Crackling/sizzling noise

[fejog]

Update:

I bought a new WR-rated GFCI and wired it in exactly the same manner as the old one and I'm still getting the same result with my circuit tester - hot and neutral reversed. Under the pretense that it is in fact wired correctly, I decided to reverse the wires on the light switch (since I observed the sizzling sound coming from the GFCI when the light was on in addition to it flickering). After reversing the light switch wires, I placed my voltage tester near the light switch and the tone was now coming from the bottom terminal, the same as the fan switch. So, then I suspected that the light switch wires had been reversed. However (and again), when I plugged my circuit tester into the GFCI I got the same result. I'm not aware if the orientation of wires on light switches matters, but in this case it didn't make a difference; the light comes on regardless.

The only other thing that I can surmise (and I suspect) is that one of the other two pairs of wires going to the GFCI should be on the Line side, and the pair connected to it now should actually be on the Load side. What say you? Otherwise, I don't know what else I can do to try and rectify the problem. When I put the voltage tester near the brass terminals on the GFCI I got a tone from both the top and bottom.

A sizzling or crackling noise from your GFCI suggests a loose connection or internal failure, which can be dangerous. The reversed hot/neutral reading, despite the wiring appearing correct, could indicate an issue elsewhere in the circuit. Turn off the breaker and check for loose connections at the GFCI and other outlets on the same circuit. If any wires are backstabbed, move them to the screw terminals for a more secure connection. If everything is tight but the issue persists, the GFCI may be faulty and should be replaced. If the problem continues after that, it's best to call an electrician.

[Bodrey]

Thanks everyone. Yes, I'm glad I was able to resolve the issue. The fix wasn't quick or easy (is it ever?), but having to replace the run between the panel and the junction box was a blessing in disguise because there was no ground. It needed to be done anyway, and as a result it also cleared up the false negative I was getting with my tester.

[JimmyMechanic]

Wonderful! So glad!

DONE!!! I finished re-wiring the entire circuit and re-installed the GFCI today. Of course, I tested it before putting everything back and it's all good. It turns out that I should not have gotten any indicators lighting up when I tested the GFCI previously when pressing the GFCI button. When I tested it on my kitchen GFCIs pressing the button caused them to trip their breakers. I'm surmising that the absence of a ground wire (from the run between the panel and the junction box) meant that the tester was not able to properly test the GFCI as intended (tripping its breaker). Instead, because of a lack of ground wire, it threw a "hot/neutral reversed" fault. Problem solved - finally!

[TinyTool]

Congrats on getting everything rewired and working properly!

[machinist]

That's awesome, Bodrey! Great job troubleshooting and seeing it through to the end. It’s interesting how the missing ground affected the tester, definitely something we can all learn from when dealing with this kind of issue. Must feel good to have everything wired up correctly and working as it should. Well done!

[Bodrey]

DONE!!! I finished re-wiring the entire circuit and re-installed the GFCI today. Of course, I tested it before putting everything back and it's all good. It turns out that I should not have gotten any indicators lighting up when I tested the GFCI previously when pressing the GFCI button. When I tested it on my kitchen GFCIs pressing the button caused them to trip their breakers. I'm surmising that the absence of a ground wire (from the run between the panel and the junction box) meant that the tester was not able to properly test the GFCI as intended (tripping its breaker). Instead, because of a lack of ground wire, it threw a "hot/neutral reversed" fault. Problem solved - finally!

[machinist]

Replacing that old cable with NMD90 should definitely clear up the missing ground issue and also help resolve the hot/neutral reversal if that’s part of the problem. Good call on running a new cable to the GFCI too. It’ll give you a clean, properly grounded setup without relying on pigtails.

And yeah, working on old wiring always has a way of turning into an unexpected adventure. Hopefully, this one goes smoothly for you, but either way, it sounds like you're tackling it the right way. Good luck! 💪🔧

[Bodrey]

I think I discovered the reason for the 75V reading. I opened the junction box under the vanity today. The cable that runs from the panel to the junction box is of the older black-sheathed type. It does not have a ground wire! So, I'm going to replace that run with NMD90. In case JM is right ("Especially in older wiring black and white might be swapped) this should also eliminate the hot/neutral reversed fault I'm getting with my tester.

Instead of pig tailing the source ground wire at the GFCI I'm just going to replace the cable between the junction box and the GFCI as well. That will give me plenty of lead to wrap around the box screw before bundling with the other two cables' ground wires.

I hope it all goes smoothly but like everything else around this old house it will likely turn into an ordeal. Wish me luck... 😏

[machinist]

That definitely sounds like a possible factor! If the source cable’s ground isn’t making solid contact with the box, it could create an inconsistent ground reference, which might explain the 75V reading. Ideally, all ground wires should be securely bonded together and connected to the box if it’s metal.

Since you can’t do full continuity tests, I’d say double-check that all grounds are tightly twisted together and making good contact. If you can, try adding a short pigtail to properly connect the source ground to both the box and the other grounds. That might stabilize the voltage reading.

[Bodrey]

There's a junction box located behind my bathroom sink. I'll check the connections (in particular the ground to make sure it's secure). However, I'm not able to do continuity tests between any of the wires except those located in the GFCI box itself. I'd need a tester with a separate transmitter and receiver to do that (the panel is in the basement).

However, I have another theory about the 75V reading. Let me know if you think this is a factor.

The source (hot) cable's ground wire was cut too short to wrap around the ground screw on the box and still have enough lead to be bundled with the other two ground wires. I had to use the ground wire from one of the "downstream" cables to ground the circuit to the box screw. I had to bundle the source cable's ground wire directly with the other two ground wires.

Could the fact that the source cable's ground wire isn't secured to the box explain the discrepancy in low output voltage?

[machinist]

The 75V reading between hot and ground is definitely a red flag. This suggests there might be a loose connection or a poor ground, which could be contributing to the flickering light and sizzling sound.

Since you've confirmed the tester is working correctly by checking your kitchen GFCI, we can narrow this down to either a ground path issue in this specific circuit or potentially a faulty GFCI unit. Before replacing another GFCI though, I'd recommend checking the ground connections at all points in this circuit - particularly at the panel and any junction boxes along the way. A compromised ground connection can create exactly the symptoms you're seeing.

If you're comfortable doing so, could you check the ground wire continuity from the GFCI box back to the panel? This might help pinpoint where the ground path is being compromised. Just be sure to have the power off when testing continuity!

Turn off the power to the circuit at the breaker panel. Double-check that power is off using your voltage tester. Then:

1. Using your analog multimeter, set it to the continuity or resistance (ohms) setting. Most meters will beep when continuity is detected, but you can also look for very low resistance readings (close to 0 ohms).
2. Connect one probe to the ground screw terminal of your GFCI box, and have someone help you connect the other probe to the ground bus bar in your electrical panel. If you get a reading of infinity or very high resistance, there's a break in the ground path.
3. If you don't want to check all the way to the panel, you can test sections at a time by checking continuity between:
   • The GFCI ground terminal and the metal box (if it's metal)
   • The ground wire at the GFCI and the ground wires in nearby junction boxes
   • Between different ground wire connections in the same box

Remember to maintain firm contact with your meter probes when testing, as even a slight movement can affect the reading. Keep in mind that a good ground connection should show very low resistance or immediate continuity.

[Bodrey]

I tested the wiring using an analog multimeter and this is what I found:

- Voltage between the hot and neutral wires was within normal range (120V)
- Voltage between neutral and ground was near 0V as expected
- Voltage between hot and ground was only about 75V! So, there's definitely something odd going on

JM - The wiring being used appears to be the newer NMD90 standard (white sheath). However, it's easy enough to reverse the wires on the GFCI temporarily just to see if the hot/neutral reverse fault disappears using the tester.

Machinist - I plugged the tester into the GFCI above my kitchen counter and both indicator lights were equally bright. So, the issue is not with the tester - it's either the circuit or the GFCI itself.

[JimmyMechanic]

I'm wondering if something "upstream" is reversed.

To test for a hot wire versus a neutral wire, use a multimeter set to AC voltage, touch one probe to a known ground (like a metal screw on the outlet box), and then touch the other probe to each wire in question; the wire that shows a voltage reading close to the standard line voltage (usually around 120 volts) is the "hot" wire, while the wire with a near-zero voltage reading is the "neutral" wire.

Key points to remember:

• Always turn off the power at the circuit breaker before testing .
• Use a multimeter set to AC voltage .
• Touch one probe to a known ground .
• The wire with the higher voltage reading is the hot wire

I read the entire thread but might have missed, have you tested the incoming wires individually yet? Especially in older wiring black and white might be swapped. That is generally because it mattered less in (much) older installations.

Another way to test is to just use a voltage tester. They look like a pen and should only illuminate (or light up, or make a sound) when you are touching the hot wire.

Let us know!!

[machinist]

The back-wire clamp system is generally reliable, but it’s good that you verified the connections. Reaching out to Klein and Leviton is a smart move.

In the meantime, if you have a multimeter, you could check the voltage between hot, neutral, and ground to see if there’s anything unusual. If everything checks out but the tester still shows conflicting results, it might be worth testing with a different GFCI tester to rule out any inconsistencies with the Klein unit. Keep me posted on what you hear from Klein and Leviton. I’d be curious to know if they provide any insight that helps clarify this.

[Bodrey]

My particular GFCI doesn't have back stab connectors. Instead, it has something called back wire clamps. You insert the bare wire behind it and then tighten the terminal screw down to hold it in place. I tugged on the connections and they all seem to be secure. I also connected a lamp to it which worked fine - normal brightness and didn't flicker at all.

I contacted Klein to inquire whether the dim right light on their tester is by design or does in fact indicate a wiring fault. I also plan to inquire with Leviton about the wiring configuration and to ensure that the GFCI is in fact labeled properly. Perhaps the GFCI itself is faulty (unlikely but who knows).