Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Why EMP pulse destroy electronics?

  1. Jan 1, 2012 #1
    Just like the tittle. I never understand why EMP destroy electronics. People talked about a nuclear explosion will destroy electronic that is on, not off. Why?

    I just never understand this.
     
    Last edited: Jan 1, 2012
  2. jcsd
  3. Jan 1, 2012 #2

    es1

    User Avatar

    I guess we have to agree on what exactly an EMP pulse is before answering, and once there is a clear definition I'll bet you'll have your answer as you know EM.

    But if it is a varying magnetic and electric field then there should exist at least one field, given any coupling constant, that can generate a large enough current to damage an IC.

    Say for example on an internal high impedance net without a high breakdown voltage. It should be not impossible to punch through a FET gate for example rendering the FET non functional.

    I've seen plenty of designs where the reset net was high impedance with rats all over the place. This would be (and in one particular case I know of was) a good victim for EM. Blow out one FET such that it now has a strong pull down and the whole circuit can't work. (In the case I know of the IC was not damaged but the reset did fire at inopportune times.)

    Is it practical to generate such a field? Guess it depends on how weak the net is and the coupling constant...

    I don't think you need hollywood style sparks to render a circuit non-functional.

    I guess you could start looking here:
    http://en.wikipedia.org/wiki/Electromagnetic_pulse
     
  4. Jan 1, 2012 #3

    jedishrfu

    Staff: Mentor

    A magnetic field has a certian eneray stored in the field. If you were construct a magnetic field via coiled wire and electric current and then abruptly disconnected it, you'd see a spark jump across the switch. The collapsing magnetic field generates a current in the wires which jumps the gap causing the spark.

    So a changing magnetic field would induce a current in electrical devices that might exceed their capacity to channel and thus burn out the circuit. Remember microcircuits can only handle so much current, the more current the higher the wire temperature and if its high enough the wire melts.
     
  5. Jan 1, 2012 #4
    I really don't know anything other than watching the movie!!!!! That anything that is powered on at the time will be destroyed and anything that is not is ok!!! Do you mean this is Holly Wood's myth?

    So a lot of the circuits that pass CE or UL test should survive because they do have transorbs and all to protect the circuits. I can see most circuits will render non function because of the EMP pulse, but a lot should be able to come back to live upon power down and power up again.

    Maybe the military pcb should demand transorb arrays even in the internal bus and have a central power reset in every piece of equipment.
     
  6. Jan 1, 2012 #5

    phinds

    User Avatar
    Gold Member
    2016 Award

    When an EMP takes out the power grid, it really won't matter if other devices COULD still work, because they won't have any power.
     
  7. Jan 1, 2012 #6
    Why? What is so different about the power grid?

    Still you can avoid the fighting jets, bombers and fighting vehicles from being toasted.
     
  8. Jan 1, 2012 #7

    nsaspook

    User Avatar
    Science Advisor

    EMP for Mil-Spec equipment is not a problem. Everything critical built since the 1970s is hardened.
    http://www.wbdg.org/ccb/FEDMIL/std188_125_1.pdf [Broken]
     
    Last edited by a moderator: May 5, 2017
  9. Jan 1, 2012 #8
    That's music to my ears!!! So the movies are just sensationalized the whole thing?
     
  10. Jan 1, 2012 #9

    nsaspook

    User Avatar
    Science Advisor

    No, I can actually spin bullets around the room to a target.
     
  11. Jan 1, 2012 #10
    Varying EM field produces current in a nearby conductor. The metals in the IC will act like small antenna and receive the signal from the field.
    Let's see what can possibly go wrong in an modern IC.
    The gate oxide dielectric may breakdown, The metallic interconnects may form discontinuity due to high current, even the field and isolation oxides may breakdown creating high parasitic. So I think all of these will occur together to bring down the IC.
     
  12. Jan 1, 2012 #11

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    completely out of my field. i claim ignorance on EMP.

    but physics-wise it seems to me it'd take a substantial antenna to pick up much energy.

    a 100 mile long power or telegraph line is one heck of antenna .

    a 10mm IC pin is not



    call me skeptical.
     
  13. Jan 1, 2012 #12
    I understand EMP is strong, but as circuit shrink, circuit loops are getting smaller and smaller. Total flux is flux density times the area inside the loop. I just don't believe there are enough energy in the flux to burn something in circuit loops inside an IC.

    The only loop that can cause enough energy are the interconnect between pc boards in the system. That's when transorbs come into play. By requirement, all I/O of the boards has to be protected by transorbs to pass CE test, that automatically protect the boards to a big extend.
     
  14. Jan 1, 2012 #13

    es1

    User Avatar

    Not much science. But is does claim to show a 1MV EMP destroying a flying toy helicopter. It just kinda turns off.
    Also he drives a car through and it kills the ignition.

    Note: you don't need to burn out an IC to render it inoperable. just inject a small current into a weak node. a consumer electronics project typically has many such nodes.

    http://www.youtube.com/watch?v=0odJKYTzXg8&feature=related
     
  15. Jan 1, 2012 #14
    http://www.thespacereview.com/article/1549/2

    The section "EMP effects upon IC-based devices" in this page may answer your question. Apparently EMP damages ICs which are already placed in PCB and turned on.
     
  16. Jan 2, 2012 #15

    es1

    User Avatar

  17. Jan 2, 2012 #16
    I used to work in EMC, and this was an interesting topic that I never had to address in design or certification, but it was very interesting nonetheless.

    Most of the movie stuff is bunk. For example, John Travolta inducing an EMP with an underground blast. That was almost as unbelievable as the people in the chamber getting far enough away in a few minutes...

    As I'm given to understand, the effect happens when a nuclear event unleashes gamma rays above the atmosphere. Gamma rays are not particles, but rather extremely high energy photons.

    When the gamma rays strike atoms in the atmosphere, they knock prodigious quantities of electrons free in a very short time. This somehow induces fields that may be as much as 25 kv / meter with a rise time down in the ns range.

    I've never heard of the effect being created for mid and lower lever bursts.

    As to how it gets to electronics, I suspect that it enters the packaging as most effects do - through the conductors, through an unshielded enclosure, or through poor attention to gasketing / seams in the enclosure. Subsequently, I suspect it destroys in a fashion similar to electrostatic discharge; it punches through gate oxide or physically damages the structure of fine structured devices (like RF amplifier transistors, low leakage devices, or ECL logic).

    Now, electronics is generally designed with extremely fine structured components. This is to get more components out of less material, and to enhance performance / complexity for a given piece of material. The net result, is that modern active devices are extremely susceptible to ESD.

    I suspect that very few bomb bursts would be required to totally destroy all civilian computing and communications throughout the developed world.
     
  18. Jan 2, 2012 #17

    es1

    User Avatar

    As circuits shrink, transistors also get much weaker and take much less J to damage. And you don't even need to fully damage it to make the internal digital gate not function properly.

    Note in the car example shown on youtube the power windows and dash lights still worked. They're probably run by just a simple switches connected to the battery. It's probably the small weak transistors in some uC in the car that died.

    And in the soviet experiments linked to by Kholdstare the old infrastructure survived.
     
  19. Jan 2, 2012 #18
    Agreed,

    The circuits are shrinking and take much much less energy to damage. The argument that circuit size leaves less loop area doesn't hold. Circuits have always been vulnerable at the interconnects. Simple current in, current out.

    As one who fought in the die size reduction wars, I became intimately familiar with the smaller chips hanging up where the larger ones didn't. Back then, few people understood how to deal with the issue. Surprisingly, McDonald's Restaurants published one of the better papers regarding the matter.
     
  20. Jan 2, 2012 #19
    McDonald's restaurant ???
    How is it related with EMP ? microwave oven ?
     
  21. Jan 2, 2012 #20
    In the late eighties, you could fall back on military documents, Guru's that gave obtuse answers, or the hard fought knowledge of the fast food restaurant world. With RS232 and RS485 cables being dropped in every other month to automate the burger delivery process, they learned that arcing contacts in fryers, fans switching on/off, or poorly thought wiring would play havoc with so many of their systems that worked great in the lab.
    So, they did the smart thing and researched the causes and migrations.
    In an age when the military was suggesting such things as the "chattering relay" test, this actually proved helpful.
    Dan White and consultants were also useful as was Henry Ott.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook