Beyond frustrated

[pas63]

I'm not a marine electrician, but I've done my share of wiring boats and houses. I have an issue that makes absolutely no sense. I have an LED light I'm connecting to factory wiring.

Wire up the light, nothing. I check the voltage, 12.7V, more than enough for an LED light. Polarity is correct, so I temporarily connect it straight to supply voltage and get light, so the light is good. Hook up the volt meter, 12.7V, lower than the panel (14.6v), but it is probably a 20-30' run over 14 ga wire. Turn the breaker off, check the VM, .1V. Turn the breaker back on and back to 12.7V. Reconnect the light, nothing. Where it gets really bizarre is I can short the pos and neg leads feeding the light and no firework or popped breaker which tells me I don't have voltage but the damn VM does show voltage. I assume it accurate because I have a panel meter and when I check the voltage at the panel with the volt meter it's within .1V of the panel meter.

None of this makes any sense. Tracing the wire is all but impossible sine it's in a bundle, routes through some inaccessible areas. I hate to run more wire. Tempted to just tie it into the courtesy light and be done with it. I'm done screwing with it tonight. Maybe clearer heads will prevail tomorrow.

[capn_billl]

A multimeter will show the correct voltage through an inch of rust, a popped fuse, and your fingers.

It only takes a few micro amps to show voltage.

an Led light takes a couple amps.

You have a rusted terminal strip, or even a rusted fuse casing.

a broken wire inside the insulation will show voltaeg on a meter, but not carry enough current to light a light.

[Reefmagnet]

Originally Posted by pas63

I'm not a marine electrician, but I've done my share of wiring boats and houses. I have an issue that makes absolutely no sense. I have an LED light I'm connecting to factory wiring.

Wire up the light, nothing. I check the voltage, 12.7V, more than enough for an LED light. Polarity is correct, so I temporarily connect it straight to supply voltage and get light, so the light is good. Hook up the volt meter, 12.7V, lower than the panel (14.6v), but it is probably a 20-30' run over 14 ga wire. Turn the breaker off, check the VM, .1V. Turn the breaker back on and back to 12.7V. Reconnect the light, nothing. Where it gets really bizarre is I can short the pos and neg leads feeding the light and no firework or popped breaker which tells me I don't have voltage but the damn VM does show voltage. I assume it accurate because I have a panel meter and when I check the voltage at the panel with the volt meter it's within .1V of the panel meter.

None of this makes any sense. Tracing the wire is all but impossible sine it's in a bundle, routes through some inaccessible areas. I hate to run more wire. Tempted to just tie it into the courtesy light and be done with it. I'm done screwing with it tonight. Maybe clearer heads will prevail tomorrow.

There's a high resistance in the circuit somewhere. Put a load on the circuit with a light bulb or similar and check the DVM readings again.

[pas63]

'an Led light takes a couple amps." I disagree, I've checked the draw of others on the boat when working out my power load and I might see 1/2amp per light

"You have a rusted terminal strip, or even a rusted fuse casing" No, I just checked and re-connected the panel side, no rust or corrosion anywhere

"a broken wire inside the insulation will show voltaeg on a meter, but not carry enough current to light a light." That could be a possibility, leads me more to connect it to the courtesy light.

"There's a high resistance in the circuit somewhere. Put a load on the circuit with a light bulb or similar and check the DVM readings again."

Hadn't thought about the resistance angle, but I did pull a bulb from another light and still nothing. I shorted the panel ends (disconnected) and then checked the resistance, I was seeing about 8 ohms which doesn't sound like for 40-60' (20-30' of wire per leg"

[Chotu]

The previous answers have it. There is a problem with the wire obviously. Or the connections on the wire, including fuses.

Check all the connections, clean them, check fuses and clean those connections, circuit breakers, then go ahead and run a new wire if nothing else works.

[pas63]

"There is a problem with the wire obviously"
I have to agree with this,

"Or the connections on the wire"
Both ends are freshly terminated and no corrosion. Since these were ran through the bilge (WHY?) I guess I have assume a pinhole that got wet, or worse, maybe a poorly spliced connection somewhere along the run from panel to light. UGGG

"including fuses"
No fuses, goes to a breaker on the panel.

[Reefmagnet]

Originally Posted by pas63

"There is a problem with the wire obviously"
I have to agree with this,

"Or the connections on the wire"
Both ends are freshly terminated and no corrosion. Since these were ran through the bilge (WHY?) I guess I have assume a pinhole that got wet, or worse, maybe a poorly spliced connection somewhere along the run from panel to light. UGGG

"including fuses"
No fuses, goes to a breaker on the panel.

Here's what happened to me on my boat once, and with an LED light, too.


I'd pulled out all the old wiring for a complete rewire, but decided to recycle the odd bit of the original wiring that presented in good condition. One circuit ran to a light in the forecabin. Hooked the light up and no go. The wire was original, but the wire at each end was as shiny as the day it was made.


The problem was someone had put a small, barely discernable knife nick in the insulation of that piece of figure 8 wire when (most likely) the boat was built, and that nick was buried deep in the innards of the boat for 30 years. This didn't stop electrolysis occurring between the two - probably due to nothing more than humidity - over the years and one wire of the pair had turned into a corroded mess at the nick. The effect was an open circuit that conducted just enough for a digital multimedia with it's megaohm+ input impedance to read a voltage with a slight drop under no loads.


It will be a poor circuit connection somewhere. The usual procedure is to start in the middle in order to isolate the location progressively at the connections. If there are no connections, you might have a situation similar to my one described above.

[fourlyons]

Your digital multimeter, unlike an old school analog voltmeter, pulls nearly no current through the circuit. So it is easily "fooled" by a high resistance connection. Hook a 12 volt incandescent bulb or small electric motor to the circuit then read the voltage, it will be too low to run either.

[chris mac]

I built a 12v plug(cigarette lighter style) with 2 long wires coming off, positive and negative. I use this as a control to test wires.
Check the led off this setup to confirm it's working. Then try with your original positive and the control negative. And of course the other way round .
This at least lets you know which wire has the issue.

[jamhass]

As an aside, it is entirely possible for a 12v incandescent light to work fine, but for an LED in the same location to fail.

2 reasons this might happen (maybe others):

1. Polarity is reversed. Incandescents don't care. LEDs do.

2. Contact resistance. The higher current draw of the incandescents can "burn through" the contact contamination. The very low current of the LED cannot.

[skipmac]

Originally Posted by fourlyons

Your digital multimeter, unlike an old school analog voltmeter, pulls nearly no current through the circuit. So it is easily "fooled" by a high resistance connection. Hook a 12 volt incandescent bulb or small electric motor to the circuit then read the voltage, it will be too low to run either.

This is the correct way to confirm you have a high resistance (corroion or a loose connection are the most common causes) somewhere in the wire to the LED. Have had the same thing happen. Read voltage at the wire and it's fine. Connect a light bulb and nothing.

Finally tested the voltage WITH the light bulb connected and it dropped to about 1V.

[JBP]

Originally Posted by skipmac

This is the correct way to confirm you have a high resistance ...
Finally tested the voltage WITH the light bulb connected and it dropped to about 1V.

Leave a bulb in circuit and on. Use your voltmeter to find a 12V drop in which of the legs which has a problem.

[fourlyons]

Originally Posted by fourlyons

Your digital multimeter, unlike an old school analog voltmeter, pulls nearly no current through the circuit. So it is easily "fooled" by a high resistance connection. Hook a 12 volt incandescent bulb or small electric motor to the circuit then read the voltage, it will be too low to run either.

I should have elaborated on this. Today's digital multimeters have high internal impedance (10 megaohms) so that they can be used on sensitive computer circuits that would be damaged by the current an analog voltmeter or incandescent test light would pull. But for the same reason they don't pull enough current to test a circuits ability to carry more than a miniscule amount of current, unless you load the circuit then at the same time test the voltage.

[Lake-Effect]

Originally Posted by jamhass

As an aside, it is entirely possible for a 12v incandescent light to work fine, but for an LED in the same location to fail.

2 reasons this might happen (maybe others):

1. Polarity is reversed. Incandescents don't care. LEDs do.

2. Contact resistance. The higher current draw of the incandescents can "burn through" the contact contamination. The very low current of the LED cannot.

#1 - agree. #2 - not likely. Think about it: if there's not enough current available to light a LED, there's for sure not enough current to "burn through" a connection issue. It is sometimes possible to "break through" contact resistance by physical force - eg inserting, twisting and wiggling a bulb into a dirty socket, which dislodges dirt and oxidation.

[Mike Burch]

8 ohms resistance is much too high. It sounds very much like a cable fault. In my experience, it's quicker and easier to pull a new wire through (using the old one as a puller) than to try and track down a cable fault—especially as the fault is always in the least accessible place!