Become Very Unpopular Very Fast With This DIY EMP Generator

Taking a break from his book, “How to Gain Enemies and Encourage Hostility,” [FPS Weapons] shows us how to build our own handheld EMP generator which can be used to generate immediate dislike from anyone working on something electronic at the hackerspace.

The device is pretty simple. A DC source, in this case an 18650 lithium battery cell, sends power to an “Ultra High Voltage 1000kV Ignition Coil” (as the eBay listing calls it), when a button is pressed. A spark gap is used to dump a large amount of magic pixies into the coil all at once, which generates a strong enough magnetic pulse to induce an unexpected voltage inside of a piece of digital electronics. This usually manages to fire a reset pin or something equivalent, disrupting the device’s normal operation.

While you’re not likely to actually damage anything in a dramatic way with this little EMP, it can still interrupt an important memory write or radio signal and damage it that way. It’s a great way to get the absolute shock of your life if you’re not careful. Either from the HVDC converter or the FCC fines. Video after the break.

94 thoughts on “Become Very Unpopular Very Fast With This DIY EMP Generator

      1. This is why the ones that work really good don’t end up on youtube…. no record of the test… builder takes a while to save up for new camera, phone, PC, cable modem, wifi router… … …

        Pro-tip, film it on a wind up 8mm in an old quarry with only the equipment you want to destroy on hand, and your vehicle parked well back from the lip… :-D

    1. Meh. The FCC doesn’t care. An article in a recent issue of QST told of a report of interferance so bad that it disabled nearby cell phones. It was reported to the FCC, who simply ignored it.

      1. Somewhere a while back I read an article on an area that was experiencing cellular outages at routine intervals, interfering with emergency communications. Finally turned out to be some old guy with a super powerful cell jammer in his car because he was tired of people driving while on their phones. Obviously I don’t care enough to find the article but I don’t think they jailed the old guy, just a firm chastising and seizure of the equipment.

  1. Isn’t the electric field generated by a coil proportional to the CURRENT? Bumping up the voltage helps overcome things like coil resistance, but adding a high voltage capacitor to increase current might help too. It would be interesting to see some measurements of peak E & M field strengths at some interesting distance to determine the effectiveness of the design. A spectrum analysis would be interesting to show where the bulk of the power resides in the frequency domain.

    1. The coil’s function is to modulate the frequency of the spark gap’s output, and the capacitance controls the power output. The spark gap itself is the only thing radiating EM. The only reason this infernal gadget works is by screwing up this power circuits in his devices with the magnetic induction from that coil. Who knows what frequencies the spark gap is emanating, but I’m sure is would at least jam someone’s Blue Tooth headphones from right beside them – until they notice the loud sound coming from the strange gadget in your hand.

          1. @paul:

            There are non nuclear EMP devices, google for a “flux compression generator”. But still a one way device, as it needs high explosives to work.

            The device in this article seems to be more of a poorly understood and tuned spark gap transmitter. Probably the air core coil resonates with the output capacitor of the “megavolt” HV generator. Up to now I have seen similar devices advertised as “400kV”. Which would already need insulation thicknesses of ~13mm of epoxy or similar and not <0,5mm of PVC wire insulation. From the spark length of ~1cm you could estimate roughly 20kV.

            So to get some higher voltage at least you should add a tesla coil secondary and treat or modify the air core output coil as the primary :-) Can need some additional primary capacitance.

        1. The last real world EMP design I saw involved destroying it’s inductor on activation. This was a nice seconday tesla coil size coil that an expendable component kinda like ammo. It used a fine silver wire that the would vaporize under the high current. This part was in a container with a suppressing medium like sand to prevent the plasma from continuing to conduct and make a sharper impulse. It would make EMPs that actually ruined electronics,not reset them. With that in mind, this is extra disappointing. Although I guess some ne’er-do-well could give someone a bad day if the person happened to use electronic locks.

      1. I strongly suspect that if you look at the spark gap, the coil, and the stray capacitance you will see oscillation and that this is a broad band, fairly modest, transmitter. We used to make them in ancient times with an LM305 to regulate 5V and a 7400 chip with the NAND gates connected in a loop to oscillate. They ran at the max rate the process could handle and the less than squarish waveform had harmonics of all sorts. It would screw up any broadcast TV and radio within 20 feet or so.

        An EMP would be a lot different, at least nuclear bomb type, since it is an excitation then decay in surrounding air molecules (they won’t work in space, only a certain altitude range where density is right). Someone correct me if there is any oscillation. As far as I know, an EMP is more like a rectangular function or Gaussian pulse.

        1. Your description is almost accurate. Do not forget the Compton effect and many other factors shown on the following web site. — Looking for folks that can redo 1974 Fortran “nuclear”r equations into Python or C++ or C# or Visual Studio or Code:Blocks ( about 2 pages ). Equations are in a doctoral Thesis.

    2. This is not very much different from the first radios made. It’s just a spark gap transmitter that blasts out white noise which over powers radios in it’s area. While technically an EMP, you’re not going to be frying most electronics.

      1. my tig welder was made in the 90’s it has a spark gap, cars prior to the distributors used spark gaps, spark plugs still use spark gaps.

        you may be right, but your phrasing is wrong.

          1. Most Tig welding, save aluminium and magnesium, is done with DC. The spark gap he is referring to is used by the high frequency start circuit…it uses a low power high voltage arc to start the high power voltage welding arc. I have seen the HF start glitch mp3 players and phones before…didn’t seem to happen with the HF start turned off.

          2. Take it a step further – my MIG, TIG, stick AND plasma cutter are all very high power spark gaps! There’s a brutal amount of sparking going on inside the above mentioned distributor cap (the rotor never contacts the cap leads).

          3. I don’t know if they “technically” fall under the category of spark gaps, but I do know that a single welder can interrupt numerous amateur radio bands from hundreds of feet away.
            Technically, it is the spark-gap transmitting radio that was outlawed 90 years ago, not spark-gaps of all kinds. The reason they were outlawed is because a single one could occupy a huge portion of radio spectrum, and the new (at that time) vacuum tube based CW transmitters could be tuned much more tightly and therefor caused less interference across the entire radio spectrum in use at that time, allowing for more (particularly commercial) users in the same amount of spectrum. Source: 400 Meters and Down by Clinton DeSoto (as far as I remember it).

          4. Yes. A colleague of mine has a pacemaker and while working in a metal processing company he avoids walking through the wielding section or at least tell the workers not to roll up their spark wielders’ cables in coil. Which is what his doc told him. True story.

          1. old fashioned radio transmitters only ran at speeds up to 1 mhz, so i will take your word for it that a welder wont be as much of a problem, that said anyone who has welded close enough to a radio will tell you that they do indeed put out a reasonable signal, enough to affect radios in close proximity at least.

          2. The conducting plasma behaves as a negative resistance and excites (parasitic) tuned circuits. In a spark gap transmitter this is the primary function. But how to tell the hot plasma in a welding arc not to do this? :-) The cables have inductance and capacitance. Probably the Q is lower, so it is more wideband noise.

          3. The most important part: In a spark gap transmitter, a lot of the design goes into blowing the plasma away and interrupting the spark. A hot plasma has a fairly low resistance and will remain conducting between the 120 pulses per second from a 60Hz AC arc welder. So some noise, yes, but strong oscillations, no.

            A spark gap oscillator works by either the capacitor charging up, then discharging through the coil with the spark acting as a switch, or by dumping power into a capacitor/inductor circuit with again the spark gap acting as a switch.

            In the first, the ringing (oscillation) occurs during the first part of the presence of the arc.

            In the second, the ringing (oscillation) only starts when the arc goes out.

            Nikola Tesla designed spark gaps that used a magnetic field created by the current to blow the arc out. Rotary spark gaps break the arc by moving the points away from each other. Sometimes compressed air is used. Or a couple of permanent magnets set up a field across the arc, forcing the arc out of the gap as it forms.

  2. Haha, Genius.,
    I was working on a driver for a coil a while back and gave up as while tuning it I kept causing my mouse, USB and keyboard to lock up and drop out. was not as simple as this one. Oh and because it was Arduino driven you can imagine the fun there.

    1. Few days ago someone asked on the electronics forum I visit how to make such a device, named slot machine jammer. It was just a single transistor CW transmitter with loosely coupled antenna via transformer wound over two 9V batteries. It actually works on some slot machines. If they are at least 20 years old. On related note I tested once zapping old payphone with piezoelectric lighter to get free calls. Worked only with very old one that had classic dial – one had to remove cover from it and zap the screw that connected dial to internal mechanism…

  3. I read once on other forums an story about a guy who made a DIY EMP for “testing and experimenting”. The guy rushed the maths part, took some parts from an old washing machine and followed some sketches he found on the internet. After assembled, he put his custom EMP device near a wall of his improvised workshop, then he put an old magnetic band card right in front of the EMP device (to see if It can destroy the info on the card) and took away to a safe distance every electronic device he had on the room.
    When he turned on the EMP Device he was on the other side of the room with all his gadgets, but he found out that he shouldn’t have rushed the maths. The EMP was WAY MORE powerful than he thought, It ended destroying the data on the card, but also destroyed all cards on his wallet, his ipod, a hard disk, some portable consoles, and other electronic devices, but the worst (or the best) part was when he heard his neighbour’s son yelling “Dad! there is smoke coming out from the TV!”.
    Yes, the guy lived in a flat and forgot that the living room of his neighbour was in the other side of the wall where he put the EMP device.
    That was his first and last DIY EMP, he realized how dangerous It was and he thanked god nobody with a peacemaker were in the surrondings.
    His neighbour never knew about the EMP and thought his new plasma TV broke due a manufacture defect.

    1. I’m calling BS on 98% of that story.

      HDD casings are built out of mu metal (an alloy of nickel, iron, & 11 herbs & spices) which ‘absorbs’ the magnetic field of the 1-1.4T magnets used in the voice coil of the R/W head.
      Furthermore your average TV is ~1m/3ft across. The field gradient needed to generate substantial enough voltage to blow caps & fry MCUs is well beyond what anyone but Tony Stark could cobble together from a box of scraps.

      Erase a few credit cards, sure. Fry electronics, not a chance.
      The 12th-hand nature of this story also casts obscene amounts of doubt on it’s veracity.

      1. This is somehow inconsistent. Why should the casing be made of expensive metal to prevent the escape of the magnetic field of the voice coil drive, while the platters INSIDE the casing could use protection?

        It is not. It is often aluminum and/or stainless steel. The field of the magnets of the voice coil drive is mostly confined between the magnets, where it can do its work together with coil. There is no other shielding except a quite close magnetic circuit made of steel.

        I still would not believe the story about fried HDs or TVs. A cheap magnetic card sure. I had data loss on a ski-lift-card one time, because the wind bounced it around over the speaker of my cellphone.

        1. Tear apart any HDD and you’ll find mu metal. I’m sure in recent years the amount has been dialed back to just around the voice coil but it’s still there.

          >>Why should the casing be made of expensive metal to prevent the escape of the magnetic field of the voice coil drive

          Again, it’s not the whole case, but the portion around the voice coil most assuredly contains mu metal or a similar alloy because of the way mu metal shunts magnetic fields. Part of the reason the magnetic field is strongest between the magnets is because the magnetic shielding shunts the other half of the dipole. Dangle a paperclip on a string near a powered off HDD and I bet you won’t see it deflect much, thanks to magnetic shielding provided by some nickel-iron alloy.

          1. The paperclip won’t deflect much due to proper design of the drive magnets. The field is mostly contained between the two poles. Not due to magnetic shielding. mu metal is not used in hard disk drives. It is not required in this application due to the heads only sensing sharply alternating fields, and the head coil amplifiers rejecting low frequency changes. Modern GMR heads are more sensitive to external fields, but as data densities increase, so does the frequency of the field changes, so the lower frequencies are easier to filter out.

    2. The guy lied :P

      With almost all modern chips containing clamping diodes to prevent damage from ESD, it’s quite hard to use an EMP to induce enough current even on very weakly driven input lines to actually deal substantial damage, all you get is a very brief spike of interference, at best.

      A design like this, does a similar job, but repeatedly and at high frequency, so it can certainly disrupt normal operations if it gets close enough, but it’s not an EMP at that point, just electrical noise, you can get the same effect going close to an electric motor, even worse if it has a bad cap. It’s also quite easy to shield against, a metal case or even something like chicken wire should do quite a good job at defeating this “weapon”.

      The only real damage comes from when/if the noise impacts the feedback for a power supply, removing its capability to correctly regulate the output. This is a long-shot at best, as the noise has to a) Reach the power supply unit itself which usually has some form of shielding to prevent *IT* from causing this problem with other devices, b) Have sufficient energy to overpower the feedback resistors for the power supply (feasible, but inverse square law means you have to be pretty close), and c) most everything uses switch mode power supplies these days – your microwave, washing machine, phone, computer, alarm clock, etc – this noise would impact all of them, so it doesnt make much sense that expensive high level digital electronics would suffer, where other more fragile hardware, such as the power supply to the unit itself, would not.

      1. Oh, I didn’t mention the magnetic stuff too – yeah, just straight up no to that. If the field was strong enough for that he would’ve blacked out his entire city from power consumption and turned all the metal nearby into a set of high speed bullets flying directly towards his generator.

        Hard drives and credit cards require a remarkably strong magnetic field to set/damage, they only work because the reading read is in actual physical contact with the media, the distance is a number of atoms, not meters.

  4. It looks similar to a spark gap transmitter from 1887 (illegal in most countries due to EMI since about 1917)

    Thankfully with the antenna chopped off and the high voltage capacitor removed.

  5. Looks to me like a design that works by accident and not in the intended manor
    You want to insert a capacitor to bump up your current, tune your coil and shield your electronics

    Looks to me like the spark gap is doing all the work and the coil is just acting like a big resistor …

    1. Looks legit to me. Assuming 6 V batteries, that’s 36 V. Keyed on creates a pulse that gets transformed up to high voltage, sparks, and the broadband signal is passed through an LC tank, radiating away out the antenna while it rings down.

      If you put the sparking part in a box so that it didn’t radiate so much, this could be half plausible. But we have a lot mellower ways of making short bursts of broadband these days, so the spark is probably a waste of time except as far as it looks cool.

      Shouldn’t this thing make a pulse on key down and key up? That must sound wierd.

      1. I would assume 1,5V (Leclanche, ‘dry cell’) or 2V (lead acid) batteries, but anyhow. I believe to see an interrupter (like a doorbell). So the coil produces HV (and the transmitter RF) as long as the key is depressed.

  6. This is why you need to go to uni and do engineering. Pathetic circuit for what he’s attempting to do. Anyone with a base knowledge of electronics knows that this wont do fuck. Pacemakers/IAD’s also have safeguards. I mean we can dump a pt with one of these in an MRI with appropriate settings and all is fine. You want to see EMP? Try dousing an MRI.

    1. “Maths” is often the british english equivilent of American “math”.

      An example in American English:
      “I couldn’t solve the problem because I didn’t understand the math”

      British English:
      “I couldn’t solve the problem because we ran out of tea”

  7. Hm… This seems way weaker then actually going to the lab and apply EN 61000-4 EMI to your devices. Just use the ESD-zapper gun, crank it’s voltage to around 15kV (thats a bit more than the standard asks, but usually what we test our devices with) and do air discharge on the vertical coupling plate or the reference plane below your device. Most badly designed/layouted microcontroller circuits (like all the arduinos…) will see a reset, devices designed for industrial compliance however should be pretty much immune to such events.
    This project is not an EMP generator, it’s just a pretty weak and random EMI generator.

  8. I found a while back that a simple regular photoflash can mess up devices, specifically cheap calculators.
    Pretty sure that its a combination of very high peak intensity of light and short rise-time that is doing the damage, google “Pi Xenon DEATH FLASH!!!” for how this can crash some laptops, SBCs and modern phones even through the LCD.

    Its interesting to note that my idea of a simple LED based photosensitivity tester based around LED strips harvested from old broken laptop panels could work here, simply put a low value resistor across each 4 LED bar to reduce the intrinsic capacitance and achieve minimum pulse width.

  9. The “Pi Xenon DEATH FLASH!!!”” was no electrical, but only an optical interference. If you covered the voltage regulator with opaque paint, it was not vulnerable any more.

  10. One of these, slightly more powerful, mounted in the trunk of a car with the coil facing rearward… Would be handy for tailgaters… Or make one powerful enough to take out those irritatingly bright eBay HID light kits and LED light bars but not nuke the car’s computer (Unless of course turning that Mercedes that’s been trying to eat your bumper for the last two miles into a paperweight is the goal)

    1. That’s a nice revenge fantasy but unfortunately – no. Notice how close he had to hold his zapper to the phones/Gameboy to get them to glitch? I suspect a layer of aluminum foil would shield most everything from the effects of similar devices. The trunk of a car would be a very effective shield as well. Even cutting a hole in the trunk wouldn’t allow enough energy to escape to cause any mischief.

  11. Funny how he claims it generates EMP without a single measurement of magnetic pulses. I’d be very hesitant to hold those spark gap wires in my hand considering said wires can’t possibly be rated for HV. While I hate to piss on some guy’s experiments, this thing is downright dangerous. Shoddy wiring, Chinese ebay HV generator and Li-Ion cells – that’s a recipe for a Darwin Awards takedown.

    Kinda disappointed he didn’t try this on a Galaxy Note 7.

  12. 1000 kV, eh? Or do you mean 1000000000000000000 fV? Maybe .001 GV. 1×10^6 V? The metric system prefixes are there for a reason. It’s common practice in applications to enumerate values in the lowest number of significant digits, per mille. So – 1 MV.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s