Dimmer controlled light never gets more than 95 volts. Is that bad?

January 18, 2018 12:05 AM

Working on my basement wiring, mostly older Romex. Light in center of living room is controlled by a Lutron dimmer. Even when that switch is on and at full power, it never reads more than 95 volts on an ammeter. Is that just dimmer's being dimmers? Or a signal of a more pressing problem. The dimmer switch literally has no screws for the neutral (white) wires at all. I plan to replace it with a binary switch, but wonder whether the low voltage is a sign of a deeper problem.
posted by msalt to Home & Garden (21 answers total)

Response by poster: To be clear, the ammeter reading is from leads clipped to the wires powering the light, after I disconnected them from the cheap ceiling light fixture.
posted by msalt at 12:06 AM on January 18, 2018

Just checking - do you have the multimeter set to AC voltage, and the probes in the appropriate sockets of the multimeter? Also, do you have any information about the age (or better still, type) of the dimmer? There have been several dimmer technologies over the years, and they can behave differently with the load (the bulb) removed.
posted by pipeski at 3:20 AM on January 18, 2018 [1 favorite]

What's the voltage at (before) the dimmer? I'd also check the voltage at other circuits in the house.
posted by exogenous at 5:43 AM on January 18, 2018 [1 favorite]

I'd try checking the voltage WITH the ceiling fixture attached. Just measure the voltage across the load (the fixture). Solid state dimmers are weird. It may need a load to properly do its thing.
posted by BillMcMurdo at 6:02 AM on January 18, 2018

Yeah, check a few outlets in the vicinity to verify you’re getting 120V. Also, turn the power off, take the cover off the switch plate, turn the power back on, and measure the voltage (a) across the dimmer switch contacts, and (b) between the incoming hot wire and neutral or ground. That will tell you what’s what.

A dimmer should let you get 120V thru.
posted by Huffy Puffy at 6:41 AM on January 18, 2018 [1 favorite]

Just to clarify, “turn the power off” means at the breaker, not via the dimmer switch.
posted by Huffy Puffy at 7:00 AM on January 18, 2018 [1 favorite]

Many Lutron dimmers have both a high side and low side trim function. The low side allows you to set the point to stop dimming to avoid CFL or LED flickering. The high side allows you to balance light levels in case multiple dimmers in a room drive different brightness lights. You can look up the programming guide for the particular dimmer you have to see if that's available and how to use it.
posted by saeculorum at 7:14 AM on January 18, 2018 [1 favorite]

Measuring mains voltage with a multimeter and no load very rarely yields sensible results. It's just too easy for a bit of stray capacitance to couple enough 60Hz to cause readings on what are, by design, incredibly high-impedance inputs.

Any two-terminal dimmer is going to need a load attached between its output terminal and neutral in order to work at all; otherwise there can be no current flowing through it, and therefore no voltage across any of its internal components. A multimeter on a highish voltage range is as near to no load as makes no difference. The 95V reading you're getting will be mostly meaningless.
posted by flabdablet at 9:03 AM on January 18, 2018

95 volts on an ammeter

You should be using a voltmeter. An ammeter measures current flow (amps).

But, depending on the quality of your voltmeter, you might get some funny measurements anyway. House electric service is 120v RMS (root mean squared). Some cheap digital multimeters don't do a great job of measuring AC voltages because of sampling issues, for instance.

Further complicating things - many dimmers work by clipping the sine wave to limit the forward voltage. This can have lots of effects - the first of which is that the switch never passes the full 120VRMS, the waveform is always deformed. The second is that it introduces harmonics, which while (mostly) harmless to lightbulbs, does all sorts of wacky shit to digitial equipment (LED lights, digital mulitimeters, etc.).

You need to measure the voltage at the switch. If you only have 95 at the switch, then you have an issue somewhere. If its 120 at the switch, then I'd replace the dimmer. But honestly, I'd replace it anyway.
posted by Pogo_Fuzzybutt at 9:05 AM on January 18, 2018 [1 favorite]

Response by poster: Sorry, I misspoke, I'm using a multimeter (Amprobe Am-420, on 200 voltmeter setting.) The wires at other lights I've tested on this circuit have all been 120V +/- 2. And yes, I do plan to get rid of the dimmer.

My concern was, if I do that, am I missing a deeper problem by blaming it on the dimmer? As I trace this circuit along, and increase the amount of sheetrock patching I'll have to do, I've found a "junction" without a box -- just four wires, capped, folded and shove up into the wood rafters. Yeesh!
posted by msalt at 2:23 PM on January 18, 2018

Response by poster: Pipeski -- it's a Lutron TG-600P (Toggler) dimmer for incandescents. I really don't know how old it is, but they seem to sell very similar models still today (TG-600PH, etc.)
posted by msalt at 2:38 PM on January 18, 2018

I think an important step is to check the voltage when there's a bulb lit, and also check the voltage coming in to the dimmer.

Just finding that crappy wiring job is worth a bit of sheetrock patching, so there's that.
posted by pipeski at 3:29 PM on January 18, 2018 [1 favorite]

Best answer: My concern was, if I do that, am I missing a deeper problem by blaming it on the dimmer?

I wouldn't think so. The dimmer works by lowering the voltage, and you're reading a lower voltage. It should be 120v instead of 95, but without an o-scope to see the waveform, who knows.

Anyway, the dimmer is going away, so we don't care. The wires that come into the dimmer from the service panel should measure ~120v across them even if you are getting 95 coming out. You can stick a probe into the cap for the neutral, and measure between it and the hot. (BE CAREFUL!)

In fact, you can positively identify the dimmer as the culprit by measuring between neutral and the hot coming in and the hot going out. I bet its 120 on one side and 95 on the other.
posted by Pogo_Fuzzybutt at 3:52 PM on January 18, 2018

Best answer: You have two issues. First, your dimmer is rated for a minimum load of 40 watts, so it isn't going to turn fully on unless you have that minimum load.

Second, your voltmeter is not a true-RMS meter. That requires a more expensive meter. Your meter will measure a rectified peak voltage, assume a sine wave, and then apply a 0.707 conversion factor to calculate an assumed RMS voltage. This will not be accurate in your case since the voltage coming out of the dimmer will not be a sine wave if not fully on.

So, as others have said, you likely don't have a problem. Just measure the voltage on the input side of the dimmer.
posted by JackFlash at 7:42 PM on January 18, 2018 [1 favorite]

Seconding what JackFlash said: the multimeter won't really measure the dimmer output accurately due to waveform weirdness.

https://www.google.com/search?q=multimeter+rms+dimmer
posted by intermod at 7:47 PM on January 18, 2018 [1 favorite]

Response by poster: Update: replaced the switch with a standard switch, 119.8 V afterwards. So it does appear to have been something with the dimmer. There are other problems with the circuit -- e.g. I opened up more ceiling and found a key junction of four 14-2 wires was done with no junction box, just a bunch of wires shoved up into the wood rafters and covered up by the ceiling sheetrock -- but the voltage was clearly unrelated to that.

Only weird thing now is that the wires to the light fixture light up a non-contact electricity tester, even if the new switch is off, though the multimeter shows 0.00 voltage. An electrician I know waved that away, though it's hard to let go of it.
posted by msalt at 6:09 PM on January 20, 2018

Best answer: Only weird thing now is that the wires to the light fixture light up a non-contact electricity tester, even if the new switch is off, though the multimeter shows 0.00 voltage.

That could potentially be a problem. Sometimes those non-contact testers can give you a false reading, but it might also be telling you that the fixture is wired so that the switch is in the neutral leg instead of the hot leg. This is not to code, and while it will work, could dangerous if someone touches the contacts of the fixture when the switch is off.

When you say your meter showed zero voltage, between what two points were you measuring? If you were measuring between the two terminals of the fixture, that tells you nothing because when the switch is open, there is no closed circuit, whether wired correctly or not.

To really tell if the fixture is wired correctly, you should carefully measure voltage from each terminal to the ground wire (assuming you have a ground wire). If the fixture is wired correctly, there should be no voltage from either terminal to ground when the switch is off.

If you find 120 V at either terminal when the switch is off, it means that the switch is incorrectly in the neutral leg instead of the hot leg as it should be.

I would not ignore the fact that your non-contact tester is seeing voltage when the switch is off.
posted by JackFlash at 6:40 PM on January 20, 2018 [1 favorite]

Also the fact that you have a junction in the ceiling with no junction box indicates that an amateur has done some rewiring which means you should be quite cautious. Stuff like this does not mean your house is about to burn down -- its not up to code and not imminently dangerous -- but it means you should be on the look out for other suspicious stuff.
posted by JackFlash at 6:45 PM on January 20, 2018 [1 favorite]

Response by poster: Thanks, JackFlash. I was measuring between the hot and neutral wires of the cable powering the light, which is currently disconnected.

I'll check the voltage to the ground from each once I rewire a proper junction box to replace that unboxed bundle. There is a ground wire at the fixture now, but the ancient Romex it branches off of has no ground wire so it's meaningless. I'm replacing the main circuit wire with new grounded Romex as part of this project.

I'm not sure I understand what you mean by "leg" here:

>it might also be telling you that the fixture is wired so that the switch is in the neutral leg instead of the hot leg

I'm wiring the hot (black) wire to the brass screws and the ground to the green screw. Would the problem you describe be =d neutral to brass? Thanks again.

PS I don't know if that unboxed bundle was amateur or just very old; the house was build in 1900, and two of the wires in that bundle are old, wooly wire. The connections under the wire nuts were soldered, interestingly.

But yeah, plenty of funkiness. This same circuit had 18 plugs and lights on it and kept popping the circuit breaker, which was what led me to start working on it in the first place.
posted by msalt at 7:46 PM on January 20, 2018

There are two correct ways of wiring a switch. The difference is if power goes to the fixture first or to the switch first. You can see the two methods here.

Note that in both cases and wired correctly, when the switch is off, there is no hot voltage at either terminal of the fixture. But if you swap the hot and neutral wires, while the switch and fixture will work, one side of the fixture will be hot even if the switch is off. Essentially, if wired incorrectly, the switch is interrupting the neutral leg rather than the hot leg.

Also note that in the first case, you connect two black wires to the terminals of the switch. In the second case you connect one black and one white wire to the switch, but you are supposed to wrap a bit of black tape around the white wire on both ends to mark that this is a wire carrying hot voltage. You can see the bit of black tape in the second picture.
posted by JackFlash at 8:13 PM on January 20, 2018 [1 favorite]

Response by poster: Thanks again, JackFlash. Final update:

I didn't have a chance to test the switch or the fixture until I had completely replaced the main circuit wire up to it. (Turns out that the cable stopped in at a handy box outlet mid-wall to switch from grounded to ungrounded Romex, for no apparent reason.)

Once I hooked everything up again, and replaced the joined wires that had just been shoved into the rafters with a proper junction box, it all operated properly with the way I had wired up the new switch. (IE 119.something voltage on hot to neutral and hot to ground, no indication of voltage from the non-contact tester when off.)

Thanks again!
posted by msalt at 10:37 AM on January 31, 2018