Troubleshooting Electrical Short

[akreel]

Im joking for those that dont know,,

I thought it was a pretty safe assumption that you were joking. Pierre's suggestion just triggered my "what if" response for some reason.

 

[DonsDaMan]

Terry,

What you're failing to take into consideration is that fuses blow because of excessive heat, not excessive current.

Once that concept is realized, it is easy to see that at T = epsilon (some arbitrarily small number), you haven't generated enough heat to cause the element to separate.

At least, that's my take on it.

 

[bwheat]

I have done the light bulb trick... Put a 100 watt light bulb in series with the fuse. Had a 4A fuse feeding that hit a terminal block feeding lots of limit switches etc. Could not find the short with an ohm meter. Pulled all of the power wires off the strip and with the light bulb in the circuit put one wire on at a time. When the light bulb got real bright we found our culprit. It was water in a level switch.

 

[icky812]

This thread seems to have been wandering a bit, so to move things along a tangent, here's a story:

<Props feet up on desk>

A few years ago I had a customer call in and state that they had blown up a power supply, a couple of operator interfaces, and a signal conditioner.

I sent them repair parts and told the tech what he should look for to find the problem.

Next afternoon, customer calls again. He stated he couldn't find anything wrong when he did the troubleshooting that i had told him to do. So he installed the new equipment and naturally enough it blew up. So again I send him the required parts.

Yep, you guessed it. Next day same thing. This time the tech's boss had come in from out of town to see why this guy was spending so much money on the same stuff. Boss goes over system and can't find anything. Again, parts are installed and a few hours later, the same parts are blown up.

By this time they have lost four days of production and spent several thousands of dollars trying to fix the machine. Boss decides it is time to involve the manufacturer (me). So I fly 900 miles to the site on a late night flight.

Next morning I am on site with new parts and my handy-dandy troubleshooting skillset. I open the cabinet up and find all of the bad stuff. I figure that I'll be a bit smarter and disconnect it and bypass the required components in logic. I added a crowbar circuit to the power supply so it wouldn't blow up again (customer had already bypassed the fuse of course.)

Lo and behold the system ran fine. Or so I thought anyway. Me and the tech and his boss are standing around discussing the problem when, Blammo, the system dies. No one within twenty feet of it.

So I get out my tools and start checking by pulling wires until the crowbar circuit let the power supply start suppling voltage again. Hmmm... a 4-20 ma level transducer. What the hay has happened here and why didn't it smoke the analog card in the plc if it is pulling too much current?

I need to add here that 90% of the system is under a covered roof outside. For some reason the cover didn't quite extend to include the whole system like it was supposed to.

I walk over to the level control (which is outside the cover) and pop the top. The sun was shining brightly and it was a really nice morning. As I am looking down at the level control, my shadow passed over it. As i am standing there looking at it, I notice the wire supplying the power to the control starting to bend back up straight from a sort of laid over position. This really hits me as odd, so I move over a bit and sure enough when the sun hit it for a few minutes it srarted laying over again. Too weird, I think. Sure enough I looked inside the cap and there is the evidence. A black electrical burn on the inside of the cover. Seems that the wire insulation sould get soft when heated by the sun and lay over against the cover. The machine vibration would make it move and it had eventually wore the insulation off of the wire causing the short.

The reason that they had trouble finding it was that the sun only hit that section of the machine for about 2 hours at a certain time during the day. Any other time, the level control was in shade and the wire was not laid over causing the short.


<kicks feet off of desk and goes back to work>

 

[rsdoran]

What you're failing to take into consideration is that fuses blow because of excessive heat, not excessive current.

That is not exactly correct. Fuses are designed using 2 factors, current and AIC (ampere interruting capacity). Excessive current passing thru the fuse creates heat which if maintained will eventually burn out/melt the fuse...how long depends on type/class of fuse.

The whole concept of fuses and circuit breakers is based on current and time, the idea is to prevent the circuit from overheating.

As for the t=o thing I am not sure I understood where Terry was going.

Since we are leaving out XC and XL I assume we are speaking strictly of a DC ckt.
1. If its a purely resistive load then current will be (more or less) instantaneous...I= V/R. There is no time constants, transients etc etc. Technically this applies to AC as well.


Fast acting (Normal) fuses are used primarily for resistive load OR to protect sensitive devices from the heat caused by excessive current, some may be current limiting. These provide overload protection at 135% and 200% of rated value. To use these with motors they need to be rated 200-300% over load current but now they only provide short ckt protection and not overload protection.

Time delay or Slow blow fuses are designed to allow temporary overloads as high as 500% for 10-12 seconds. These fuses can be selected closer to a ckts actual operating current which provides overload and short ckt protection.

All fuses have an inverse time characteristic...ie as the current increases the opening time decreases.

In the US NEC an UL governs classification and actions for fuses.

 

[plc_padawan]

Good story icky, that would be a tough one to find, for anybody.

 

[elevmike]

AHHH Haaaa!

So there! AD, can, will and does produce a tech support body on site!!! Good work Icky!

 

[icky812]

Actually, I was with a previous company when this occured and the customer paid through the nose to get me down there the same day. It was far cheaper than the fines that they would have had to pay if the system had stayed down longer.

There are very few problems that I can't solve over the phone, as long as the person on the other end gives me all the info I request.