Finding a Short

| One of our office circuits apparently has a short, and I would like to know | if there is a straightforward way to detect at which segment of the circuit | the short is developing. | | The electrician trace the wire from the end user's workstation to to the | panel, and he could not find any obvious short at any segment of the cable | run. There are many junction boxes in between the panel and the end user's | station. The breaker is tripping about once every two weeks, and the load | on the 20A breaker is only about 8A.

It could be any number of things, including:

1. A loose connection at the breaker itself, heating up the thermal element in the breaker.

1. A bad breaker. A thermal element that has been damaged could be getting hotter than normal, effectively derating the breaker.

2. Someone plugging in an extra load somewhere for a while, like a laser printer. Or maybe just turning it on.

1. An intermittent short in a computer power supply or other load device. Faulty rectifiers have been known to do this internally.

2. An arc fault due to bad insulation, which won't be readily visible. There are many ways such damage could be caused.

| For various reasons putting in a new circuit would be expensive (around | $2K). If possible, I would like to do some detective work to determine | which segment has the developing short, and then use that information to | replace just the cable in the affected run. Is there any device that would | allow us to determine which part of the cable the short is in? For | example, maybe this device could be put next to one of the junction boxes | and actively connect both ends of the circuit, and at least tell us which | side of the wire the short developed on? The device has to be safe for | use in tile ceiling area.

If the short is not there all the time, this gets hard to do.

Try swapping that breaker with an equivalent one elsewhere in the panel. That will do two things. One, it gives the electrician the opportunity to make sure the connections are nice and tight on the breakers. And if the breaker is at fault, the problem will move to another circuit.

If the problem remains in the same circuit, you will need to isolate where. But don't necessarily jump on the circuit wiring as the only possible cause.

Are the PCs in this circuit on UPSes? If not, put them on one that has a good circuit breaker in it. If a PC is at fault, it should trip the UPS breaker first (and only that breaker if it is a "full time" type UPS).

Anything else on the same circuit?

sounds like something intermittent. very difficult to find. usually these things are NOT in the power circuit itself, but in the equipment connected to it. You could try powering the workstation off another circuit and see if the same thing happens. that would tell you the workstation is at fault.

"CHANGE USERNAME TO westes" wrote in message news: snipped-for-privacy@giganews.com...

The electrician already moved the circuit to an available breaker on the panel, and took some of the load off the circuit.

Effectively we only have three PCs on the circuit, all protected by UPS systems.

Isn't there some kind of "breaker in a box" that could be introduced at some leg in the circuit? At least that way we could deduce if that breaker in the box trips that the problem is from that breaker somewhere down to the load. Then we could begin advancing that breaker in a box closer to the actual end stations to determine which one is at fault.

I still suspect something from the wall plug toward the breaker since the equipment is behind a UPS, which I assume has its own breaker.

Sounds like it may be a weak breaker instead.

As I understand the thread, there are 3 PC's on the circuit now, and NOTHING else - and the breaker still trips. Take the 3 PC's that are currently on the failing circuit off of it. Move three other PC's on to it. If the trouble follows the PC's, one of them is bad. If the trouble stays on the circuit, then it is the circuit itself that is defcetive.

If the circuit itself is bad, some re-wiring will be necessary, regardless of cost. This could include abandoning part or all of the original circuit.

I don't know of any device that will do what you described. You could load the circuit with a heavy draw device like a toaster or hair dryer to see if that brings on the failure any more frequently, but I don't know if that might bring about a safety hazard given that we don't know the "health" of the circuit.

On Thu, 16 Sep 2004 11:16:21 -0700 CHANGE USERNAME TO westes wrote: | Isn't there some kind of "breaker in a box" that could be introduced at some | leg in the circuit? At least that way we could deduce if that breaker in | the box trips that the problem is from that breaker somewhere down to the | load. Then we could begin advancing that breaker in a box closer to the | actual end stations to determine which one is at fault. | | I still suspect something from the wall plug toward the breaker since the | equipment is behind a UPS, which I assume has its own breaker.

I saw many many years ago a circuit breaker constructed in the form factor of a light switch. Same yoke dimensions and fit in a 2x4 box. Now days I cannot find such, although I did find a plug fuse in a receptacle form.

If someone can find one of those, maybe that will help.

I'd put a breaker protected power strip between the UPS and its outlet to see if it is a UPS causing the problem.

"CHANGE USERNAME TO westes" wrote in message news: snipped-for-privacy@giganews.com...

Along with the other good suggestions, I'll add one more.

Carefully take the wall plates off of each outlet (one at a time is fine) and, with a strong light, look inside each one. In most commercial buildings the wiring is in conduit so it isn't likely that a fault develops

*between* junction boxes. But if there *is* a fault developing, such as damaged insulation shorting to the box or some such, a good visual inspection will find it. Look for burn marks or blackened wires. This is pretty easy and cheap and it may find the problem.

If this doesn't, then you'll have to resort to the other methods mentioned by others.

daestrom

And it would make sense to do some wiggling of the plugs, and some pushing/pulling/tapping, etc., of receptacles that have nothing plugged into them. It seems VERY unlikely to me that you have a "developing" short circuit, unless you can get it to act up by doing some mechanical aggravations of the wiring. It seems much more likely that you have a piece of equipment that is either malfunctioning or drawing more current at times than you think -- or a defective breaker as someone else said. --Phil

"CHANGE USERNAME TO westes" wrote in message news: snipped-for-privacy@giganews.com...

It's easy to rig up just such a "breaker in a box". Either get an inexpensive power strip with integral (15 amp) circuit breaker or have your electrician rig up (safely) a 15 amp circuit breaker (preferably a "feed through" type) fed from a cord w/male plug that feeds a recepticle (or however many you need max. at each location). Use a 15 amp breaker since you'd like it to trip before the permanent one and you said that the normal load is about 8 amps which I assume is usually fairly constant. I'd use a power strip - they're cheap and I'm lazy. But there is a better way to trap intermittent overload trips.... see below.

Intermittent shorts, especially with a two week period of occurance, in a circuit with "many junction boxes" and probably run through inaccessable spaces will be a challenge to trap. Here's a solution that will work but will cost you about about $100, involve a little minor electrical wiring, and may take quite a few those two-week-breaker-trips to finally reveal the faulty section.

Purchase a "current sensing relay". A good example can be seen at

. Choose a relay with "Latch on High", zero time delay, 3-30 amp adjustable trip. Like the model CR4395-LH-24D-330-X-CD-ELR - see :

After wiring it up, you will need to adjust the setpoint (see below) to somewhere between 10-15 amps (some value substantially lower than the 20 amp breaker but higher than normally present) and locate it at the most convenient places first. The circuit will have to be opened up temporarily to install the device. One wire of the circuit must be passed through the c.t. (current transformer - the doughnut in the picture), then re-connected. We chose a latch type unit so that when it trips it remains tripped until you reset it. The relay could just light a small 120vac lamp to show that it's tripped. Just rig it into the circuit and leave it there until your 20 amp breaker trips - then see if the light is on. If so, the fault current passed through your short circuit trap. You should also prepare an ordinary extension cord for use with the relay to test your individual devices and for initially setting the trip setpoint. You'll need separate the individual conductors at some convenient point so you can pass one of them (the black or white - not the green ground wire!) through the c.t. doughnut. You can later use this cord to test each of the three UPSs. Also, the relay/pilot light should be powered from a different circuit than the one you are testing so the 20 amp breaker tripping doesn't interupt power to the relay circuitry.. See To initially adjust the trip setpoint you need to plug in a test load that gives you between 10-15 amps. That would be some resistive device (like a hair drer or small space heater, toaster, etc - but no motors) with a power rating between 1200 watts (10 amps at 120 volts) and 1800 watts (15 amps at 120 volts). Just dial the setpoint to maximum, plug the known load into your rigged extension cord, plug the other end into a kitchen outlet at your house (so you don't trip the circuits at work), and adjust the setpoint down until the relay trips.

There are cheaper varieties of current relays out there but the CR Magnetics are nice since they already have a built-in c.t., latching if you want, and adjustable or zero time delay. You could shop around for others on Google or maybe even ebay. Don't be intimidated - current sensing relays are really pretty easy to rig up. It took me longer to type this than to wire one up.

Good luck, Perion

We had a similar problem a few years ago. Moved the circuit to a different breaker and the problem persisted. Did an inventory of devices on the circuit and total current should have been well under the capaabilities of the 20 amp breaker. Finally mounted a 10 amp fuse on a J-box cover. This was moved around to different locations until we found a place that was blowinng the fuse. One of the ladies that worked there had a small blow dryer that she would plug in to dry her nails. This thing pulled about 12 amps..