ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: Pulsed Current tracing - divide and conquer  (Read 1244 times)

Mike Sokol

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3354
  • Lead instructor for the No~Shock~Zone
    • No~Shock~Zone Electrical Safety
Pulsed Current tracing - divide and conquer
« on: September 26, 2014, 01:32:59 pm »

I posted something about this a while back (before this AC Power & Grounding Forum started, I think). A buddy of mine just called to tell me he used my pulsed current testing method to locate a short circuit in a large building. See below:

I have a pretty cool trick to figure out where the current is flowing in grounds and determine if they're bonded in multiple locations. Nearly 40 years ago I was an IE working for Corning Glass and we had a single 120 volt AC control transformer that fed a separate alarm circuit that went to a dozen different packing lines spread across 50,000 square feet of production floor. The control transformer had lost its G-N bond, and somewhere in the thousands of possible connections one of the packing machines had shorted a wire to the chassis. We knew this had happened because all the white neutral wires now measured 120 volts and all the black hot wires measured zero volts. When we reconnected the G-N bond in the 2KVA transformer, it would just blow the fuse, so that was no help. I tried adding a static resistive load between the G and N connections that I could trace with a clamp-on ammeter, but there were so many other currents flowing in the wires that it was hopeless.

The plant manager was not very happy with the idea of shutting down the entire production floor for a day just to find a single short circuit somewhere in hundreds of machines, so I came up with the idea of using a time delay relay wired to click on and off once a second, and used this to connect a coffee pot across the Ground and Neutral bus. This gave me a nice 10 amp, 1 second pulse we could easily trace with an analog clamp-on ammeter. However, most of the wiring troughs were overhead and required moving pallets of glassware out of the way to reach them, I came up with a divide and conquer strategy. Since there were 12 main production lines, we went to the middle of the production floor and tested to see which way the pulsed current was flowing. That quickly eliminated half (25,000 square feet) of the possibilities. I then divided the remaining half into quarters, found the current, then eighths, found the current, and so on until just a few tests later we located the exact machine where the pulsed current was sinking to. As predicted, someone had pinched a wire under a bolt, so we just cut out the damaged section of wire and all was well. This test only took a few hours total, including the coffee break we took in the middle after the coffee pot had brewed properly. That was good coffee...

The point is, my boss was sure it would takes weeks of work to find this problem, but with a little novel thinking I did it with a single electrician to run the fork truck in under two hours. I have a similar trick that can be done to locate multiple G-N bonding in buildings even when you don't know where all the wire has been run. That will be the subject of a future posting.

Some fun, eh? Mike Sokol



Logged
Mike Sokol
mike@noshockzone.org
www.NoShockZone.org

Jonathan Johnson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2990
  • Southwest Washington (state, not DC)
Re: Pulsed Current tracing - divide and conquer
« Reply #1 on: September 26, 2014, 02:18:32 pm »

When I was growing up on the farm, I would use a portable AM radio tuned to static to help find shorts in the fence. As I approached the short, the pulsing audible in the AM radio would increase in intensity. This is similar to how lightning strikes can be heard on AM radio; the arc generates loads of broad-spectrum radio frequency interference.

An electric fence charger that delivers multi-thousand-volt pulses (it's a high impedance source, so current is limited) may not be the best tool to use for detecting short circuits, but it certainly can make the shorts obvious. However, if it's a low-resistance dead short, there may not be sufficient arc to create the RFI.
Logged
Stop confusing the issue with facts and logic!
Pages: [1]   Go Up
 


Page created in 0.062 seconds with 24 queries.