Rules and Guidelines
Want to participate in the discussions? Please read our Community Rules and Guidelines. Need some help? Visit Support and Answers.

  • State Not Answered
  • Locked Locked
  • Replies 40 replies
  • Subscribers 32 subscribers
  • Views 3499 views
  • Users 0 members are here
Options
You may also be interested in
This discussion is locked.
You cannot post a reply to this discussion. If you have a question start a new discussion

When is a spark an arc? OR - When is an arc a spark?

whjohnson
Just watched some chap on the E5 youtube channel visiting Eaton in Austria. Eaton AFDDs have been something of a subject of ridicule in youtubeland, with various respected electrical content providers demonstrating through various real-world means that they don't function. So, off this chap goes to Eaton's HQ in Vienna where they provide him with a aluminium case full of test kit, complete with the Eaton logo and fitted out with various Eaton devices inside.

One of the devices is the Eaton AFFDD which has famously failed to operate on numerous youtube video presentations.

Of course, it trips when tested with their own test kit. After all, no point in trying to sell something which isn't really needed unless you can demonstrate that it actually works, so Eaton helpfully provides the 'right' arc signature so that the device can trip on command in front of all those cynical doubters.


Apparently, all those heath robinson youtubers have been getting it wrong because they have unhelpfully been simulating real world arcing events which these devices won't actually pick up. You see, according to the 'experts' you need an arc instead of a spark to trip the device! What the hell is the difference?


Oh how I laughed! Is this how far they'll go to flog you some old tat you don't really need?

Just how many different arcs and sparks are there out there? Has anyone told David Attenborough of all these new species to explore?


Feel free to jump in!
  • 0
    In normal high voltage parlance a spark is pre-avalanche, and is based on simple breakdown, say ~1mm/kV for sub 10kV sort of voltages.

    Once current starts to flow however,  the gap rapidly fills with ionised material - that is atoms from the air, or metal torn from the electrodes, but with more or less electrons than would make them neutral, and they move towards one electrode or the other. At high current densities, these ions collide with more atoms, and create more ions, and the amount of plasma that can be created by this avalance  reaction is only really limited by the power available to tear electrons off atoms, and this means that if enough current is flowing, the spark can be drawn into an arc that is many times longer, and is not directly related to the initial voltage. A classic must be a welder, to strike the electrodes must practically touch for we only have 50V or so available, but once struck the arc length, and also the weld depth, is set by the available current.

    A similar thing occurs with spark gaps, or even the small neon tubes, in low current glow, the voltage is almost fixed, but as current rises, the arc forms, and the voltage falls dramatically . At 10A or more, the neon that is 90V and orange at a few mA, will only drop only about 5V and glows blue white. Such tests must be limited to a very short duration - typically  hundreds of microseconds at most, total energy is limited with capacitor discharge circuits., as the arc is hot, and the total energy will cause the envelope to shatter.


    For AFDs however, I am not sure these definitions stand up.
  • 0
    I don't think "arc" or "spark" are defined in Part 2 definitions of BS7671 (given the time I am cheating and looking at an online pdf copy of the DPC for convenience rather than my official copy which isn't next to me right now).


    It is possible that IEC 62606 (standard for AFDD) has a specific definition for these terms*, if not then for an IEC based standard the definitions in the International Electrotechnical Vocabulary usually apply.

    This can be searched free online here: http://www.electropedia.org/


    Searching for "arc" includes the following result:

    arc
    arcing




    luminous discharge of electricity across an insulating medium, usually accompanied by the partial volatilization of the electrodes

    Note 1 to entry: A complete sinusoidal current half-cycle is not considered to be an arcing half-cycle.






     

    [SOURCE: IEC 62606:2013, 3.1]





    and searching for "spark" includes the following:

     




     



     

    (electric) spark




    small luminous electric arc of short duration


    so from that it would seem that a spark is a short arc (which is effectively what Mike said earlier)


    * it turns out that the primary definition for "arc" is actually within IEC 62606 as this is referenced in the IEV entry.


  • 0

    You see, according to the 'experts' you need an arc instead of a spark to trip the device! What the hell is the difference?



    So is a spark a 'click' and an arc a 'crackle'?


      - Andy.
  • 0
    And is Arc the plural of Spark?

    Alasdair
  • 0
    An arc is what a spark grows into if you feed it, and once established an arc can outgrow its containment, as a large volume of sizzling plasma, and is more flame like

    Or a spark is all you have if the energy is limited.

    https://upload.wikimedia.org/wikipedia/commons/thumb/c/c3/Spark-plug01.jpeg/200px-Spark-plug01.jpeg

    https://upload.wikimedia.org/wikipedia/commons/thumb/f/f8/Lichtbogen_3000_Volt.jpg/220px-Lichtbogen_3000_Volt.jpg


    In reality there is a fairly blurred intermediate region.
  • 0
    Mike,

    While I have seen some (single) sparks with more energy than some sustained arcs I have seen, I think your explanation is a pretty good summary.

    Alasdair
  • Former Community Member
    0 Former Community Member

    whjohnson:

    Just watched some chap on the E5 youtube channel visiting Eaton in Austria. Eaton AFDDs have been something of a subject of ridicule in youtubeland, with various respected electrical content providers demonstrating through various real-world means that they don't function. So, off this chap goes to Eaton's HQ in Vienna where they provide him with a aluminium case full of test kit, complete with the Eaton logo and fitted out with various Eaton devices inside.

    One of the devices is the Eaton AFFDD which has famously failed to operate on numerous youtube video presentations.

    Of course, it trips when tested with their own test kit. After all, no point in trying to sell something which isn't really needed unless you can demonstrate that it actually works, so Eaton helpfully provides the 'right' arc signature so that the device can trip on command in front of all those cynical doubters.


    Apparently, all those heath robinson youtubers have been getting it wrong because they have unhelpfully been simulating real world arcing events which these devices won't actually pick up. You see, according to the 'experts' you need an arc instead of a spark to trip the device! What the hell is the difference?


    Oh how I laughed! Is this how far they'll go to flog you some old tat you don't really need?

    Just how many different arcs and sparks are there out there? Has anyone told David Attenborough of all these new species to explore?


    Feel free to jump in!






    The "chap" is actually a Chartered Engineer, and one of the youngest IET Fellows of all time, if it makes any difference, and the video of the test rig was pretty much all we were allowed to film.  I know because I was filming.

    When we went into their R&D laboratory, understandably to be honest, we had to leave our phones & cameras in the board room.

    What you have to realise is that they have to build the devices to the EN standard, 62606, and that defines the requirements that they have to detect, and that is pretty covered by the video I think.

    One of the things that they look at in the analysis of the arc is repeatability, and the way that the other youtube video's seemed to have addressed the issue does't give the repeatability of the arc over several cycles of the supply, this combined with the characteristics, which both Eaton & Siemens agree over, are critical to the detection algorithms that are used, even though they may both do the specific detection in differing ways.

    We have been to see both companies and reviewed a lot of their information and products.

    There is a limit, obviously, to what they would show us as it is their intellectual property.

    The random generation of sparks by touching a few bits of wire together when wiggling them around doesn't give you the arc stability that is required for detection.

    Eaton had the bit of wire to be wiggled together like John Ward (JW on YouTube) originally had, and demonstrated the arc pattern on their test rig, which unfortunately didn't fulfil the required criteria.

     

  • 0
    That is interesting to know, thought it simply transfers the question to be asking if the EN standard 'arc' fault waveform actually represents the sort of current waveform  you would get with a real fault of the kind you should disconnect.

    This is very different to testing RCDs where we are all agreed that the test rig and the faults are similar, here I suspect there are real loads that look like arc faults, and arc faults that have signatures  nothing like the standard test waveform.

    With modern electronics it is possible to interrupt and reconnect in almost arbitrary speed and precision,  so making an AFD  tester that spits out a very complex set of test waveforms is a technically trivial matter - the hard work is agreeing what the waveforms  should be.

    I'm not sold on the idea myself, I thing we are some way from a useful item.
  • 0
    Whilst I appreciate that all devices have to have some standard they are tested too surely in the real world  the arcs won't be sustained or have the consistency that these devices seem to need I realise you don't want these  things  tripping when you get  an arc across switch contacts  but surethe best time for them to operate is when you get a series of arcs from a loose lead or similar. I've used the term arc to mean any event where juice flashes across a gap so technically could be an arc or spark
  • Former Community Member
    0 Former Community Member

    mapj1:

    That is interesting to know, thought it simply transfers the question to be asking if the EN standard 'arc' fault waveform actually represents the sort of current waveform  you would get with a real fault of the kind you should disconnect.

    This is very different to testing RCDs where we are all agreed that the test rig and the faults are similar, here I suspect there are real loads that look like arc faults, and arc faults that have signatures  nothing like the standard test waveform.

    With modern electronics it is possible to interrupt and reconnect in almost arbitrary speed and precision,  so making an AFD  tester that spits out a very complex set of test waveforms is a technically trivial matter - the hard work is agreeing what the waveforms  should be.

    I'm not sold on the idea myself, I thing we are some way from a useful item.




    It's not so much a standard arc waveform but a description of the criteria that need to be met, & a standard test "wire" to simulate the fault.


    Along side their development rig Eaton had racks of different products which they had been using to develop the firmware, as in to recognise their arc signature as friendly arc's Siemens have been doing the same.

    I think the issue with developing a standard test device is that the specifics of the detection mechanism, i.e. how the hardware and firmware respond the same to the same stimuli, or just give the same end result.


    As far as RCD's go, with all the new non linear loads that have developed and the RCD's that have been developed to identify those leakages and react to them, are our "standard" RCD tests adequate any longer?...

    As far as "some way from a useful item" goes, I'm not going to argue, especially when you think that the minimum current threshold for detection is around 2.5A, but, being worked on by the manufacturers to reduce it I believe.

Join us to get the best from IET EngX.

Joining EngX lets you personalise your experience so you stay up to date on the topics that interest you, plus you’ll be able to make connections who are looking to collaborate, exchange ideas and more.

Join us Not now

Community Rules & Guidelines