This discussion is locked.
You cannot post a reply to this discussion. If you have a question start a new discussion

The £1300 AFDD consumer unit

Should be good this one!
https://www.youtube.com/watch?v=YDGeyJnoqZQ
  • I wonder how many SPD have been checked since they were installed in consumer units?
  • Sparkingchip:

    I wonder how many SPD have been checked since they were installed in consumer units?


     


    I wonder how many ever will be?


    (The majority of folk don`t know what an RCD test button is so never test it)


  • An interesting video. I had not appreciated that AFDDs available at present use dedicated microprocessors. I have my doubts about incorporating complex technology into a consumer unit whose lifespan should be much, much longer than the typical consumer goods cycle that leads to new releases currently every 18-24 months for phones etc. Will these devices be reliable enough in practice? If they aren't, will they be directly replaceable on failure or will the latest 'update' force a wholesale replacement of all other devices (as David Savery found he had to do as the bus bar connections on the new AFDDs were not compatible with the other devices in his DB).


    Despite his ribald humour, David Savery has put out an interesting series of videos and at least is doing something to advance the discussion. I commend what he has done to move the debate forward.


    Anyways, I remain unconvinced that there is any need for AFDDs at all, but that's a view I've already made in a response to the DPC.


    -Stewart
  • If it's done properly, the microprocessor software in such devices shouldn't need updating - no internet (yet!), so no security issues!.


    The devices themselves should be good for more than 10 years of service life, with that being the life at which a certain percent are expected to fail with most lasting much longer.
  • In 10 years time the Regs will have changed that much that we will have to fit something else anyway. Probably a similar product, with questionable figures about how it improves everyone's safety, that the manufacturers have just invented and pressured into BS7671, as appears to be happening with these overpriced inconsistent products
  • Stewart Ross:

    An interesting video. I had not appreciated that AFDDs available at present use dedicated microprocessors. I have my doubts about incorporating complex technology into a consumer unit whose lifespan should be much, much longer than the typical consumer goods cycle that leads to new releases currently every 18-24 months for phones etc. Will these devices be reliable enough in practice?


    Good question. Computers are all over the place in cars now, but at 14 years the only problems that I have with my car are electrical. Nobody knows what the fault is in the sound system; the firmware in 3 out of 4 window modules needs to be updated; the clock no longer updates itself; and worst of all, I think that there may be a poor joint somewhere in the engine management unit. If the car doesn't pass it's MOT on Friday, it has scrap value only. ?


  • If it's done properly, the microprocessor software in such devices shouldn't need updating - no internet (yet!), so no security issues!.

    Software is the marketing departments dream - now you can just manufacture one module design and just before you print the case, flash it with the desired RCD sensitivity and delay, and maybe even have a programmable  MCB sensitivity.- after all the big ones already do.


    Shortly after that happens a shadowy market emerges of after market modders who will chip it for you to get 40A out of a 32A device, and various attacks to disable or enable features from the original spec - you only have to look at engine management software and the hacks to de-restrict small motorbikes to see this sort of thing. I'd not look forward too it with that great an enthusiasm.  Certainly I know companies that take the tops off chips and bypass the code protection for a living, and I suspect it can be done outside military circles as well - once the algorithm is liberatated, writing a similar but not identical one is not so hard. as this slightly out of date  article describes

    mike
  • The other thing of course, is that with the mass growth in consumer electronics security market, the product life cycle is even shorter and the initial products, bug-ridden and full of security 'holes' never get patched, thus leaving them insecure. Ring doorbell cams, CCTV products, amazon echos smart TVs with always-on  cameras and microphonesand the like,  All are left behind because the manufacturers can't be bother to release retro-security patch upgrades because the 'newest latest and greatest are about to hit the marketplace.

    No wonder I won't entertain any such devices in my own home.
  • mapj1:

    Software is the marketing departments dream - now you can just manufacture one module design and just before you print the case, flash it with the desired RCD sensitivity and delay, and maybe even have a programmable  MCB sensitivity.- after all the big ones already do.


    Shortly after that happens a shadowy market emerges of after market modders who will chip it for you to get 40A out of a 32A device, and various attacks to disable or enable features from the original spec - you only have to look at engine management software and the hacks to de-restrict small motorbikes to see this sort of thing. I'd not look forward too it with that great an enthusiasm.  Certainly I know companies that take the tops off chips and bypass the code protection for a living, and I suspect it can be done outside military circles as well - once the algorithm is liberatated, writing a similar but not identical one is not so hard. as this slightly out of date  article describes

    That can be done, but it's also possible to secure the firmware so that it cannot be tampered with - common where tampering is shown as an identified risk related to functional safety. However, that does often require a more expensive processor (but only a few £s)...

    Even engine management system are starting to do this, especially now that OTA (other the air) updates are becoming more common and vehicles are starting to have paid for after-market upgrades/options.
  • Stewart Ross:

    An interesting video. I had not appreciated that AFDDs available at present use dedicated microprocessors. I have my doubts about incorporating complex technology into a consumer unit whose lifespan should be much, much longer than the typical consumer goods cycle that leads to new releases currently every 18-24 months for phones etc. Will these devices be reliable enough in practice? If they aren't, will they be directly replaceable on failure or will the latest 'update' force a wholesale replacement of all other devices (as David Savery found he had to do as the bus bar connections on the new AFDDs were not compatible with the other devices in his DB).


    Despite his ribald humour, David Savery has put out an interesting series of videos and at least is doing something to advance the discussion. I commend what he has done to move the debate forward.


    Anyways, I remain unconvinced that there is any need for AFDDs at all, but that's a view I've already made in a response to the DPC.


    -Stewart


    I agree with the posting above. I was amazed at the complicated internal construction of an AFDD and wondered what the electrical strength/resistance to surges was. I wasn't successful in tracking down the surge withstand strength information on the internet. It would be ironic if the introduction of AFDDs required the addition of SPDs! 

    Regards