New ( to me) pin plug failure

[JChenault]

I had an interesting failure today. I'm trying to understand the physics of how it occurred.

I was doing some work and needed more light. I hooked up a 400Watt scoop and stuck it near my work space. Note that the scoop was not attached to a metal pipe, it was effectively sitting in a free air space.

After 30 seconds or so of use, I heard a loud arcing noise, saw sparks, and got some flames coming out of the pin plug. After what seemed like 10 seconds ( but was probably more like two) the breaker tripped and all was quiet.

I examined the plug and here is what I found.




Note that the arcing burned the ground and hot pin more than half way through.

When I put a meter on the fixture and discovered a short between the hot leg and ground ( not sure where that short is as yet as I need to take the fixture apart. I suspect that it is somewhere in the whip). I also discovered that the neutral in the male plug was not tightly connected.

Now I have no issue with the fixture developing a dead short between ground and hot. It does not happen frequently, but it does happen. My surprise and concern is that I got arcing between the hot leg and the ground on the pin plug. I am also surprised that the breaker did not trip sooner.

Anyone had a pin plug do this before? Any theories why I got such a spark between the pins?

 

[Nelson]

Holy ####, never seen anything quite that interesting!

 

[JD]

Holy ####, never seen anything quite that interesting!

I have. Usually starts with some type surface burn that leaves carbon behind. Once enough of a trace is down, power will migrate between the pins, burning more of the surface and leaving more carbon. Just goes downhill from there. Not enough current will pass to trip the breaker until the final 4th of July happens.

The question is, how did it start? It can happen from moisture if the water has enough dissolved conductive solids. Somehow the connector got contaminated which started its destruction.

Not normal for heat to do this as the amount of heat needed to start the carbon cycle is very high and the rest of the connector looks ok.

(Of course something conductive (like sheet metal) could have slipped in there if the connector wasn't totally seated, but you would have seen a lot of other evidence at the scene of the crime!)

Somewhere, it's mate is waiting to self destruct as well.

 

[Nelson]

Interesting, so are you saying that this could have been caused only by carbon between the connector pins, not by a fixture fault?

 

[JD]

Interesting, so are you saying that this could have been caused only by carbon between the connector pins, not by a fixture fault?

Can't think of any fixture fault that would do that. The pins do not look discolored by heat. Something got on that connector that caused a part of the surface to conduct. The blowout happened at the surface of where the pins meet the connector housing.

Dead shorts on a fixture would simply trip the breaker. The 30 second delay is the clincher.

 

[JChenault]

Can't think of any fixture fault that would do that. The pins do not look discolored by heat. Something got on that connector that caused a part of the surface to conduct. The blowout happened at the surface of where the pins meet the connector housing.

Dead shorts on a fixture would simply trip the breaker. The 30 second delay is the clincher.

I will re-test the wiring with the pin plug disconnected. Could be the short I was detecting was due to the carbon build up.
The pins were not notably discollored before the event.
Just to be clear - we did not get the arcing for 30 seconds. The fixture worked successfully for about 30 seconds.

 

[IEGPR]

Carbon is nasty nasty stuff peeps !! JChenault hit this one on the head !! I have seen it do big damege. This is a good Thread for everyone to read and learn. If you ever see any carbon on anything clean it off !!!!!

 

[JD]

I will re-test the wiring with the pin plug disconnected. Could be the short I was detecting was due to the carbon build up.
The pins were not notably discollored before the event.
Just to be clear - we did not get the arcing for 30 seconds. The fixture worked successfully for about 30 seconds.

Probably a nice mix of metal in with that carbon now. May not read as a dead short when cold, but you'll get some resistance. Would also hunt down it's mate as that female may be a ticking bomb. If nothing else, it needs to be inspected.

Also, the act of opening it (the male) or moving the pins a bit may make it read ok.

 

[JChenault]

I think JD has this figured out.

I removed the whip from the pin plug, and found that there was no short in the fixture. I then took a ( crappy ) multi-meter and measured about 20 - 40 ohms between the ground and hot pin.

So if I am doing my math right ( assuming it is really 30 ohms when cold) that would imply a starting current of about 4 amps. If the resistance goes down as it heats up - or if moving the pins changed the numbers, then what we saw seems to match JD's scenario.

JD - I did a visual examination of the female. No arcing visible. I will clean any carbon. Anything else you can think of?

 

[mstaylor]

As noted, find the female it was plugged into, but also if you have since plugged another male into the suspect female change it too. It will just keep transferring back forth and kill many plugs.

 

[venuetech]

Probably a nice mix of metal in with that carbon now.

In photo #1 you can see the copper deposits between the pins.

 

[Beans45601]

I never realized that carbon was conductive at such low voltages. Back in my Tesla Coil days, it was really easy to get carbon to conduct at 15kv (or from the 1.2 million volts from the coil its self), but I thought that only applied to hgiher voltages. Hm.

 

[Nelson]

I suppose if you think about it, carbon is the "active ingredient" in old telephone carbon transmitters, and they work at very low voltages.

 

[MPowers]

Thomas Edison's light bulb had a Carbon filiment (carbonized bamboo)

 

[JD]

I never realized that carbon was conductive at such low voltages. Back in my Tesla Coil days, it was really easy to get carbon to conduct at 15kv (or from the 1.2 million volts from the coil its self), but I thought that only applied to hgiher voltages. Hm.

That's why I suspect it got something else on it that started the whole thing. Plastic is a hydro-carbon, so as it burns it rapidly adds more carbon to the trace. Carbon also drops in resistance as it heats up. A contained burn like that can also ionize the air allowing a very large flow of current. (Think arc lamp)

Surprised the other connector was not as damaged. I would replace it anyhow.

 

[ship]

Main question for me, breaker didn't trip in getting this far why?

Fixture problem before the lamp to be sure in short, resistance in the plug in melting away.

Some stuff to investigate and inspect but why if this much damage was done didn't the breaker trip? Should have.

I also have these breakered followed in the chain GFCI test circuit outlets for any gear tested. Everything that goes out in prepping gets plugged into these at least in the 120v world. Test outlets where you flip a switch in not hot patching anything live or touching it.

For 2Kw loads especially ACL lamps, there is a 2Kw dimmer that will induce load slowly but on the same protection.

This failing in getting that bad.... a number of things I would be looking at in how it got this far including loose screw terminals that could melt away with time on a connection.

 

[FMEng]

Main question for me, breaker didn't trip in getting this far why?

Fixture problem before the lamp to be sure in short, resistance in the plug in melting away.

Some stuff to investigate and inspect but why if this much damage was done didn't the breaker trip? Should have.

I also have these breakered followed in the chain GFCI test circuit outlets for any gear tested. Everything that goes out in prepping gets plugged into these at least in the 120v world. Test outlets where you flip a switch in not hot patching anything live or touching it.

For 2Kw loads especially ACL lamps, there is a 2Kw dimmer that will induce load slowly but on the same protection.

This failing in getting that bad.... a number of things I would be looking at in how it got this far including loose screw terminals that could melt away with time on a connection.

The breaker didn't trip because the arc didn't draw enough current to trip it. In a small area like that, quite a bit of heat can be generated without much current because it is concentrated.

Using GFCIs on your test bench is a great idea for safety. In this instance, a GFCI would have sensed the difference in current between the hot and neutral and would have tripped immediately. The current difference occurs because some current was flowing through the ground wire instead of the neutral, where it belongs The downside of a GFCI tripping is that troubleshooting would be more difficult as there would not be the telltale damage. Burned parts usually make troubleshooting easy.