Inside a recalled prominent brand socket
Added 2025-01-02 02:06:33 +0000 UTCI opened this socket expecting to see low separation between poles in a double pole switch, but instead it is a single pole switch with good contact clearance.
The internal design didn't really seem to have any obvious flaws at first glance.
I'll let you take a look and judge for yourself.
Comments
There's a messaging system on Patreon. Or you could ask here.
Big Clive
2025-01-08 06:57:51 +0000 UTCHey is there a way to email you? I wanted to pitch an idea about stage lighting dimmer rack solid stage controls and solid state relays
James Hamilton
2025-01-08 06:13:03 +0000 UTCI would never design it with the two switches so close together ! I would stay away from making cheap products MK makes good products too.
Richard Lackie
2025-01-08 03:19:41 +0000 UTCThank you for replying! Unfortunately, it confirms some of my fears, but also alleviates some. You, at least have the UK safety people looking out. I'm in the USA and as long as the design, with correct components, pass the testing, the UL sticker goes on. But like all things, caveot emptor!
Mike Bird
2025-01-04 19:20:56 +0000 UTCThe government report covers product shipped from 2018 to 2024 (see Simon Allen's post).
Mike Page
2025-01-04 18:36:26 +0000 UTCI design electrical test equipment manufactured in China and sold in the UK. As an engineer I'm one step removed from commercial and production operations and it's really hard to guarantee what I've put on the Bill Of Materials (BOM) is what gets used. There's a lot of secrecy around component sourcing. There's a lot of pressure on BOM cost and delivery. Nobody wants money tied up in stock they can't sell; and most shipments are by sea (~6 weeks). Foreign imports aren't banned in China but are punitively taxed. It's not uncommon for purchasers to "nip round the corner" to grab reels of whatever they need to fulfil an order. Counterfeit semiconductors must be anticipated. The factory are generally OK about me specifying "critical components" as long as it's not everything on the BOM. Which means everything else is up for grabs. And then you have the approvals "process" which as far as I can gather consists of handing over a moderate sum in exchange for a report which says "PASS" at the bottom. I like to have safety behaviour 100% tested. I don't always get my way; but fortunately the UK safety culture takes up a lot of slack as my designs are used by professionals rather than Joe Bloggs. It sounds cynical but in the end there is no such thing as perfection. It's a quite different way of working. As I'm not the one selling the products, I don't bear legal responsibility. After shipping something close to 50,000 products nothing's gone wrong enough to end up in court. There are always chancers of course: people who screw up and try to shift the blame. They usually get found out. If they were smart they wouldn't cock up in the first place.
Mike Page
2025-01-04 18:27:15 +0000 UTCI think the biggest problem with products from that area is the lack of quality control. Some companies have figured that out and added real QC and that's how we have reliable power supplies in our computers and some really good machining. But it seems as if other companies just buy from the cheapest supplier this week and these companies think QC is a stamp that has to be applied before selling.
Mike Bird
2025-01-03 13:03:47 +0000 UTCWild guess, but these sorts of recalls aren't always a design issue. Sometimes a machine is misaligned or not calibrated correctly, and that results in problems that the as-designed part wouldn't have. If it's only on one assembly line, or only for some period of time, the units with the issue might make up only a tiny percentage of the number covered by the recall, so you don't see the problem with a random sample. e.g. the bus bar passing over the live could be pressed incorrectly so that it didn't have the design clearance from the other bar, and could flash over with an inductive load or something.
Charles
2025-01-03 03:21:09 +0000 UTCUK sockets are excessively complex. That's my only take away.
Dustin S Cochran
2025-01-02 23:19:32 +0000 UTCWith the switch pin near the earth bar, I wonder if flicking the switch too fast could bump the pin high enough for an arc to flash over to the earth bar, with a large inductive load like a large building site transformer.
Seán Byrne
2025-01-02 22:44:03 +0000 UTCIf this is the case it won't be the only Chinesium mains outlet (but not from a major brand) to be flammable.
Alice Chapman
2025-01-02 16:13:46 +0000 UTCAre they saying the plastic is not flame retarding and there is enough energy to ignite it (when hot for example after running full load on both sockets)?
Mike Page
2025-01-02 12:43:17 +0000 UTCIn my line of work I've always valued double-pole switching. You can nuisance trip an RCD by shorting neutral and earth on a single pole switch.
Mike Page
2025-01-02 12:36:45 +0000 UTCI'll have to test that.
Big Clive
2025-01-02 10:21:15 +0000 UTCIronically - no.
Big Clive
2025-01-02 10:20:55 +0000 UTCI found the recall notice. https://www.gov.uk/product-safety-alerts-reports-recalls/product-safety-report-mk-electric-mk-essentials-electrical-three-pin-sockets-with-switches-2411-0130 My guess is that if you moved the switch slowly under a high load it might flash but really I cannot see an issue with it. You made me laugh when you said that you went to the site where you could buy recalled products, you paused and said Ebay.
Simon Allen
2025-01-02 09:35:28 +0000 UTCThe problem is the plastic is flammable and doesn’t meet UK regulations. To say it’s not what you want in a socket would be an understatement.
Gadgetman
2025-01-02 09:03:09 +0000 UTCLogic Plus it ain't. That's for sure.
Gadgetman
2025-01-02 08:30:37 +0000 UTCThe 2747DPWHI/2747DPMBK are the double pole versions. By the looks of the recall notice they've just gone with a blanket recall of the whole range.
Alice Chapman
2025-01-02 08:29:51 +0000 UTCAre they a good price on eBay?
YouKnowHowYouAre
2025-01-02 07:38:15 +0000 UTCInteresting to see what the actual issue is since it looks like it's ok...
Doug Merrett
2025-01-02 03:34:12 +0000 UTC
