strange problem connecting light to dimmer

[John Hunter]

I am trying to change the light fixture on a ceiling light controlled by a dimmer that has an on/off push control, and a rotary dimmer control. I've changed this light twice before with no problems (dimming worked as expected), but this 3rd time when I threw the breaker back on I got no light. I'ver verified that the light works by connecting it to another power source.

For the record, on this 3rd try to change the light, I was working with the breaker on and and one point caused a brief short that caused some sparks to fly and the breaker to trip. Since then I've always turned off the power when working. Also, the original wiring coming through the ceiling is very old and brittle with lots of exposed parts. That's how I caused the short; by moving the wires around some exposed wires must have brushed up against one another.

Tonight I tried to diagnose what was wrong, pulling out my trusty fluke voltmeter. This surprised me: when the power is on, the minimum voltage across the two wires I connect the light to is 60V. I get this when the dimmer switch is off, or on at its minimum position. When the dimmer switch is on and at its maximum position, I get 120V. I was expecting to measure something close to 0V in the off position.

Now more wierdness. When I connected the light to the ceiling wires (white to white, black to black, same ones that have worked for me twice before without incident), I got nothing (as described above). At first I assumed it was just a poor connection (I'm using wiring nuts) but undid and redid everything and got the same result. When I measure the voltage across the wires the light is connected to with the light connected I get 0V, whether or not the switch is on or off, minimal or maximal.

I have repeated all of these measurements multiple times (ie, turning off the power, unwiring the light, powering back up, measuring, turning off the power rewiring the light, powering back up, measuring again). Always the same result: 60 to 120 V min to max when the light is not connected, 0 volts when it is.

Can anyone explain to me what is going on, or what I should do to diagnose this problem?

[indago]

Turn off the circuit; take the dimmer switch out of the wall and disconnect it from the circuit and wire nut the two wires together and connect the fixture to the wires in the ceiling. Then turn the circuit back on again and see if the light works. I am thinking that the dimmer has failed and you need a new one.

[Jeff Wisnia]

You have probably fried the dimmer.

If your Fluke is a digital meter it probably has an extremely high input impedance and any voltages you measured were obviously present across the meter leads, but might not have more than a miniscule amount of current capability, hence dropped to zero when loaded with a lamp bulb.

Try what indago said to try.

Jeff

[Chris Lewis]

According to Jeff Wisnia :

Yup, probably.

If he pulled enough current thru the dimmer to blow the breaker, the dimmer triac almost definately blowed up. Triacs, being semiconductors, are relatively fragile that way.

[Anthony Diodati]

[John Hunter]

That did it, thanks. New dimmer and all is well.

John Hunter

[Chris Lewis]

According to Anthony Diodati :

It _could_, but dimmer triacs are MUCH more sensitive to high overloads than house wiring. Semi-conductors (even power devices) are excellent "very very fast blow" fuses on high overloads. They can blow in microseconds, whereas breakers take considerably longer even in the highest of overloads.

If you get a short on a dimmer controlled device, even a momentary one, chances are that the dimmer is dead even if the breaker doesn't trip. Indeed, I'm a little surprised that the breaker _did_ trip. Somehow that triac held the current high long enough for the breaker to notice...

[Likely the junction in the triac vaporized and maintained conductivity just long enough for the breaker to start moving. Then the conductivity failed when the vapors condensed.]

When I saw "nonfunctional", "short circuit" and "dimmer", "fried dimmer" is the conclusion I immediately jumped to. The breaker tripping was essentially irrelevant in determining the most likely cause.

The idea course of action is to test the dimmer ("Do I have good voltage on the input and not the output?"), but, without some knowledge and proper instrumentation (a high impedance voltmeter isn't it), swapping the dimmer (or simply taking it out of the circuit temporarily) is probably easiest for most people.