determine device that is arcing

[dmcguire56]

I often must troubleshoot high-voltage systems that fail by arcing. The system tells me when an arc occurs but does not provide clues as to what part of the HV system arced. The basic components of the HV system are an inverter connected to a voltage multiplying tank and finally a x-ray tube. All this is moving while in operation. I was wanting to make a device to determine the source of the arc. The components are connected by cables. The tank is about 24 inches from the inverter and the tube is about 30 inches from the tank. I was hoping that a device with three equal length leads, running to each component could detect which one arced. Even though the arc would be cascaded to each component there would be a few nanoseconds difference in time. Does anyone have an idea how I might approach this? Thanks

 

[user_88]

Does anyone have an idea how I might approach this? Thanks

One interesting phenomenon related to spark generation, and possibly the arcing to which you are referring is the generation, or radiation of wide band radio frequency noise .....

In fact, the original radio transmitters ... circa 1900 ... were actually just big spark makers ... There was not any particular radio frequency back then....just a lot of RF noise that could be detected at a distance.

It would be interesting to take a portable AM radio .... with a basic internal antenna ....and see if it was possible to discern any directionality as related to your particular arc events.

.... Even if the AM radio is not helpful, there may be some way to construct a specialized antenna .... maybe a few turns of wire wrapped in a coil .... and try to detect any direction or orientation effects .....

Do you have an oscilloscope? This device might be helpful .... An oscilloscope could be used with a wire coil antenna .... in place of the transistor radio .... much more sensitive...

... One other question ...can you approximate the current magnitude involved in the spark events? What sort of power levels are you working with? If there is a large current flowing through the arc gap, there may be a localized magnetic field to consider, in addition to the Electric field .... generally these two fields are geometrically orthogonal to each other.
The magnetic field has a very short effective distance, while the electric field can actually be detected at great distances. ... It may be that the magnetic field of your arc event would be more easily detected with a coil antenna/transformer of some sort.

 

[dmcguire56]

thanks

i will consider your ideas -- thanks again