How to measure high current pulses without the noise?

[pomodoro]

Guys, I need help. I've read a mountain of research papers from the wire explosion days and have tried to make my own sensors for measuring fast pulses of 100A plus from discharging 4-10 uF caps at 1000V plus into inductors.

I have tried building a Rogowsky Coil with coax, as these are supposedly the most noise immune sensors, but have found that pulses from the discharge enter the oscilloscope even when the sensor is a few feet away from the discharge.
Not only that, but a coax cable, terminated with 50 ohms at the scope and soldered short at the other end still picks up the discharge.

Even a bnc hookup coax simply connected to the scope picks up a pulse.
But without a connection to the input, the scope pick up nothing.

I cant figure out how these guys in the 60's faithfully measured kiloamps , which gave only a few mV from the Rogowsky coils, and here I am picking up 100's of mV with a shorted coax.

How do  I make something totally immune to pulses that are not from a current flowing through it , yet retains a good bandwidth?

[MarkE]

Guys, I need help. I've read a mountain of research papers from the wire explosion days and have tried to make my own sensors for measuring fast pulses of 100A plus from discharging 4-10 uF caps at 1000V plus into inductors.

I have tried building a Rogowsky Coil with coax, as these are supposedly the most noise immune sensors, but have found that pulses from the discharge enter the oscilloscope even when the sensor is a few feet away from the discharge.
Not only that, but a coax cable, terminated with 50 ohms at the scope and soldered short at the other end still picks up the discharge.

Even a bnc hookup coax simply connected to the scope picks up a pulse.
But without a connection to the input, the scope pick up nothing.

I cant figure out how these guys in the 60's faithfully measured kiloamps , which gave only a few mV from the Rogowsky coils, and here I am picking up 100's of mV with a shorted coax.

How do  I make something totally immune to pulses that are not from a current flowing through it , yet retains a good bandwidth?

Yours is a basic shielding problem.  Noise can get into your oscilloscope through several paths.  One path is cross talk between oscilloscope channels.  Another path is whatever exposed loop area is at the end of your coax.  A third is the less than infinite shielding effectiveness of braided coax.  A fourth is the effective antenna formed by the AC power cable. 

If you shield the experiment by putting it in an RF screen room then junk won't get into your oscilloscope through the AC power.
If you can't do that, you can put as many ferrite clamp on filters over the coax and AC power leads as possible to mitigate issues with the cables.
You can troubleshoot using just one oscilloscope channel to eliminate cross talk issues while you fix the other problems.

[minnie]

Sounds like you could use a Krytron!!
            John.

[pomodoro]

Cheers Mark, your idea of using those ferrite noise suppressors worked quite well. I used some from old monitor cables.

Unfortunately the Rogowsky coil simply had way too little output and was swamped even by the greatly reduced noise.  However, the same ferrite toroids came in handy as I used one to make a current transformer instead of the Rogowsky coil.  Similar to the Pearson coils but at no cost compared to $ hundreds.

This simple device I made has a lower 3db limit of 2.5kHZ and the upper beyond 5Mhz. It outputs 1V per amp and is very linear .  This huge voltage output  is not affected by any stray pulses and easily measures 300A , which is more than I need for my experiments.

[MarkE]

Cheers Mark, your idea of using those ferrite noise suppressors worked quite well. I used some from old monitor cables.

Unfortunately the Rogowsky coil simply had way too little output and was swamped even by the greatly reduced noise.  However, the same ferrite toroids came in handy as I used one to make a current transformer instead of the Rogowsky coil.  Similar to the Pearson coils but at no cost compared to $ hundreds.

This simple device I made has a lower 3db limit of 2.5kHZ and the upper beyond 5Mhz. It outputs 1V per amp and is very linear .  This huge voltage output  is not affected by any stray pulses and easily measures 300A , which is more than I need for my experiments.

Pomodoro, I am glad the ferrites did the trick.  I would take the transformer down so that the output remains under scope input amplifier maximum: typically ~+/-40V.  I would aim for +/-10Vpp for your hottest signal into 50 Ohms which for a sine wave would limit the dissipation in the oscilloscope front end to ~1W.