Locate broken conductor in shielded cable

[superUnknown]

Any ideas on locating a broken conductor in a braided shield installed under tiles in the floor?
-Tried the 120VAC deathstick (capacitive coupling pen); no dice
-Tried the tone tracer; no dice
Because of shield I think the best chance is high voltage arcing across the fault and trying to locate with IR camera. Foolish idea?

[dfmischler]

?

[Tomorokoshi]

Something like this?

http://www.l-com.com/test-equipment-tripp-lite-multi-function-network-phone-coax-cable-tester#

[Muttley Snickers]

That would also be my first choice of tools, a TDR should locate the fault in a heartbeat, your toner/ tracer can also be used in a number of ways other than what is written in the manual, I would not even consider the arcing technique as a valid option and I would be surprised if it has broken by itself in a fixed or confined space unless it was either shorted out or damaged in another manner such as someone going berserk with a power drill.

I do have a few questions if you don't mind:

1. What is the application for which this cable is normally used for ?.
2. Is this a live cable and can it be disconnected from the feed ?.
3. Can you provide the particular cable specifications as a better guide ?.
4. Do you have access to one or both ends of the cable in question ?.
5. Does the cable utilise solid or stranded cores ?.

[superUnknown]

1. What is the application for which this cable is normally used for ?. Heat trace, resistive cable (inconel? hastaloy?) for in flour radiant heating 1ohm per foot
2. Is this a live cable and can it be disconnected from the feed ?. it is disconnected from mains at the moment
3. Can you provide the particular cable specifications as a better guide ?. it's nuheat's proprietary blend of 11 herbs and spices. 120V hot and neutral each insulated and shielded with tinned copper braid then another shield for the overall cable then a jacket.
4. Do you have access to one or both ends of the cable in question ?. yes
5. Does the cable utilise solid or stranded cores ?. unsure.
N00b Question: TDR what if I omit the "R" and just use the rise time from my 1kHz ref from the scope? The cable has a resistance of 50 ohms and I have a good known cable.
The good cable has a rise time of 135us
1kHz from the scope has a rise time of 4us
One end of bad cable has rise of 53us
Other end of bad cable has rise of 69us (53+69 = 132us, pretty close to good cable, no?)
Is the rise time directly proportional to the length/resistance of wire?
Too simple?

[Muttley Snickers]

Are you sure about the 50 ohms resistance ?, anyway to properly locate the fault with a TDR you still require the NVP or nominal velocity of propagation or known length before you start cutting open floor tiles, use your toner or a multimeter to probe core one and shield then do core two and shield to identify the damaged core, also check to verify that the cores are completely open from each other and the shield.

As mentioned if the cable is snaked around the room then you really have little hope and even then you would still need to expose the break and a reliable repair may not even be possible or practical long term. These pictures below maybe what superUnknown is working on.

[blacksheeplogic]

That would also be my first choice of tools, a TDR should locate the fault in a heartbeat, your toner/ tracer can also be used in a number of ways other than what is written in the manual, I would not even consider the arcing technique as a valid option and I would be surprised if it has broken by itself in a fixed or confined space unless it was either shorted out or damaged in another manner such as someone going berserk with a power drill.

A TDR would not be my preferred method as it only gives distance to fault and that's not much use if you don't know how the cables been routed under the floor as you will have no way of translating the distance to the actual break. Also, most of the cheap TDR are going to be +/- 3 feet.

Use a circuit trace like the Ideal 61-957 to trace the cable and pinpoint the fault. They are a good tool for this type of work.

[vk6zgo]

Your cable is evidently intended to be a complete resistive circuit.
The link that Muttley Snickers attached,refers to measuring from the "yellow to black wires" & mentions that should be the "correct" resistance.

If you have a "good" cable of the same length, there should be a short circuit at the other end.
The cable resistance may stuff things up,but if you get a real display,that will allow you to calculate the velocity factor of the cable.

You may still be battling to get a good location on the faulty one,as "distance to fault" will be complicated by the resistance of the cable,& the fact that one leg may travel all the way to the normal shorted end,then all the way back to the fault.

You can probably find out the route your cable takes by attaching your toner to the cable shield & searching with your tracer.

It's a shame you're not looking for a short,as that would be simple-----just compare the resistance.

[nali]

Your cable is evidently intended to be a complete resistive circuit.
The link that Muttley Snickers attached,refers to measuring from the "yellow to black wires" & mentions that should be the "correct" resistance.

Trace heating cable normally has 2 parallel bus wires carrying the main power with the heating element connected to them at regular intervals to form a distributed load. They definitely shouldn't be shorted at the end! The heater can either be in the form of a coil of resistive wire or a polymer semi-conducting material between the bus wires in the case of self-regulating cable.

It may be possible to measure the resistance between the bus wires from each end and use the ratio of the two measurements to get a ballpark area if you know or can work out the cable routing. Or hire a thermal camera, turn on the power and find the spot where the floor stops getting warm.

Note if this is constant-wattage cable then it'll be segmented (segments identified by notches in the cable) so the offending segment will need to be replaced.

[Rerouter]

I would say float the shield, and hook up the wire to a much louder signal source, Common one is spark plug feed for a motorcycle or chainsaw, with the shield floating the volume of the signal source will cause less attenuation, and should allow better localisation,

[superUnknown]

Forgive my ignorance, by float you mean remove the connection to earth? Then zap the broken conductor with high voltage sparks and I might be able to find the break even through the shield? With an AM radio? 

[Brumby]

The good cable has a rise time of 135us
1kHz from the scope has a rise time of 4us
One end of bad cable has rise of 53us
Other end of bad cable has rise of 69us (53+69 = 132us, pretty close to good cable, no?)
Is the rise time directly proportional to the length/resistance of wire?
Too simple?

I would expect the rise time to be more a function of capacitance (how's your transmission line theory?)  Presuming capacitance is fairly evenly distributed along the cable, you could work out the distance of the break from either end.  Then it's a matter of finding where that point is under the floor.

I do like the thermal camera idea - but that relies on being able to run power through the cable to the break point.  If that's possible, a thermal camera could guide you to within an inch or two of the break, where other methods will be lucky to put you within one or two feet.

[SeanB]

Is your cable a single shielded wire, with a break somewhere in the length, or is it a pair of wires with the heater resistance between them.

If it is a single wire  then the element itself is the resistance, and will be a fine wire spiral wound around a core, likely of glass fibre cord, and terminated at the ends by crimping to a terminal wire. Generally it fails because the end fails, as this is the most stressed part in use, all the rest is well damped with a good heat path to keep it from getting hot. Getting a fault in the mid point is generally a failed outer sheath ( installer kinked it or cut it or dropped a hammer on it) letting moisture in. Easiest way to find the break is to high pot test it using a high voltage, and listening for the arc across the break, as it will be open at 120/240VAC, but apply 2kV and it will arc over so you can use a stethoscope to find the buzzing noise.

If it is the 2 wire type then the thermal camera is going to show you the layout, with the cold spot showing the break. Then you mark it on the floor and power the two wire cable from the other end, and again look to see where the hot ends. Hopefully they are close so you know where to start with the jackhammer.

Hopefully you have a few spare tiles to fix the carnage after the fix. Most likely point for failure is a doorway or floor joint, where the cable has either been flexed by point loads or by the floor expansion and contraction.

[Kleinstein]

Similar to TDR the capacitance can give a hint on how long the cable is up to the broken part. A metal finder might work to find where the cable is running.

Often it's the ends and maybe bends that fail.

[blacksheeplogic]

I do like the thermal camera idea - but that relies on being able to run power through the cable to the break point.  If that's possible, a thermal camera could guide you to within an inch or two of the break, where other methods will be lucky to put you within one or two feet.

A lot of the time it comes down to how badly do you want to find the break verse cost. A circuit trace is the correct tool for this work and is quite cable of pin pointing the fault in the cable. however a good circuit trace is not cheap and the reality of these type of situation where the failure is not close to an access point your better off spending the money and having someone with the correct tools come in and locate the break. Unless it's a very small area and ripping up the floor is an option, it's going to be more cost effective to have someone come in than investing in tools you will use only once.

[superUnknown]

I think I found the break; Reversed a 240V to 24V isolation transformer and fed it with a variac. Started "arcing" at 500VAC RMS 40 mA. But couldn't detect anything with the AM radio even with shield floating. After a few minutes saw a warm spot in the floor with the IR gun. Same approximate spot as I guessed using the rise time from the 1kHz off the scope. Pulling tile tomorrow. Thanks Everyone!!!!

[superUnknown]

In floor radiant heating broken wire RESOLVED. It wasn't exactly where I guessed, but it was under the same tile. Off by 12" as the crow flies or 48" by length of wire. Even a blind pig...  Thanks again guys!

[SeanB]

Pics of the carnage please......