demand across phases, don't. Modern meters to net billing so you're only charged for your net import (or paid for net export) across phases at any given moment in time. Couple more things to note: - you can't buy a 3 phase inverter now "just in case" and upgrade to 3 phase supply later. A 3 phase inverter needs 3 phases to output onto. - A single battery often has a power limit lower than it's capacity, e.g. a 16kWh sized battery may only be able to supply 5kW at any given moment. This also limits it's usefulness for high demand users - do you have DNO approval for 14kW PV ? They may look more kindly to this if on 3 phase. - have you asked for an indicative price for 3 phase? This maybe prohibitive and will answer your question - "The 16kw was advised as part of a recent quote I had." - this is obviously just guess work with so much future expansion of use planned. tldr you can always more than you have, the optimum size depends a lot if you just plan to time shift your PV generation or use it for cheap overnight rate time shifting too. With my tariff (Eon Next) it's actually more financially rewarding to sell all PV generation rather than store it, and then fill the battery from grid each night. But that may change when the tariff is updated next year.....

A pair of relays and a dedicated permanently powered spur/radial https://knowthecode.io/labs/basics-of-data-and-program-circuitry/episode-6/or-gate-relay-logic But honestly if implanting moods I'd use a set of lighting controls designed for the job, but I would say that 😂

An advert for 1GB in early to mid -90s is very plausible. My first PC in 1995 had a 1GB HDD... that was my second hard drive, just a few years after purchasing that 20 MB drive for my Amiga. A period of insane growth, bit like AI expansion is now.

I doubt this very much. 1GB would have cost about $1M in 1985, so 1TB would be more like 1 billion dollars. (Or $100M for 1Tb if you mean bits not bytes - not conventional*). https://humanprogress.org/dataset/average-cost-of-hard-drive-storage-per-gigabyte £6000 would have maybe get you 10 MB mid 80s. In the early 90s I paid £200 for 20MB hard drive - £6000 would have bought about 600 MB. By the end of the 90s I paid about £150 for 4Gb - so £6000 would have got 160GB. It's almost the mid-noughties before £6000 would purchase 1TB - around the time I started working for Google which is where I first started hearing TB referred to in casual conversation. So off topic.... * - of course we should be using MiB GiB and TiB for storage, not MB GB TB but honestly that never caught on did it.

Isn't this topic about a new build self build where UFH can be installed and running costs will be much lower. I really can't see how debating hydrogen in some far off future is a useful thing to consider when building a brand new house at this point in time

No, still evil, because - the lamp can't turn on until some low-threshold is reached (not enough power available when the phase cut ratio too high) - the power level and lux brightness it comes on at will vary from one manf to next, even from one bulb to next within make/model, so you get random spotting on of lights on a slow fade up - it's very laggy - it's inefficient Fundamentally, communicating direct to the bulb's low voltage driver and telling it what PWM to drive at (e.g. via DALI, KNX, or yes even some godawful wireless protocol) is going to give an electrically and visually better result. We're just now in the unlucky state that phase cut is the lowest common denominator communication protocol and commoditised solution for controls to speak to lamps.

Agree their doorbell is a bit of an electrical-mechanical-firmware fail. At least you don't have the previous version with PoE over USB-C (a horrible connector for outdoor installation use), and pisspoor electronics that would melt down and release magic smoke if you enable "mechanical chime" mode when powering it direct from a 12V psu. Do you need a cloud key to run it, or just to setup? In the old days you could use a laptop to run the cloud config software just when needed. And most the "smart" features (like AI object detection) are built into the doorbell itself. You should be able to configure webhook(s) on it to have it ping HA (or whatever) when object spotted or button pressed, and this part at least doesn't need the cloud key. I guess remote answering via app does though. Anyway good news is now you have cloud key you're one of us, and lower marginal cost to buy your next shinny unifi object.

My system low limit is calculated dynamically based on indoor temp and RHI, using https://www.loxone.com/enen/kb/dewpoint-calculator/ However in practice I could hard code it at 13deg and be about the same result.

Are these specifically for LED strip tape? If so, yes I prefer using one central 24V PSU (or one per floor), PWM dimmers on the low voltage side of the PSU and then radials to the LED strips. Size the cable appropriately (1mm2, or 1.5 if the runs are long or LED strips large) and never seen any problematic voltage drops. Put in a maintenance free junction box near the fitting to swap from thick T&E to finer flex cable to the strip itself. This is going the other way: hide the driver inside the light fitting. (It still has one, just integrated into the GU10 bulb). Not really an option for LED tape. But I'm increasingly favouring this in all but high end designs (preferring DALI there), not because it's technically elegant (dimming via mains phase cut is evil) but because the longevity of both the LED bulb, the fitting it is in, is bound to be higher than proprietary driver/emitter combos that may well be impossible to replace like for like in a few years.

What emitters is that with, and where? I've been doing this with UFH (GF) and fan coil (upstairs), I've never found the UFH cooling especially useful, but if you have UFH upstairs then perhaps it is more useful. Our FCUs work a charm though. (Or did, until motor on one failed.... ebay special was not so special, sigh)

Lifespan will be much more than this 10-15 years, it's just the capacity reduces. (Same as PV the peak power degrades over time). So a 10 year guarantee normally guarantees something like 90% capacity remaining. I expect the 15 year guarantee wording says 85% capacity remaining. If you have space and dno approval it makes more sense to plan to add additional battery in a decade or whenever, rather than replace.

I'm having moderate success with the waveshare mmWave module, per https://community.home-assistant.io/t/waveshare-mmwave-module-any-successful-implementations/747900/2 At £3, plus the price of an ESP, it's excellent value, and I have total control over where I send the data. The movement proximity sensing is much more accurate than I would have expected. Downside is the calibration is a bit painful, and everytime I update settings it seems to get in a muddle and need a reboot to resynchronise where it is in Presence sensing. (Also it seems ESPHome has bloated quite a bit and getting a bit unwieldy to use on ESP8266, which is most of what I have in my parts box. ESP32 doesn't cost much more so probably time I bin my spares and start over... getting better SSL support is never a bad idea.) Anyway it was a fun project and working pretty well as an office presence detector. (I'm planning to put it in a case with a big thump button on top to act as quick access to mute music in the room!)