Local factory possibly causing premature failure of LED lights in nearby house ?

As the subject says. Is it possible that spikes of some sort from a local factory can cause premature failure of sensitive equipment.  

Over the last few years i have regularly fitted a certain circular LED mains under cabinet light . Out of probably fitting well over a hundred, i know not a massive amount, i have never had a call back to replace any.  I fitted 6 in a property local to me last year and have so far replaced 2 and just had a call back as another has failed.  So dodgy batch possibly or is it possible that a local factory could be spiking the mains somehow ( literally next door but one to the house) ? I have replaced a couple of external LED lights at the same house.  The factory is a wool mill or something similar making tennis balls and snooker table clothes amongst other things and runs all day and all night.

Any thoughts please, is that feasible ? I don't currently know if the latest failure is an original or a replacement .

Gary

  • In principle it is certainly possible, but I'd like to think it not all that likely, and I think it would be very hard to prove. I presume the factory is 3 phase, and could be pulling the supply to the house either up or down as loads come on or off, and it could be that there are mains born spikes or high levels of RF energy from plastic 'welding' or wood drying equipment

    Equally there could be things in the house itself, or perhaps another house on the same phase that  are causing problems.

    I presume you have eliminated he simple and confirmed that the mains RMS voltage is in spec. but fast spikes or RF riding on the back of the 50Hz will not be seen by most meters and in any case without 24 hour monitoring a spot measurement may not be at a time some offending machine is actually in the problematic state.

    A scope may show something but the same risk of missing the effect on the day applies,

    Another  approach would be to have a mains filter to the lighting circuit so that anything coming down the wire from anywhere is attenuated before it reaches what seem to be rather delicate fittings.

    Note that a surge arrestor that only cuts off anything above about 500V may well not do a lot to help in case where there are a lot of lower level spikes, but they may still affect operation. 

    Before doing very much - it may be worth asking do the neighbours have a problem

    Mike

  • Thanks Mike.   As you say, hard to prove, just got me thinking thats all.  i did think about getting them to talk to the neighbours. i'll call in next week and see if it was an original fitting first as well.  

    Gary

  • I know they aren't perfect,but if not already installed would it not be a good idea to install a SPD in or near the CU and for good measure a type 3 SPD where the lights are connected. Both don't cost that much and can at least reduce any spikes coming from the factory or other sources inside or outside the house.

    Most silicon devices designed to be connected to mains voltage are designed to withstand 450v or 500v, so a spd that custs off spikes above 400v will  certainly help if there are spikes.

  • Quite a few things to consider on this one.

    Are LED lights the same quality all the time.  I would say not and a faulty batch is more possible now than lets say 10 years ago.  Eg Osram/Sylvania are now Chinees brands and not made to the same high quality as before.  However most people remember them as very good brands and continue to purchase them.  The internet is full annedotal evidence of LEDs in homes and outside homes failling.  Some fail in strange ways like a PIR flood light that stays on or even start to flash.

    Might be worth getting the customer to speak to the local neigbours to see if they are experiencing some thing similar?

    Is there SPD type 2 and 3 at the property?  Could it be fitted?

    Hire or purchase a PEL (power energy logger) from Chanvin Arnoux and leave a the property for a week or 2.  CEF and others hire them out quite cheep.

    Has an EICR been done at the property recently?

    How old is the current install of electrics at the house?

  • Be aware that some makers (Fluke for example)  make a distinction between energy monitors and power quality monitors - the latter record more events, like fast spikes, while the energy monitor is more concerned with energy use, power factor, harmonics  and brownouts. worth seeing what is and is not logged by any particualr model.
    For spikes and RFI, you need an instrument with a high sampling banwidth, ideally ten MHz or more (the first hundred harmonics of 50Hz only get you to 5kHz - a sharp spike can easily be 1000 times faster. )
    The good ones save it all to a memory card so you can look at the waveform, or at least inspect the statistical data created from it later.
    Mike.

  • I'm sure you used to be able to get "surge clocks" - I've a feeling RS used to do one - basically a big 13A plug with a a ring of LEDs - every time it saw a surge it advanced the lit LED one place. Might have been useful for monitoring a site to see if it had an on-going risk. I can't seem to find any now though ...maybe they weren't reliable?

       - Andy.

  • These sort of surge montitors  are really equivalent to a spd in series with a pulse counter, and record how often the SPD has conducted more than say 1mA. For what they are they are quite dear, due to the interfacing and all the software.
    I suspect the plug in ones were simpler, and really where MOV and a counter ;-0
    mike

  • I have seen similar issues caused by adjacent premises.  A client of mine was having endless grief with LED lights, as well as expensive computer equipment, burning out (sometimes fairly dramatically).  He was noticing his lights always tended to flicker in the evenings, whenever his neighbour across the road would be using a welder.  I fitted an SPD almost a year ago for him and he has had no problems since.

    Another issue I once had was after fitting a large number of new LED lights to a building, only to find that they kept flickering when I turned them on (last thing on a Friday evening!).  After some investigation, it turned out that the voltage coming into the building was considerably lower than it should have been.  There was a restaurant next door, supplied by the same transformer, and it was in full operation at the time (with a lot of cooking appliances etc. running).  When I measured the voltage again the next morning, with the restaurant closed, it was well within specification.