Pulsed Power and its Applications

Pulsed power is a technology that consists in accumulating energy over some period of time, then releasing it very quickly. Since power equals energy (or work) divided by time, the idea is to emit a constant amount of energy in as short a time as possible. It will only last for a fraction of a second though, but that instantaneous power has very interesting applications. With this technology, power levels of more than 300 terawatts have been obtained. Is this technology for unlimited budgets, or is this in reach of the common hacker?

Consider for example discharging a capacitor. A large 450 V, 3300 uF electrolytic capacitor discharges in about 0.1 seconds (varies a lot depending on capacitor design). Since the energy stored in it is given by 1/2 CV², which gives 334 Joules of energy, the power delivered will be 3340 watts. In fact a popular hacker project is to build large capacitor banks. Once you have the bank, and a way to charge it, you can use it to power very interesting devices such as:

A portable, 1.25 kJ coilgun by [Jason Murray]

Railguns in particular are subject to serious research. You may have read about the navy railgun, capable of reaching a muzzle speed of more than 4,600 mph (around Mach 6), more than any other explosive-powered gun. Power is provided by a 9-megajoule capacitor bank. The capacitors discharge on two conducting rails, generating an electromagnetic field that fires the projectile along the rails. The rail wear due to the tremendous pressures and currents, in the millions of amperes range, is still a problem to be solved.


Another device that uses capacitors for high power pulses is the Marx generator. It is a very simple circuit that allows you to charge a number of capacitors in parallel and then suddenly discharge them in series using spark gaps. Very large Marx generators have been built, for high voltage component testing and other purposes, but it’s also very easy to make a small lightning simulator in under an hour if you have some high voltage capacitors and resistors. Marx generators are in use in the Z machine, a Sandia National Labs project for fusion research, that is capable of shooting 26 million amperes in 95 nanoseconds. Temperatures of 3.7 billion kelvins have been obtained.

The Marx generator is a particular case of a pulse forming network, or PFN. Capacitors, inductors and transmission lines, or a combination of them are used for energy storage in various topologies. Then, the network is discharged into the load via a high voltage switch

Transmission line PFN. By Chetvorno, via Wikimedia commons.
Transmission line PFN. By Chetvorno, via Wikimedia commons.

such as a spark gap or a thyratron. The transmission line PFN is interesting because the capacitance of the conductors in the line is used for both transmission and energy storage. When the power supply is connected it slowly charges up the capacitance of the line through RS. When the switch is closed, a voltage equal to V/2 is applied to the load, the charge stored in the line begins to discharge through the load a current of V/2Z0 and a voltage step travels up the line toward the source.

Compulsators (a portmanteau for compensated pulsed alternator) are another way of delivering high current pulses. They convert rotational energy from a flywheel directly into electrical energy. The compulsator works in a similar way as a normal alternator, but is designed with minimal inductance windings to deliver extremely high currents in very short time periods. There is little information on compulsator design and, as far as we know, no hobbyist has ever made one. You have your homework assignment.

Alternator vs compulsator designs. From Weldon et al.
Alternator vs compulsator simplified designs. From Weldon et al.

The explosively pumped flux compression generator, or EPFCG for short, is a device that generates a high power electromagnetic pulse using a high explosive to compress the magnetic flux. Million of amperes and tens of terawatts of power are produced by the EPFCG in a single pulse, since the device is destroyed in operation.

Croquant, via Wikimedia commons
Steps in flux compression. By Croquant, via Wikimedia commons.

The three basic steps in flux compression are shown above.

  1. An external magnetic field threads a closed ring conductor.
  2. The ring’s diameter is reduced by the explosive. The variation of the magnetic flux induces a current in the ring, which in turn creates a new magnetic field, so that the total flux in the interior of the ring is maintained.
  3. The external and induced magnetic fields add up so that the total magnetic flux remains constant, and a current is created in the ring.

The compression process allows the chemical energy of the explosives to be (partially) transformed into the energy of an intense magnetic field surrounded by a correspondingly large electric current. There are several designs of EPFCG’s. The figure shows the hollow tube type.


Pulsed power is also used in particle accelerators and high power lasers and the technology is rapidly evolving.

If you’re starting out, you may want to experiment with capacitor banks which are a relatively simple way of obtaining pulsed power. But if you do, take all necessary precautions. The power levels can be extremely dangerous.

39 thoughts on “Pulsed Power and its Applications

  1. The Australian railgun program (yes, Aussies did have a railgun program with a bit of funding in the 80s) was powered by the largest homopolar generator ever made. These can generate stupidly high currents, definitely worth looking at.

    1. The basement of the old EE building (ENS) at UT Austin used to contain a homopolar generator used to power the tokamak in the basement of the neighboring physics building (RLM). It was made from a 1920’s era power plant generator and used the mass of the rotor alone.

      It was in the process of being scrapped when I attended, so I wasn’t around when it was in operation. Rumor is that the whole building would shake when it was fired.

  2. I’ve had some fun with supercaps, dumping all the energy into Nitinol wire to get it to reshape in a second or so. It’s one of those things that is unexpectedly fun.

    Another pulsed energy thing I read up on a bit was the National Ignition Facility, which fuses hydrogen with 192 optically pumped lasers. Neat stuff.

      1. I would not use a mercury relay. If that relay blows up, you’ll have a big mess on hand. It is not even a good switch. Too expensive and too slow. Also too much inductance. I’d use an SCR, a triggered spark gap, or make my own switch. You can layer Cu films with kapton inbetween, and then drive a tack through the stack. That will be extremely fast and give the lowest possible inductance. For the triggered spark gap, you can experiment with overvoltage protectors.

          1. An SCR or thyristor is used extensively in pulsed power.
            Do a Google search on thyristor pulsed power. Commercial switches can operate north of 50 kA. They do have rate of rise issues for current so they are not good for sub microsecond pulses. IGBT’s are much better for electric vehicles because you can turn them off.

          2. Clarification, not saying they don’t exist, just they are unavailable to non-commercial customers in single unit and small quantities, without signing a contract for x amount of product plus promise that you’re not going to build a nuke bomb trigger or other weaponised use.

            Best source for the experimenter of high power capable thyristors, triacs, trigatrons etc, is surplus and surplus equipment.

          3. The problem with SCRs is they are painfully slow, most of the energy need for such a project as mentioned above would end up as heat as the SCR was turning on.
            You are better off exploring mechanical switch alternatives. Mercury is a relatively poor conductor.

            At one time I had explored building an aluminum can crusher that would use a magnetic field to crush aluminum cans. I am surprised this article didn’t mention the one here on Hack A Day I found several years ago. Like your wire vaporizer a capacitor bank would be discharged into a load, which in this case would be a single turn loop of buss bar. The magnetic field would cause the can to implode leaving a thin cylinder of aluminum.

            Besides the noise, safety issues, a fast switch, all of which present a significant engineering challenge. I set aside the project as it’s impractical at this time. Also given that my neighbors beginning wetting themselves when a spark travels from my 75 meter dipole to a branch on the yucca tree supporting one end, I thought about how they would react to my using a ballistic switch to dump the current into the one turn buss bar coil in my can crusher design….

            Living in The People’s Republic of Kalifornia just isn’t any fun anymore…


        1. OVPs are not really a _triggered_ spark gap. Once I have seen a design, perhaps in a marx generator. There were several spark gaps made with semi-spherical electrodes, I think with something of 5 to 10cm diameter. The lowest was a “triggered” one. In the middle of the lowest potential electrode an automotive spark plug was mounted. Firing the spark plug with an ordinary ignition coil triggered the sparkgap and that triggered the other gaps, probably by a combination of overvoltage pulse and UV radiation.

  3. I have not seen that footage of the navy railgun tests before. Incredible! The projectile is fishtailing which I think would be horrible for long-range accuracy. I wonder if they have figured out a method of rifling the rails (like a double helix?) to add rotation. Perhaps at these speeds it’s impossible. Maybe the projectile needs to be spun up before launching?

    Also, my favorite capacitor bank discharge application is still Krux’s hard drive platter launcher.

    1. While I recall reading that the problem has since been mitigated, I could not find the article referencing it. Regardless, the problem is not as awful as one may think. The MDA (Missile Defense Agency) partnered with the Navy on rail gun development for the purpose of developing Rail Gun based ABM systems. These are designed to intercept a missile outside of the atmosphere, at apogee of the incoming missiles trajectory. There are many practical reasons for this, one of the most important is TIME. It takes a bit of time, even at rail gun/missile speeds to make it to the edge of the atmosphere, and you need to launch with enough time to meet your target there. Due to limitations of sensing equipment and accuracy of projectiles, what this means is you have to take a guess at where the missile you are tracking will be at apogee, and fire your Kinetic Kill Vehicles/Slugs in anticipation. The inherent error in tracking estimation means that you have to have terminal guidance on board your KKV to make the final intercept course correction. For the MDA rail gun, as I understand it, they placed small explosive charges and a control circuit in the slug which would be used to make terminal course corrections for the KKV. Don’t ask me how you get an explosive and any sort of circuit to survive being fired from a rail gun, I have no idea, and my guess is its Classified. Anyway, all of that is to say that, even if it is fishtailing, its not a big problem because they only need approximate accuracy for the MDA application, and rifling would probably cause more problems because you want to time your course corrections perfectly, which I imagine would be much more difficult if your directional charges keep changing relative orientation to the flight path.

      On the other hand, if you want to hit a target in atmosphere, like another ship, you would probably be better to use an APFSDS projectile and rely on a really good computer to work out the firing solution.

        1. And this was done with “hollow state” electronics (tubes)! And at least in one system with a battery activated by the centrifugal force of the spinning shell which squeezed the acid between the plates. Nice way of arming it to prevent any premature detonation.

          Regarding rifling and spin: I don’t think they would do this. Even the high speed armor piercing kinetic projectiles shot with conventional chemical propellant are fin stabilized. And I am sure this would be more easy in this case. Probably the wear of the rails is more than enough even without the projectile spinning like mad.

          It’s not so much a problem for the guidance system. AFAIK some missiles exist, which have a control vane which can only move in one axis. But this shell is spinning and the actuator is fast enough to move the vane when it’s looking in the right direction.

    2. I suspect rifling will not be possible, certainly not in the near future.
      Rail erosion & rail warping hasn’t been sufficiently solved with straight rails adding rifling to them will likely only compound these problems.

    3. That projectile is not designed to be aerodynamically stable (at least not at supersonic flight)…
      If you do a bit of searching, there was slow-mo footage of a test shot with an aerodynamically correct projectile.

    4. Whilst rifling probably isn’t viable, I wonder if you couldn’t use magnetic fields to impart a spin? Since they’d be perpendicular to the direction of the accelerating fields it shouldn’t slow down the projectile much. You could also use such fields for micoadjustments to aiming (such as counteracting vibrations on the rail).

  4. KEMA in The Netherlands used (or still uses?) a couple of heavy fly-wheel generators that were powered up during a couple of hours to reach maximum speed. Then, they were connected to a shortcut or switch that had to be tested and a huge amount of current was produced which slowed down the generators in seconds. They were burried in concrete to prevent movement of the housing.
    Very odd sound everytime they tested something. The high voltage was produced by the Marx generator.
    Nice stuff, was about 15 years ago I walked there for my internship.

  5. I wonder if that kind of quick release is hard on the longevity of those caps? Also wouldn’t be surprised if someone harnessed the power of a nuclear explosion. The ultimate pulse.

    1. I’d be incredibly surprised!

      Still, it seems to be about the speed of the explosion, not the overall power. Far as I remember a nuclear explosion isn’t all that fast, as explosives go. Plus there’s the problem of your hollow tube being vapourised before it gets a chance to shrink.

      Still, would be great as a way of launching stuff into space, a railgun powered by a nuke-pumped flux compression generator. Your payload would probably vapourise too.

      1. If you have a nuclear device at hand you have probably all the EMP you want. No need for any hollow tube or flux compression generator, which is needed to turn the energy of a chemical explosive into some kind of EMP.

  6. Isn’t there an emerging use of pulsed power in radar and lidar systems for self driving cars? At the distances involved, I’d imagine you’ll need a lot of peak transmitted power to get back a usable signal.

    Also interesting to note that another use of pulsed power in automotive – CDI – is slowly fading away as EVs become more and more ready for the mass market. Ironically, it’s competition from EVs that pushed automakers to continue developing that technology in an effort to stay relevant.

    For something more like what most hobbyists might want to experiment with, what about a pulsed power light source that synchronizes with a video camera to more or less eliminate motion blur? A Xenon flash comes to mind for a pulsed power light source except it doesn’t work very well at the high repetition rates required.

    1. In case of radar, it’s not emerging, but has been in use well before WW2…only difference now is that (nearly) everything is solid-state. Back then, thyratrons were used to trigger the pulse.
      Big radars have peak pulse power over 1MW.

      As for the synchronized light – is incredibly difficult to get a sync. signal out of a consumer camera (or to get it in :P). Only speciallised lab stuff and DIY projects have that feature.

  7. I was a little disappointed with the article for failing to mention inductive storage – which is very useful in some physics setups, and any serious mention of the tons of exotic switches – from spark gaps, to inverse pinch switches, pyrotechnic exploding links (for inductive storage). While I haven’ done anything at large as the Z machine pictured, I have done physics where 6 kJ was discharged in way under a uS…
    The switch is the hard, and therefore interesting part…Always the crucial and usually limiting factor in pulsed power.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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