Andrew Pickett

How a page here gets written

Every fault code page on this site is written from the manufacturer's own documentation for that appliance — the installation and servicing instructions, the user guide, a published service bulletin — and it says what that document says. Not what a repair forum says the code means, and not what it looks like it ought to mean.

The second half of the method matters more than the first. Where a manufacturer has published no meaning for a code, the page says they have published nothing, rather than quietly filling the gap. There is a value in this site's own data for recording exactly that: the manufacturer's documentation was reached, the code is not in it, and the page tells you so in as many words. Seven pages carry it. It is the least commercially useful thing I can put in front of somebody and it is often the only honest answer available.

What checking actually finds

The reason to work this way is that it keeps finding things, and what it finds is not what you would expect. When a page here has been wrong, it has been wrong while reading perfectly well.

All eleven boiler brands on this site have now been checked against manufacturer documents. Thirteen manuals went into the last three of them — Vokera, Ariston and Ferroli — whose literature I had recorded as unreachable across three earlier passes and which turned out to have been reachable the whole time. Four passes over that corpus have not once found a page that was merely thin. Every one has found pages that were wrong.

Viessmann publish E3 as a fault in the safety chain. This site had described it as a flow temperature sensor fault and called it one of the less alarming codes on the list. Alpha publish E20 as false flame detection — the appliance seeing combustion when there should be none. This site had it as weak flame sensing, rated as something to get looked at, with a dirty electrode and a cheap fix among the causes. Both were the exact opposite of what the manufacturer publishes.

Wrong procedures are worse than wrong meanings, because they travel further. Three reset instructions have been corrected in that corpus, and not one of them had come from a manufacturer's document. Two of them told you to hold the reset button for five seconds. On an Alpha, holding it for eight with no demand present starts chimney sweep mode, which fires the burner at a fixed output for commissioning; Alpha's own instruction is a plain press. On a Worcester Greenstar i, holding the two keys for about eight opens a menu where one further press returns the appliance to factory defaults, taking the installer's commissioning settings with it; Worcester's own instruction is a short press until the word Reset appears. All three had been written by somebody reasoning about what a reset probably looks like, and all three landed within a few seconds of an engineer mode. That is exactly the failure this method exists to catch.

Those are corrected, and the corrections are recorded rather than tidied away.

What I will not fill in

The same method decides what is not here, and those decisions are on the pages instead of behind them.

Beko publish no fault code table for their tumble dryers. That is not an inference from a gap: forty-two of their own manuals, Beko- and Blomberg-badged, contain no occurrence of "error code", "fault code" or "error message" anywhere in them. So the Beko dryer pages say so, and explain the one alphanumeric prompt those machines really do show, rather than reprinting a list from somewhere else. Of 311 Candy and Hoover dryer manuals, 233 document no codes at all.

Where two of a manufacturer's own documents disagree, the page carries both and tells you how to work out which one applies to you. Whirlpool describe F4E9 as a blocked sound channel and a sensor cable in one document and a microwave inverter in the other, so that page carries both and uses one symptom — does it still heat? — to tell you which one you are looking at, because being quoted for an inverter on an appliance that heats perfectly well is the specific harm there. Bristan give E1 as a failed thermistor needing a phone call on one shower and a filter you can clean yourself on another. Both documents are current, so the page says to check which model you have.

And a citation is a claim rather than evidence. Six source notes in the boiler corpus credited a manufacturer for numbers that manufacturer does not publish, several of them on pages that were otherwise right. A false citation is worse than no citation, because it stops the next person looking.

What I am not

I am not a Gas Safe registered engineer. I am not a registered electrician, and I am not F-Gas certified. I hold no trade qualification in any of the fourteen appliance categories on this site, and I would rather this page said so plainly than leave you to assume otherwise from a confident tone.

No engineer has reviewed any page on this site — not for gas, domestic appliances, refrigerant work or electrical installation — and the team page says so. No page here carries a reviewer byline, nothing says it has been checked by anyone, and no piece of structured data on this site claims a registration.

What sets the severity on a page is the manufacturer's own instruction. Where their documentation says isolate the appliance and call somebody, the page says that and publishes no steps to try, whatever that costs in traffic. Where they publish a remedy an owner can carry out, the page publishes that and nothing past it.

That is a defensible way to work. It is not the same thing as a competent person having read the page, and nothing here describes it as though it were. How we verify sets out the whole of it, including the parts I cannot do.

If something here is wrong

Every code page has a link at the foot for telling me so. It writes to a queue in the site's own admin, with a target of five working days, and when a report is right the page changes.

If you work on these appliances daily and I have a procedure wrong, that outranks what is on the page — the editorial policy commits to it in as many words, and it is the reason the link is there. Where you and the manufacturer's own document disagree, the page will carry both; how we verify publishes the order it resolves them in, and the manufacturer's documentation is first on that list.