Thought experiment - break 1000Vdc & 1000Arms in <1mS !!

[MagicSmoker]

In my opinion, to break 1000Vdc § 1000A (dc and not rms because rms is only if there is ac current or ac+dc current) in <1ms, that's means that this current must be interrupted in the load.

Interrupted = no current at all

Then you can't Use a suitable commutation diode to provide a path for the inductive current to flow after the IGBT is turned off, (this what we call "free wheeling diode")  because inductive current will flow in this diode and in the load still after IGBT has interrupted the current, during a time depending of the time constant L/R of the circuit.
This can be more than 1ms.
In an inductive circuit, you must make a "hard" breaking of the current to minimize the current breaking time.

You can use a capacitor on the supply side of the IGBT to provide a pathway for energy stored in the supply inductance. On the load side a gas arrester tube or MOV or the like can be used to allow the flyback voltage to climb to a value that both the IGBT can tolerate and which provides for rapid demagnetization of the load side inductance.

EDIT - corrected "freewheeling diode" to "capacitor"

[oldway]

I agree but with 1000Vdc and 1000A, we deal with high power.
I don't know if there are gas arresters tubes or MOV's for such a high voltage and high current.

[MagicSmoker]

I agree but with 1000Vdc and 1000A, we deal with high power.
I don't know if there are gas arresters tubes or MOV's for such a high voltage and high current.

Oh sure, there are plenty of MOVs and GDTs that can handle this level of power which you can buy straight from the usual distributors. E.g. - a Littelfuse BB60 series MOV or Bourns 2097 series GDT.

[max_torque]

Lots of great ideas, and the testing of those ideas would be a LOT of fun!  (from a suitable distance away!  )

The problem with the "explosive" options is sourcing the explosives, but then i wondered about using a gaseous explosive, say Hydrogen/oxygen or even easier to source Oxyacetylene.  iirc, gaseous explosions can be very quick indeed, and are easily triggered, using perhaps a modified flash tube or something similar?

Hows about a gaseous explosive ram, ie a piston in a cylinder, being used to fire a projectile through a conductor, or using it to break the conductor?  A secondary compressed air blast could be used for arc extinguishant for a few sec after the conductor is severed?

 O0

[MagicSmoker]

...Hows about a gaseous explosive ram, ie a piston in a cylinder, being used to fire a projectile through a conductor, or using it to break the conductor?  A secondary compressed air blast could be used for arc extinguishant for a few sec after the conductor is severed?

I seem to vaguely recall that flame front propagation isn't nearly as fast in a gaseous mixture as it is in solid explosive compounds. Maybe somewhere around 20-40m/s inside a piston engine, depending on the fuel, compression ratio, etc.

Also, I thought the whole point of this thought experiment was to completely disconnect a 1kV/1kA load in 1ms or less - I would think that any kind of chemical reaction - whether an explosion or merely "rapid deflagration" - would be too slow by an order of magnitude or two.

[T3sl4co1l]

I really wonder how you would do this with an hydrogen thyratron .

As far I know, the thyratron has no turn off capability, nor possibility to interrupt a dc current.
It should be a forced comutation circuit as used for fast scr's, but deionizing time is greater than Toff of SCR's (Thyratrons :50 to 500s , fast SCR's 5 to 100µs)

Food for thought.

But no, you don't use SCRs for this, you most likely use IGBTs, a clamp diode, and a large enough snubber cap and resistor.

As mentioned earlier, the problem is underspecified.

As soon as you have an estimated maximum inductance figure, we can choose the IGBT, diode, capacitor and resistor, and all that's left is for you to build it (and not screw it up by adding more inductance ).

Tim

[Alex Eisenhut]

I really wonder how you would do this with an hydrogen thyratron .

As far I know, the thyratron has no turn off capability, nor possibility to interrupt a dc current.

As a shunt, to short out whatever a slower pass-device won't cut. Thyratrons have tremendous peak current handling, subby didn't specify how often he wanted to do this...

[Cerebus]

Hmm.  What about a meter long 'fuse' carrying the current made of heavy wall copper pipe, packed with copper thermite and hermetically sealed, with det cord strung below it?   Fire the thermite and the det cord, and the pipe departs explosively as copper vapour. while the blast from the det cord blows the vapour out of the gap.

Interesting idea, we though about something similar, but actually, we weren't sure how quickly something like that could be detonated?  I know detcord burns at a fanastic rate (km/s iirc), but initiating it is actually quite slow i think?  (speed for typical primer cap?)

Search for "exploding bridge wire detonator". Very fast and potentially very accurately timed. Will quite happily detonate high explosives like detcord directly without any intervening pyrotechnic help. Be aware though that this is exactly what you need to trigger the conventional explosives used to generate a compression wave for an atomic bomb. Hence the high voltage parts necessary to control triggering them are tightly controlled and trying to obtain them can bring you to the attention of the kind of three letter agencies you don't want looking at you!

[LeonV]

How about a large copper rod, say 1 or so meters long, clamped between two electrodes. And some impact device to whack it out.
Like a rocket sled, but more practically, a spring loaded hammer.
Or have the copper rod spring eject-able, and make a release mechanism on the electrodes.

You could have the copper rod get slammed into the other current path, but the electrodes that it gets slammed into would need to be shock absorbed, with water barrels or something, Maybe a sand dune.

[Ian.M]

The problem with anything mechanical that doesn't involve lots of high explosives, is the acceleration required.   To clear a 1m gap in 1ms, assuming uniform acceleration from rest requires a final velocity of 2000m/s (Mach 5.8 ), and an acceleration of 2*10⁶m/s² (>200000g).  Assuming there was some method of suppressing the arc so that a 10cm gap would be sufficient, the situation improves slightly, as the final velocity is now subsonic, and it *ONLY* requires a little over 20000g acceleration, comparable to a rifle bullet being fired.

[tautech]

The problem with anything mechanical that doesn't involve lots of high explosives, is the acceleration required.   To clear a 1m gap in 1ms, assuming uniform acceleration from rest requires a final velocity of 2000m/s (Mach 5.8 ), and an acceleration of 2*10⁶m/s² (>200000g).  Assuming there was some method of suppressing the arc so that a 10cm gap would be sufficient, the situation improves slightly, as the final velocity is now subsonic, and it *ONLY* requires a little over 20000g acceleration, comparable to a rifle bullet being fired.

Most seem to have overlooked the old style of arc suppression using contacts immersed in an oil bath. 

[zl2wrw]

At least in this part of the world, "Exploding Bridge Wire" detonators (and explosives in general) are probably much harder to get hold of than the electronic components required to fire them.

A big problem with anything explosive, or even high speed deflagration (eg "smokeless" gunpowder) is that the hot gas given off by the chemical reaction is already ionised, and thus electrically conductive (google "flame rod sensor"). The nitrogen gas released by an automotive airbag gas generator (decomposition of sodium azide) *might* be cool enough to not be ionised (relatively low energy reaction compared to other pyrotechnics), though an airbag inflator will likely be far too slow a source of gas: https://en.wikipedia.org/wiki/Airbag#Operation

[Berni]

Explosives are actually being used for interrupting large currents in certain fuses.

Interrupting 20kV or more on high voltage lines is not an easy task for a fuse. To help it they put an explosive charge inside the fuse that upon tripping pushes the hot arc out the bottom of the fuse and away from the terminal on the other side. They are usually used on top of poles so that if it trips it wont hurt anythyng or anyone.

Testing this is actually doable since a capacitor bank the size of a night stand could provide the juice, and more importantly it becomes empty at some point in case the device failed to interrupt as maintaining the 1000A for 1 second would be pretty spectacular.

I been in a lab where they test fuses and similar high power equipment, they are next to a hydroelectric dam where they are wired in to there local substation where they can pull the huge thousands of amps currents. Or they can put the power in to a giant 2 or 3 story tall power supply that generates 100s of kV for insulation testing high voltage equipment.

[LaserSteve]

A large array of "Hard" modulator tubes in parallel.  As "Hard" modulators (Big Triode or Tetrode)  are in the realm of accelerators, nuclear research,  classified "old school" military widgets, and old style large radars, you may not find much on them these days.

There are some other technologies that can do this. ;-)
Steve