Connect with us

Surge Protection

Discussion in 'General Electronics Discussion' started by shiekh, Oct 11, 2010.

  1. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    I'm a little surprised that people don't seem to be trying their hand at surge protection, though it might be interesting to design filters that would send higher frequencies back down the line and not let them through (for mains protection).

    A little more difficult for signal protection, when higher frequencies need to get through.
     
  2. Militoy

    Militoy

    180
    0
    Aug 24, 2010
    I'm fairly sure many Engineers on the site are incorporating some type of surge protection in their designs. It is surely a consideration in every product I work on. The idea of a filter isn't to 'send higher frequencies back down the line" - but to mismatch impedances of undesirable electrical energy. Presenting a low impedance in parallel with, and a high impedance in series with undesirable signals attenuates their effects on the desirable part of the signal or power.
     
  3. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,192
    2,694
    Jan 21, 2010
    In fact shiekh, I think you'll find that whenever someone talks about electronics in noisy environments (car electronics for example) it is something that comes up.
     
  4. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    Attenuates

    I think that is the point, one really does not want to 'attenuate' large amounts of energy, one wants to reflect it away...
    I perhaps should have said I have large energy surges in mind here... (my bad)
     
  5. Militoy

    Militoy

    180
    0
    Aug 24, 2010
    Perhaps you would like to elaborate on how transient energy might be "reflected away", without the need to attenuate or absorb it. In military systems - we attenuate and absorb transients from load shed and lightning strike, etc., using a combination of filtering and TVS protection devices. The lowest level of protection is typically a varistor (MOV) - somewhat problematic, as their degradation over time with transient events makes them a maintenance point. At higher levels of event frequency, we use a combination of TVS protectors, gas tube surge arrestors, compliance inductance, etc. For MV direct lightning strike protection - we start considering more serious protection.
     
  6. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    Impedance mismatch seems like a good start.
     
  7. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,192
    2,694
    Jan 21, 2010
    I am seated comfortably and have popcorn.

    Please start you lecture when you are ready...
     
  8. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    Caramel Popcorn?

    It is just the quarter wave filter, certain frequencies are let through, the others are reflected back, since the impedance is frequency dependent; so I am not suggesting anything new. I just don't want to have to dissipate the energy; that way maybe one can deal with more...
     
    Last edited: Oct 12, 2010
  9. davenn

    davenn Moderator

    13,328
    1,769
    Sep 5, 2009

    well thats the best bit of rubbish I have ever heard..... ya better go and read up some transmission line data mate

    Impedance of a line , be it waveguide, coax, twin cable, or openwire is totally freq INDEPENDANT. that is its impedance doesnt change with freq if it did we would have to produce coax cable for every different freq that was in use

    If you do some reading, you will discover that impedance is a function of the sizes, spacing of the conductors and dielectrics used in a given cable (transmission line)

    Dave
    VK2TDN
     
    Last edited: Oct 12, 2010
  10. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    Exactly... and we are talking a quarter wavelength insert, not a uniform cable.
    http://physics.usask.ca/~hirose/ep225/animation/emreflection/anim-emreflection.htm
    so the matching only happens for certain wavelengths; impedance mismatching for other wavelengths, that is why I made the original suggestion (impedance mismatching).

    i.e. wavelength (frequency) matters, or we wouldn't make the insert a quarter wavelength.

    But my real point is to follow the philosophy of not dissipating the energy, the above was just one such example.
     
    Last edited: Oct 12, 2010
  11. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,192
    2,694
    Jan 21, 2010
    No, check out my steely expression, it's irony flavoured :)

    Perhaps you can start by explaining how to implement a quarter wave filter for equipment designed to run on voltages from say 90 to 270 volts 50/60Hz. Let's assume (for simplicity) the device has a power factor of 1 and consumes a constant 25W.
     
  12. davenn

    davenn Moderator

    13,328
    1,769
    Sep 5, 2009
    Now back to the actual topic.... Surge Protection

    not really sure why you are relating freq to hi energy surges would love to hear your ummm explanation ;)

    The most common way of dissapating hi energy pulses (>200V) on any sort of transmission line will vary with the expected level of the surges.....
    for example on RF feed systems to antennas spark gap and gas discharge devices connected between the cables and ground are the most common as they dont affect the impedance of the cable.

    on, say household and other general equip across 110/240VAC mains supply a mix of gas discharge, MOV's or PTC's can be found.

    a lot of incorrect assumptions.... much equip (specially test gear) already has mains filters. The last thing you want to do is reflect it back down the line it doesnt really achieve much... well nothing good anyway

    there are things called Hi pass, Lo pass and band pass filters for filtering out unwanted SIGNALS.... NOT Surges 2 very different beasts


    Dave
     
  13. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    Good point, and again my bad; I also had in mind protecting data cables, where the quarter wave might work.

    For much longer wavelengths one might want to use an inductive approach; what are those filters called that use what looks like a transformer but turned sideways, as is often found at the input to a PWM power supply; they also don't dissipate.
    I need to catch some sleep now, but I can continue to entertain with my stupidity tomorrow if requested...
     
  14. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    I agree, two very different beasts; but perhaps one can learn from one beast and use it to design a circuit that will be very selective about say 60Hz and throw back all else.

    Reflecting it back down the line might protect the equipment; and the surge will be spread out for a mains feed (dispersion). True the other users on the line might not be too happy with me, but we all just got hit... at the end of the day the hit needs to be dissipated, but that might best be done on the feed cable.
    I'm just playing with ideas, and have already been beaten black and blue for my ignorance... I'm just playing with ideas...
     
    Last edited: Oct 12, 2010
  15. Militoy

    Militoy

    180
    0
    Aug 24, 2010
    Those filters that "look like a transformer but turned sideways" are called "baluns" or "common-mode inductors". They block common-mode currents by dissipating them as core losses. If they had infinitely high Q (no dissipation), the CM current would go on ringing away on the "dirty" side of the circuit forever.
     
  16. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    How can one ring on the dirty side forever? Would one not just be sending a wave out... i.e. the pulse is reflected; and thanks for the name.

    I'm new here; I thought noobs like me were allowed to ask questions, especially in this chat area.

    Now how to build one with low loss so it rings the wave back... without heating up too much. I'm worried about what might happen if the core saturates... then the inductor shorts, and the surge flies straight through... not good.
    I really got to sleep now, and thanks for all the help.
     
    Last edited: Oct 12, 2010
  17. Militoy

    Militoy

    180
    0
    Aug 24, 2010
    Good night.

    Tomorrow - something to think on - conservation of energy. If nothing dissipates the energy of the wave set up by the transient, it will "bounce" back and forth forever on the cable, or in the circuit. You have repeatedly referred to the "reflection" of the transient. Problem is, if it can be "reflected" at one end of the circuit - it can be reflected back again, unless it is dissipated. When a transient pulse "rings down" from its peak level to zero volts - that's because the energy was absorbed - dissipated - by some element in the system.
     
  18. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,192
    2,694
    Jan 21, 2010
    Sure, ask questions as much as you like. I was actually going to note earlier that you would be better advised posting your thoughts as questions rather than statements.

    You're not really to know, but you've managed to engage 2 of our very knowledgeable members here. I'd listen to them if I were you -- I always do.

    Go and have a read of this thread and note that you've strayed into Militoy's strong suit. I don't have a thread handy, but rest assured, your mention of quarter waves places your topic well into davenn's as well. Me, *shrug*, I'm just an intellectual pigmy in comparison -- but when you're in to your 4th decade of experience you can just sorta tell when things don't sound right.

    Oh, and this isn't rough... You shoulda been here 6 months ago :D

    Seriously though -- starting with a statement rather than a question sets you up to be shot down in flames.
     
  19. davenn

    davenn Moderator

    13,328
    1,769
    Sep 5, 2009
    Thats ok Steve, you blow me out of the water when it comes to digital electronics
    thats definately my weakest area in electronics :rolleyes:

    Its just great that there's a bunch of good guys on here, each with their specialties and a bit of overlap. I have learnt a hell of a lot in areas that Im not too familiar with

    cheers

    Dave
     
  20. shiekh

    shiekh

    77
    1
    Oct 11, 2010
    Many good points here; I'll be sure to put a question mark in next time.

    I agree that at some point the energy is going to dissipate, my thought/question was would it not be better to do this outside of the equipment?

    A further thought; I assume (wrongly perhaps) that a lot of pulses are picked up from the EM pulse, as opposed to a direct hit; if so, then the same loop might just as well radiate the energy away. I guess I'm not being clear when I have an exposed mains line in mind, and when I have a data transmission line in mind.

    Again, I am just messing with ideas, I should not have made things look like statements.

    It is just my luck to have engaged the strongest members, and I don't mean this sarcastically.
     
    Last edited: Oct 12, 2010
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-