Designing a Burn Down Transformer

[sinB]

Hello EEVblog, this is my first post.

I've had this project in mind for a longer period of time.
My goal is to build a crude, but safe, Burn Down Transformer for locating faults in 240V heating cables in combination with a thermal camera.

Example:
A heating cable is damaged somewhere in the floor, causing the fuse to trip. The cable is then hooked up to the Burn Down Transformer, and a thermal camera is used to observe the exact position where the heating cable gets cold. The heating cable can then be repaired at that exact point.

There is of course commercial equipment available, like the Baur ATG2 Burn Down Transformer
(https://www.alpha-electronics.com/power-engineering/cable-fault-location/baur-atg2-burn-down-transformer.html).
Those are expensive, but looks relatively simple to build.



Here is the problem: I can't seem to find any designs or hardware documentation about Burn Down Transformers, not even simple sketches.

Does any of you have experience related to this or other inputs that can lead me in the right direction?

[TERRA Operative]

I'd start with an isolation transformer (To get around earth leakage faults) and a beefy dimmer/current limiter.

It looks like a chunky PSU with a few mods to make is specific to this application.

[digsys]

Agree with the Isolation transformer - A second option, which I've used in the past is a say 5A Variac, with VI meters. Just dial up slowly until you get to a reasonable current, then wait till the wire heats up. IF the short is very close by, and you can't dial up enough voltage, just wire in a series lamp etc It may even be good practice to leave it in. You'll soon find out how much current you need for detection.

[MagicSmoker]

Calling something a "burn down transformer" sounds like the marketing fail of the century to me so I clicked on the link. It appears to be a current-limited, AC voltage source which can output switch-selected combinations of voltage * current that maxes out at around 2.2kVA, ranging from 60VAC * 32A to 10kVAC * 0.16A - yes, 10kVAC...

I can think of a few ways to do this off the top of my head, but none of them are cheap. How much is one of these things, anyway*?



* -  one must request the price, which is always bad news for the budget...

[Marco]

Calling something a "burn down transformer" sounds like the marketing fail of the century

Since they actually burn the broken cable to create a more conductive path it seems accurate. It seems a bit foolhardy to me, but I don't do power electronics.

That said, it's massively overdesigned for finding a short in a heating cable. Using an isolation transformer with a lightbulb (and possibly a variac for more control) seems the most straightforward method.

[OwO]

This is a cable for a heater, so it normally carries ~10A. I'll assume the cable normally dissipates much less than 20W of power, so its total resistance is at most 0.2ohms (10^2 * 0.2 = 20W).

If we assume 20W of dissipation is enough to find the fault, then at most we need to output 2V because 2**2/0.2 = 20W. There is no need for the current generator to be directly mains powered. I would simply build a 12V to 2V 20A current limited buck converter and use an off the shelf 12V 3A power brick.

A transformer is the last thing I would consider for an application like this. Do not play with mains when not absolutely necessary.

[Marco]

If it's a mains powered heating cable it's a little higher resistance than that.

[MagicSmoker]

Calling something a "burn down transformer" sounds like the marketing fail of the century

Since they actually burn the broken cable to create a more conductive path it seems accurate. It seems a bit foolhardy to me, but I don't do power electronics.

Oh, the name is certainly accurate, but it still doesn't sound like something you'd want to leave powered on and unattended... 

That said, it's massively overdesigned for finding a short in a heating cable. Using an isolation transformer with a lightbulb (and possibly a variac for more control) seems the most straightforward method.

Not necessarily - if the fault is an actual short between two wires (from, say, the insulation melting or breaking) then your approach would work, but if the fault is a partial insulation breakdown from a transient on the mains then you'd likely need to impress a similarly-high voltage across two otherwise open-circuited conductors to cause the partial breakdown to turn into a total one. That's the impression of the intended use I got from reading the OP's link and the product datasheet.

[sinB]

Wow, a lot of feedback, thank you - i'm excited!

To clearify around the heating cable values:
Normal twin conductor floor heating cables in scandinavia is rated at 10-20W/m.
Normal elements at 17w/m varies from 200W/264,5ohm up to 3300W/16ohm (in a non-broken state of course).

My understanding by now is that the Burn Down Transformer is a device that replaces a variac/high voltage insulation tester combo as described in this guide
https://www.warmlyyours.com/en-US/publications/TROUBLESHOOTING-KIT-INSTRUCTION-C.pdf.

The guide states that:

• If there is continuity between one or both of the cores and ground shield, you have a complete circuit present and the variac may be used.
• If there is no continuity between the core and ground of either wire, the core is likely broken or vaporized.
  The Hi-Pot device is then used to establish continuity between the core wires, or from the core wire to its respective ground shield by “fusing” the wire together sufficient enough so that it may carry the low voltage current of the Variac transformer.

Now i'm actually a little confused, but I guess the Burn Down Transformer, for this use, has to both supply sufficient power to heat the cable and to supply hight DC voltages to successfully fuse the wire together if it is entirely broken.

This understanding kind of correlates with the voltage steps on the Baur ATG2:
AC 60V/32A
AC 230V/8.5A
DC 800V/2A
DC 2600V/0.6A
...and so on


As some of you have suggested, I think an isolation transformer and current limiting is a good idea.
Now, high voltage DC. What do you think is the best way to integrate this?

I can think of a few ways to do this off the top of my head, but none of them are cheap. How much is one of these things, anyway*?

I contacted a dealer to get the price as of 2019.
The list price of the Baur ATG2 is EUR 13,500 before taxes.
What did you have in mind?

[Marco]

So what does the hi-pot device from the heating wire company cost? (Though I don't like the sound of the warning that it might break if run longer than 3 seconds, poorly protected device.)

These are quite dangerous devices by the way, that's a lot of power at high voltage ...

[MagicSmoker]

Yeah, I definitely agree with Marco that a hi-pot tester that can't be used for more than 3 seconds is a fail. For a quick-n-dirty solution I would use a Variac to supply either an isolation transformer or an old-school neon sign transformer (not the new style with built-in protection against ground faults and shorts). Use the neon sign transformer to encourage a partial insulation breakdown to finish the job, so to speak, and the isolation transformer to energize the heating cable at a reduced power level. Old-school NSTs have magnetic shunts in the transformer that turn them into current-limited power sources so are almost ideal for hi-pot testing as-is, while supplying an isolation transformer at less than its rated primary voltage will also limit the maximum current it can deliver, though additional resistance or reactance might be warranted for extra safety.

But as Marco already pointed out, there's nothing safe about the equipment needed to do this kind of testing, and while there are other means of finding faults in cables that are intrinsically safe - like time-domain reflectometry and RF injection/detection - they aren't accurate enough to pinpoint where to dig/rip up the floor/etc. like a hot spot on a thermal camera.

[digsys]

Thinking about it - a long time ago a friend had that problem with his heated concrete slab. The only way we could find the break, and luckily it was a ring loop, was to inject a good quality RF signal into the cable and trace out the circuit - then looked for obvious "dips" , "inconsistencies" - unfortunately there were none - then drilled "breaks" in the cable run - re-checked with the RF, and when that didn't help, injected 24VAC and tested a load at the break. A slow process, but worked out ok.
We found the crack

[james_s]

It's a valid technique. I once used a 9kV neon sign transformer to find a break in the winding of a HID ballast that had taken a whack. The high voltage instantly arced across the fault, clearly visible in subdued lighting and I was able to locate and repair it with minimal invasiveness.

[IanB]

First of all: this is an underfloor heating element designed for direct connection to the mains. I do not see the point of an isolation transformer here?

About how such things may apparently work:

Feeding a constant current source into a wire loop will produce local hot spots at points of high resistance.

If the buried heating element has a break and has gone open circuit, then a constant current source with a high compliance voltage can arc across the gap and produce a hot spot where the gap is. A thermal camera may then be used to locate the hot spot under the floor and let you know where to drill.

As was previously mentioned, a neon sign transformer may be a good option for doing this.

[IanB]

A heating cable is damaged somewhere in the floor, causing the fuse to trip. The cable is then hooked up to the Burn Down Transformer, and a thermal camera is used to observe the exact position where the heating cable gets cold. The heating cable can then be repaired at that exact point.

More likely the heating element has gone open circuit and the heating has stopped working.

Those are expensive, but looks relatively simple to build.

Whenever you have "expensive" and "simple" in the same sentence, it is probable that you are missing something important...

[james_s]

There are lots of expensive but simple things out there. The market is small so there's little incentive for anyone to jump in and drive the price down and most of the people using them lack the skills to build one themselves.

[sinB]

Thank you all for sharing!

So what does the hi-pot device from the heating wire company cost?
These are quite dangerous devices by the way, that's a lot of power at high voltage ...

I could not find the exact one from the heating wire company, but cheap chinese ones is sold for ~500EUR on EBAY.
I may look more into that later, cause.. what's the fun in that.
Dangerous devices indeed. I do not have any plans of playing with this without necessary safety measures.

Use the neon sign transformer to encourage a partial insulation breakdown to finish the job, so to speak, and the isolation transformer to energize the heating cable at a reduced power level. Old-school NSTs have magnetic shunts in the transformer that turn them into current-limited power sources so are almost ideal for hi-pot testing as-is, while supplying an isolation transformer at less than its rated primary voltage will also limit the maximum current it can deliver, though additional resistance or reactance might be warranted for extra safety.

That's a good idea, my first thought was a current limited microwave transformer, but an old school neon transformer sounds better.
Though my guts are telling me that HV DC is better as it is generating a more stable arc while "burning", thus making it more effective.
What's your thoughts on that?

Also, I'm not sure about how much current is needed to generate a sufficient "weld" for lower voltages to pass through, but I guess that will be one of the things I will test on some scrap heating cables when that time comes.

I completely agree with the additional resistance, I'll have to pick up some high voltage resistors.

The only way we could find the break, and luckily it was a ring loop, was to inject a good quality RF signal into the cable and trace out the circuit - then looked for obvious "dips" , "inconsistencies" - unfortunately there were none - then drilled "breaks" in the cable run - re-checked with the RF, and when that didn't help, injected 24VAC and tested a load at the break. A slow process, but worked out ok.

Sounds like a lot of fooling around, but interesting anyway, will look into it.

First of all: this is an underfloor heating element designed for direct connection to the mains. I do not see the point of an isolation transformer here?

One practical reason:
Since the heating cable often is buried on top of a reinforcing mesh which often is earthed (in scandinavia atleast), the current does not have a long way to go before tripping the RCD under these conditions.

If the buried heating element has a break and has gone open circuit, then a constant current source with a high compliance voltage can arc across the gap and produce a hot spot where the gap is. A thermal camera may then be used to locate the hot spot under the floor and let you know where to drill.

That sounds right, though I think a "permanent weld" (like the method in the guide I posted) is better when possible. That way I can supply lower voltages while locating the exact point without having to be using safety equipment. To me it sounds better, although it's a theoretical approach.

[IanB]

If the buried heating element has a break and has gone open circuit, then a constant current source with a high compliance voltage can arc across the gap and produce a hot spot where the gap is. A thermal camera may then be used to locate the hot spot under the floor and let you know where to drill.

That sounds right, though I think a "permanent weld" (like the method in the guide I posted) is better when possible. That way I can supply lower voltages while locating the exact point without having to be using safety equipment. To me it sounds better, although it's a theoretical approach.

If a "permanent weld" means to somehow repair a break in the heating element by melting it back together, then I have my doubts about this being effective or reliable. If you heat up the break in the wire to the point where the ends melt, then surface tension will tend to cause each end to ball up and recede from the gap, making the gap bigger.

IMHO the only reliable way to effect a repair is to dig up the floor and splice the broken ends back together with a proper crimped join.

[james_s]

Yeah I certainly wouldn't trust a non-invasive weld like that. That said, maybe it works, if the wire is buried into the floor then breaks would likely be hairline fractures from thermal cycles. I have "fixed" incandescent lamps before by applying power and tilting the thing around until the filament touches and welds. It was a favorite trick of mine when I was a kid.

[sinB]

If a "permanent weld" means to somehow repair a break in the heating element by melting it back together, then I have my doubts about this being effective or reliable. If you heat up the break in the wire to the point where the ends melt, then surface tension will tend to cause each end to ball up and recede from the gap, making the gap bigger.

IMHO the only reliable way to effect a repair is to dig up the floor and splice the broken ends back together with a proper crimped join.

Those were not so carefully chosen words from my part.

What I meant was to spark the gap with high voltage to make enough "permanent" conductivity for lower voltage to pass through before finding the exact point of the damage with the thermal camera while the heating cable is energized with lower voltage. And then of course chisel down at that point and repairing the heating cable.

I may be misunderstood, but the guide I linked to suggested so.

I guess it also depends on factors such as depth. If the cable is not too deep the high voltage arc should leave a thermal footprint alone. Then if the cable is deeper it would be best to have enough conductivity to supply the cable with lower voltage over a longer period of time and then see where the cable gets cold (and of course a brighter spot as a result of higher resistance at the end).

[james_s]

If the arc is AC it will make a buzzing noise, that on its own is likely going to be enough to get a pretty good idea of where the break is.

[coppercone2]

Calling something a "burn down transformer" sounds like the marketing fail of the century

Since they actually burn the broken cable to create a more conductive path it seems accurate. It seems a bit foolhardy to me, but I don't do power electronics.

That said, it's massively overdesigned for finding a short in a heating cable. Using an isolation transformer with a lightbulb (and possibly a variac for more control) seems the most straightforward method.

that belongs in a factory or with a trade professional. There is nothing overdesigned about it. You don't want to be wiring those things together incase you move a work bench etc. Think about this, at the point where you designed a chassis, you might as well put some features.

I hate this line of thinking because you get assembly line **atrocities**.

"it's just a....(looks like a bundle of organs)... its ok" (this machine ends up making/saving the company 10,000$ each time its used and may cost crazy amounts if a bad wire due to ad-hock assembly goes bad, also accumulates dust, dropped screws, etc)



and it actually can save a factory or workplace money because its no longer 'that bench we use once in a while but its really important you can't move it incase something comes in'. Slapped together things turn into area denial VX nerve gas in a factory. Paralysis and movement denial! Then you get cardboard box men living under the bench because its seldom used. Like bridge trolls.

And the bench is always built by 'that guy' that has untouchable stuff, so if that guy is not there the bench gets ARMED and you are not allowed within 10 feet because it might break and looks flimsy. Fully armed management battlegroup is deployed so it does not go off during cleaning or nearby work. Suddenly it makes its way into floor plans as a permanent structure.

wheras with a nice cube, people think 1) we can send it to be fixed, 2) its a damn box with a dial and 2 knobs, I can do this (and the cash flow continues, management is undisturbed, then men in black are not there to neuralyze the tour groups, etc). 

[Marco]

It's overdesigned for finding an existing short, it's purpose designed for creating a short.

[coppercone2]

The burn down transformer ATG 2 enables the user to change a high resistance fault into a low resistance fault.

Thus allowing an impulse reflection test set to be used for low resistance fault pre-location. The portable burn down transformer is enclosed in a sturdy 19 "steel housing. The well-proven power electronics allows individual control of output voltage and current. A vital step to a practical burn down technique.

It's still designed with RF reflectrometry in mind but still.

I would imagine proper use means hooking it up to a LF Rmeter and messing with it till the short gets to some particular impedance to minimize risk of fires.


What he is doing may get a fire department call if it turned out the wire got routed into a weird place.

[NiHaoMike]

What about connect a micropower AM transmitter to it and then use an AM radio to track where the signal suddenly attenuates? Or connect a flyback transformer and find the point that's operating as a spark gap transmitter.