IEC lead,Bang Flames Smoke

[wasyoungonce]

Posted for interest sake....moving computer monitor, it went off, I checked IEC lead...bang, fizzle, flames smoke out lead around 4 inched from connector.  Ok I've seen things go bang before, this was a little unnerving I had to shut off power I couldn't yank the lead out.

Pics.....it appears to be a no names IEC lead "made in Wang Yip"...what ever that means,  Supposed to be .75mm^2 conductor (which are ~1mm dia).  I cut it apart...conductor is barely .6mm dia which works out to be .282mm^2.  Even then it probably shouldn't have gone bang.  I've seen this before in oscillating fans, the wire works then snaps then similar pop crackle smoke.

So putting it out there for others, yes stoopid me using no names IEC...probably had it for years.  Its not one of those Aluminum conductor IECs.

[Ian.M]

You don't *know* its not CCA (copper clad aluminium) or CCS (.. .. steel) until you do the flame test for CCA and magnet test for CCS.

[wasyoungonce]

Ian....you are totally correct it is CCA...falls apart under flame.   I should have known that.  Sigh!

Now I need to burn test all my IECs?

Might look at getting IEC plugs etc wiring my own...except even those come from China.  My older Australian IECs are ok but I have a few I'm now suspect.

[Ian.M]

No. You can non-destructively resistance test them.  It doesn't matter whether its got broken, too thin, or CCA/CCS wires - a resistance test will catch all of them unless the CCA has 160% of the nominal CSA.

Make up a probe with two tinned bronze leaf springs to go into one contact of the IEC connector so you can do four wire measurement to eliminate contact resistance, and if the resistance is more than 20% over that expected for that length of pure copper of the rated CSA, its either worn out or a chinesium P.o.S. that needs to be condemned and cut up before disposal.   

[wasyoungonce]

Many thanks IAN....I'm on to it tomorrow. 

[Ian.M]

If you don't have a four wire capable bench ohmmeter, build a circuit to pass 100mA through the cable (e.g. carefully trimmed and well heatsinked LM317 current source) via the forcing contacts, and  configure an OPAMP as a x10 differential amplifier for the sense contacts for direct readout of milliohms as millivolts on a DMM.   Make sure the OPAMP input circuit protects against the full forcing voltage with the load O/C, and inductive spikes.

You are looking for 16.8 milliohms per meter for 1mm² copper wire. 

It *may* be helpful to go up to a 1A test current for heavier cables to get ten times the output voltage, *if* you can avoid excessive heating of the probe to IEC contact.

[Halcyon]

I keep a box full of old IEC power leads for this reason, even though the pins aren't insulated (insulated pins on 240v plugs have been mandatory in Australia since the early 2000's).

Without chopping up my leads, the older ones seem to be of far better quality and in many cases heavier than the crap that comes with some products these days.

[vk6zgo]

I keep a box full of old IEC power leads for this reason, even though the pins aren't insulated (insulated pins on 240v plugs have been mandatory in Australia since the early 2000's).

Without chopping up my leads, the older ones seem to be of far better quality and in many cases heavier than the crap that comes with some products these days.

Decently rated IEC cables were commonly used on electric kettles, & were still quite useable when the kettle died, so, I have quite a few.
Unfortunately, since all the kettles went to the "cordless" style, that avenue of supply has dried up.

The "cordless" things are, in my opinion, dangerous, as the base has raw mains easily accessed via the quite open construction of the necessary connecting socket.
This can also fill with water if there is a leak, which is not uncommon with Chinese kettles (& they all are Chinese, no matter what fancy brand they have).

It is amazing that the licensing authorities get all worked up about putting a bit of insulation on the power plug pins, & let these nightmares be generally available.

[MrMobodies]

I am so paranoid when I order network cables that I tear one apart and test it under a flame whilst taking pictures incase there is a problem.

I came across Excel counterfeits from when I ordered a batch of CAT5e cables some years ago.
I saw packet losses on the switch statistics from a batch of those to tidy up an old rack.
Got suspicious and tested one under the flame and it flopped.

Why would they counterfeit such an old low spec of cable when they could make more money flogging cat 6/7a cables.
Looking at the detail of the cable I suspect the factories that make the genuine stuff also trade CCA variants out the back door to so called "middle men".

For any mains cable I get I test them for their rated capacity and toggle them about just incase they cause problems down the line as I might think that the appliance is at fault.

[Cyberdragon]

Just separate the indevidual strands and try to bend them slightly, if they resist or spring at all they are bad (courtesy of Bigclivedotcom). Not sure if it works on larger solid core strands though.

[Berni]

Now that's a dangerous cable, didn't think it could just burst into flames mid way.


I do have a non destructive test for IEC cables that look suspicious. I simply put a short in the end of the cable and hook up the other end to my high current PSU. First i put 16A trough the cable (That's the max current from a standard outlet here) and let it sit there for a few minutes to see how much it heats up, if it gets just slightly warm to the touch its fine, if its actually getting pleasantly warm or even hot then something is wrong(Too thin wires, or aluminum wires). You can use a known good IEC cable to see how little heat should come from the cable. Then i tend to torture it by cranking it up to past 25A. Its normal for it to become a nice hand warmer, but what i look for is if the connectors on the ends are getting any warmer than the cable, if they are becoming hot this would point to shitty connectors or bad cable joint on it. In this test i also move the shorting jumper to earth to verify the earth conductor is of the same size and has good connection.

Seeing how shitty cables can be on the internet made me worried about such a cable starting a fire in my house so i started testing any that look too chinese for comfort.

Just recently i found a unknown origin USB-B cable that turned out to have the pinout reversed on one of the ends so that 5V and GND become swapped on the other side, Luckily it has not blown up anything. But in terms of wire size yeah the cable would easily loose 1V along the length for a 500mA load.

[amyk]

Repeated movement fatigued the conductor until it broke and the ends started arcing.

I suspect the reason why we don't hear more about these things causing major fires is because in the majority of cases they aren't used at anywhere near 16A. How many average users have equipment with an IEC socket that draws more than 1A continuously? That's 120W in 120V countries, and 240W in 240V countries.

[coppercone2]

somewhere in my lab/home modernization strategy is to buy quality heavy gauge shielded IEC cables for all equipment.

[wasyoungonce]

Trying to ohm compare the IEC cables I have.  I have quite a few old and new.  The cable that blew wasn't bent or worked in any way.  This has got me paranoid.

[wasyoungonce]

Ok quick incomplete update.

Been testing some IEC cables I have, many to do.  So far I found good copper IECs and some CCA IEC cables measure resistance wise reasonably closely to calculated predicted values.
R=Rho(L/A) in ohms, using L = std copper value resistivity, length m, A = conductor area m^2 Of course you have to know the conductor size which is not always obtainable.

But, I found throwing the cables on a Current limited PSU in CC mode I could input a set voltage and see the current drawn over its length.  I used IEC sockets to feed each wire back thru each conductor so it was 3 times the length (just to make values greater to see).  Anyway CCA copper drew double to 3 times the calculated expected value in amps (using 100-300mA) in each case.

So I have more to do but this looks a promising method.

[Berni]

Nice to hear the difference is clearly visible. Tho you can probably do the same thing with just a multimeter if you got one that can measure down to 0.01 Ohm resolution (Most handheld ones don't but quite a few bench ones do).

The reason i test the cables by overloading the cables to twice the rated current using a big PSU is to also find bad connections. If the joint at the connector is weak it might just make the cable look like it has 20% more resistance, but this is 20% extra power loss happens instead of along the cables whole length is happening inside a few millimeter large spot. This causes that point to get much hotter than everything else and could eventually lead to failure. The failure could be the cord getting too hot when used with a large load or as the cable is bent the bad connection could get even worse, turning into 1 Ohm in that point. Send 10 A trough 1 Ohm and you get a 100W heating element inside your IEC plug. As you can tell the IEC plug wont take that for long and will melt or catch on fire.

Now that i have a 300A 3.5kW PSU im thinking i might build a jig that tests cables by applying all that current full wack trough cheap power cables for half a second or so. This would not only show up bad connections as hotspots but could even blow apart such bad connections. Its better that it gets blown up on the test jig rather than in the wild. A fault in equipment could short L and PE sending a massive fault current trough PE and possibly blowing it open, when the beakers are flipped back up this would make the metal case of said equipment live and not trip any breakers.

[MrMobodies]

Never had that problem like that before but I did threw away some cables in a job lot of used IEC cables that I didn't bother to test as they were too thin and when I broke one up it was aluminium not copper.

I got a pat tester and I check every cable I come across for that and for it's rated capacity.

[IanMacdonald]

I've often wondered when upgrading computers if it's better to leave the old IEC leads or fit the new ones. I think I have my answer. 

[MrMobodies]

I don't trust some of the new IEC cables especially the ones especially when they are without the logos on or the styling of with the weird Chinese names on it.

The thinness of the cable and the imperfections on the moulding on the plugs makes me suspicious where I may open them.

Some the older ones look better built and I keep them after bending them a bit during testing to try and find weaknesses.

I had two atx power supplies that gradually blew (Got them RMA'd) about ten years ago and a third one didn't turn on until I swapped the cable over.
So now I take it seriously.

[MrMobodies]

Seen these figure of eight cords before?

came across many over the years and binned them.
I see advice to change the plug on it but it is not worth it.
Didn't know they also had CCA in it.