joth

I connected mine from softened water, because that's how @Jeremy Harris did his. (His reasons were to avoid cost of another descaler ,and it makes a better cup of tea, IIRC) After 5 years I've not had any issues with it. Ours is a combi tank so the main kitchen tap is all softened, and we have a dedicated hard water spigot next to it, mostly used for cold drinking water We're in a fairly hard water area - borderline for being within the safe guidelines for drinking the softened water. If I have four mugs of tea (1L water) it's about the same as one bag of crisps. If someone has a medically prescribed low-sodium diet then they should avoid it still.

It's relevant for any external pipework. I think this is more common in F-Gas installs as they tend to be retrofits (in domestic installs at least), and tends to be easier by running pipes all over the outside of the building. (Ugly for sure, but easier & cheaper). I expect retrofits to increasingly use a2a for this very reason. This is the key decision point. If you're happy to pay for it, then an a2a (i.e. some sort of refrigerant based) system is the only way to get the "arctic gale" of an aircon system. Whether you combine one system with UFH & DHW or do two systems (potentially one exhaust air for UFH & DHW) comes down to shopping about, supplier selection and what they are comfortable working with. Cost wise it's probably a wash (one complex system vs two more conventional ones). A a2w heat pump with FCUs is likely cheaper to install and can perform very well (comfort and cost wise) if setup correctly. And is much more DIY-able if that's your truck. But it will never give the same frozen air blast of a2a

If you ask OpenReach they'll tell you only they are allowed to move the master socket. That detail aside... I'd say what everyone else said. Wrap some electrical tape to insulate the wires from each other before pulling back out through the wall. Else if they short they'll spark rather worryingly.

I have a Texecom Premier system, professionally installed, and honestly I can't recommend it. It's hugely antiquated. I posted a short-list of my grumbles with it on the Loxone mailing list 5 years ago -I'll copy below for posterity... The key question is do you need a "Graded" alarm system? this is never financially worth it for "insurance premium savings", the monitor/support contract costs more than the savings. It's really only worth it if you have high value items that the insurer refuses to cover without a graded alarm. Or, you just want the highest level of system for peace of mind. A graded alarm allows for automatic police notification. I found that a waste of time, as the ARC would call me and my 2 other keyholders before doing police callout, and if anyone answers they would basically just say "your alarm is going off" and hang up, no information on what cause or support in figuring out if it's a false alarm. Which was oh so helpful when I was camping on a mountain in another country... So I've now terminated that contract and do my own self-monitored alarm (via Loxone, and a network of mutually supporting neighbours) which is cheaper and far stronger protection. If you DO need a graded alarm, I keep hearing AJAX.systems is the much more modern system, and uses cameras for much stronger confirmed alarms. (Texecom has a camera offering now but honestly I wouldn't touch it given my experiences) Still looking for some first-hand feedback from someone that's lived with AJAX for a few years and see if it lives up to the marketing.

The SolarEdge modbus interface is not great - presenting scalers as value + exponent in separate registers is inherently racy and can create large glitches in readings. I extensively hacked the Loxone integration to work around this -- basically read multiple registers in a single atomic 64-bit register read, then crack apart the words and recalculate the correct values. BUT it's still better than their very low frequency cloud API. What do you mean by At the moment still time based?

Seems like the research was published in January but a load of AI regurgitation has sprung up about it in the last few days. Original paper https://www.nature.com/articles/s41598-026-35389-6 Not sure, the reference "problematic" LEDs in the study were 3000K so it's already comparing incandescent vs warm white LED. Skimming the article all LED lights roll off at about 670nm but you need about 1000nm+ for the IR spectral component they talk about I've always felt the CTT [colour temperature tuning] is bit of a fad, especially in a domestic setting. But I can see a variant having a large benefit of having WW / IR dual channel emitters, and dynamically increase the amount of IR emitted based on various factors. e.g., more IR in winter when the inhabitants will have exposed to less of it naturally, and also when the loss of lighting efficiency is not so problematic as the IR is at least contributing to useful space-heating.

And down-sizing the ASHP to allow these long runs at a low deltaT without short cycling, isn't it? Our ASHP is way oversized so will bordering short cycle (5 min runtime, 10min period) when just cooling the slab to ~16deg. I'm still getting COP of 3.5-4. But it's powered from PV - if I could double the COP it would give me about 30p more SEG payments per day on these hottest days.

This is how I set mine up (6 years ago) on our ecodan, the problem is it's not just the FCU loop that needs condensation proofing but also the primary pipework that includes the filter, pump, sensors, divert valves etc. None of those devices are designed for being insulated, indeed there's a risk of pump burning out if you really did, and they gather condensation as easily as the pipework. Adding a second, much larger, FCU and keeping it at dew point is how I've used it for past 3 years.

Yeah by hot i mean "not cold", c.f. typical (historic) uk climate averages: 15-35°C is the ideal temperature range for LifePo4, which is much hotter than the uk average of 10°C

It's an interesting trade-off, as batteries like it hot and this move will reduce the efficiency and lifespan of the batteries when used in the cold, in winter. If you're only using them for PV self-generated storage, that's probably moot, but if like me you fill from overnight cheap rate then it's rather a hit. That said with 27kWh you have a lot to play with. I hadn't realised batteries could generate so much heat - presumably you moved the inverter too, and that was the actual heat-source? Unlike batteries, inverters do like it cold.

It's not so useful if you want to add on automations of any type - even a button to "turn off all the lights" is much trickier as it needs to inspect the current state and only pulse those that are on. Depends on what the end goal is really

My go-to is the Waveshare 32ch modbus RTU relay - been running a a few for over a year now, very stable. However i can't see evidence of CE mark now. Not even the old incorrect kerning "China Export" variant

Final follow up on this L9 error: it was not (really) the flow sensor or the kinked up flexis, but a load of limescale build up in the ASHP heat exchanger. Full write up in new thread:

OK early signs are (very) promising! the "bucket test" delivered 24 l/min, and I've refilled the system (adding a bottle of F1) and it's now doing a DHW cycle at 16 l/min. Still not quite where I'd like it to be, but far better than the 10 l/min or less it's been hobbling by on for the last 24 months .... Interesting... I'd always thought using such occasions as an opportunity to flush out the crap and refill with clean. But yes, or course ... if scale is the problem refilling it is introducing a whole load more calcium carbonate ontop of what's already built up. I guess running the drain off through a fine filter would help sift off the worst of the mess coming out. Challenge is where to put the 100l of water when temporarily draining down the system ... although in most cases I can isolate the volumizing tank and leave the UFH loops untouched, so probably a quite doable. I'll read the heatgeek page and ponder how to attack this next time I do some service on it. Good news is I do now have this nifty wee pump so cycling water through it is much simpler now. 15 quid well spent https://www.ebay.co.uk/itm/187759301764

Thanks @JohnMo that's super useful info Today's project will be just to descale (the second batch of DS3 is currently resting in the PHE), flush, flow rate test, lob in a bottle of Fernox F1 and set the system going for a few days to validate I'll definitely look into the VDI2035 for a follow up project once the basics are back working