Help with understanding noise

[caspercody]

Here is picture of the water pipe. If I take out the bottom wire I see sparks, with bottom one out still and I remove top wire there is no sparks

[Rob Strand]

Here is picture of the water pipe. If I take out the bottom wire I see sparks, with bottom one out still and I remove top wire there is no sparks

That makes sense. 

The source of the problem is up in the house and the current wants to flow to earth.    The earth I'm referring to here is to earth outside of the house and that's where the lower connection goes to.

When you break the bottom connection you are breaking the earth side.  So you are breaking the current flow to ground and hence seeing a spark.   

When you break the top side the current flow to earth is not broken.  The current can still flow through the bridge wire and down the lower side to earth.   What you are doing by breaking to top connection is disconnection the house side water pipes from the external earth connection.

[caspercody]

I was hoping to see if it could help explain if the problem is in my house or coming from outside my house?

That copper wire is still connected to earth ground by a delegate copper rod connected to the electrical panel going outside to ground

[Rob Strand]

I was hoping to see if it could help explain if the problem is in my house or coming from outside my house?

One point to realize is the presence of the spark might be totally unrelated to the noise problem!

The spark is likely to be caused by devices (or a device) in the house.  No guarantee this is 100% the case though.  You would have pull everything out of the power points to narrow it down.   It could be a cumulative thing as well.  However if you fixed devices like a water heater or stove have a problem you will still see the spark as you cannot disconnect them.  You would need to pull the fuses feeding those devices to see if that had an effect.

I would recommend borrowing a clamp meter to measure the AC current going down the earth wire instead of looking at the spark.  That would also give you an idea of scale of the problem.   Electricians use these things all the time.

The noise could be coming from outside and not causing the spark, or, it could be coming from outside *and* causing the spark.   The last case being the least likely case.

[caspercody]

Maybe tomorrow I will try turning off the main breaker to the house and remove the wire and see if it sparks.

I do know if I remove the wire from the bottom I do not get the noise in my amp when it is removed

[amptramp]

This sounds like the problem you would get if someone had interchanged the ground and neutral wire at one of the sockets.  You would not necessarily see it with a socket tester that has three neon lamps since the voltage may not be high enough to light up the lamp for ground to neutral.  You may be able to localize it by successively removing fuses and putting them back again and seeing if the hum goes away.  Do not leave this situation as it is - it could be dangerous.

[caspercody]

Thanks everyone for your help!!

I turned off the main breaker in house, to see if the spark would be gone. It was gone, darn it. So I started turning on breakers one by one to see what circuits cause the spark, I am down to one that causes the biggest spark, and two that have a tiny spark. Unfortunately, these (3) circuits cover a lot of my house for receptacles and lights

When I turned off the main breaker the EMF meter was still showing EMF. Still with power off to the house, I removed the wire, no spark, and EMF meter quiet. I reattached the wire, still no spark, but EMF meter indicated EMF present. Again all of this is still with main breaker off to house.

[Rob Strand]

When I turned off the main breaker the EMF meter was still showing EMF.

Maybe there's two problems! One inside and one outside.

amptramp's swapped wire theory is possibility for the in-house issue for sure.

[andy-h-h]

You should probably call an electrician as that sounds potentially dangerous.    I'm in Australia and power here is a little different to USA as we run at 230v 10amps.  We used to use plumbing as part of the earthing system, but stopped as a result of the related deaths...   if there's current on earth, something is wrong.   

[caspercody]

I measured 1.6 volts from the UN-clamped ground wire to ground (water pipe). I want to measure current but I am not sure how to hook up a multi meter into this?

[EBK]

I measured 1.6 volts from the UN-clamped ground wire to ground (water pipe). I want to measure current but I am not sure how to hook up a multi meter into this?

A clamp meter could be used to measure the current.

You could borrow one, buy one,

or DIY one:

[PRR]

That DIY clamp looks frikkin' dangerous. The open-circuit voltage could be VERY high.

Clever though.

I would NOT put a multimeter in-line with a ground jumper. While current "should be small", even that is more than some DMMs are good for. And you have a Problem. While unlikely, you could have A Hundred Amps in there. Car battery, not bench-work.

This really is a job for Professional Electricians. Both Power company and you-pay. (They will point the blame at each other, but the one you pay helps keep the company honest.)

[Rob Strand]

I would NOT put a multimeter in-line with a ground jumper. While current "should be small", even that is more than some DMMs are good for. And you have a Problem. While unlikely, you could have A Hundred Amps in there. Car battery, not bench-work.

Agreed, I thought the same thing.   Even on a 10A range it's not the thing to do these.days *especially* with a weird problem going one.  I still can't work out exactly what is going on.

That DIY clamp looks frikkin' dangerous. The open-circuit voltage could be VERY high.

Clever though.

Would need calibrating as well.  My feeling is the *measured* voltage would be small.

[PRR]

> Would need calibrating as well.

If well clamped, the current ratio is simply the turns ratio. Maybe 100:1? Eddy currents and poor magnetic contact effects would be large for precision metering (I have the book....zzzzzz.....) but non-issue for getting an idea of the current.

[Rob Strand]

If well clamped, the current ratio is simply the turns ratio. Maybe 100:1? Eddy currents and poor magnetic contact effects would be large for precision metering (I have the book....zzzzzz.....) but non-issue for getting an idea of the current.

Yes, it might give you an idea.      Perhaps it would be more sensitive with a voltage output but it would need a load resistor.   That would also reduce the open circuit voltage problem.

[PRR]

> more sensitive with a voltage output

Current transformers MUST work into a "short". The proportions don't work as a voltage output. (Side-thought: transformers only work predictably if at least one side sees a lowish impedance.) If you disconnect them from the meter, you first short the CT *then* take the wires from the meter off.

"Short" is of course some very low resistance, known in the trade as "burden". How big a meter or breaker-coil it can power without loss of rated accuracy. (Handheld digital meters can use much smaller CT coils than ones used for power metering with traditional spinning-disk power meters.)

The ideal open-circuit voltage is "infinity". In real life limited by magnetic saturation. The big pole-top meter CTs would deliver a fatal shock if un-shorted and "any" current was flowing. The tiny doughnuts I used in my digital house demand meter apparently can't make 10V open-circuit. The OC Voltage of a G-clamp CT is ? ? ?

[Rob Strand]

All I was implying in my post is you can play with the load resistance if the sensitivity isn't enough.

Current transformers MUST work into a "short". The proportions don't work as a voltage output. (Side-thought: transformers only work predictably if at least one side sees a lowish impedance.) If you disconnect them from the meter, you first short the CT *then* take the wires from the meter off.

"MUST" is a bit too strong.  Resistively loaded current transformers are pretty common.

https://www.hioki.com/en/products/listUse/?category=17

This is a tradesman type meter where the current transformer essentially works as a "current to voltage" adaptor (1A AC = 1mV AC) ie. the meter is set to the AC volts range.    The clamp is entirely passive.

"Short" is of course some very low resistance, known in the trade as "burden".

Yes, it's generally a low value resistor.

If you make the load resistance too high the current-transformer will start to high-pass filter the AC waveform and you get errors. 

While this example is for a high frequency current transformer you will still see the same low-frequency roll-off effect when the load resistor gets too high.
https://www.researchgate.net/figure/Transimpedance-gain-of-the-current-transformer-as-a-function-of-the-value-of-the-load_fig12_3280950

For the "G-clamp" the roll-off point will be end-up corresponding to a low resistance value.

[stallik]

Did you find a solution?

[caspercody]

No, I gave up. I called the city because according to the electrician in our office they are required to supply clean power, but they had no idea what to do. I just try to find that one quiet spot and do not move.

Thanks for asking!

[stallik]

That's really frustrating. Some years ago, all the PC's in my office would loose every usb device. They would reconnect after a few seconds but large format printers would drop the job and that was embarrassing during a demonstration.

There was a cell mast literally outside the room which was the easiest thing to blame till we noticed that the issue only occured when the air conditioning kicked in. The air con engineers gave up after a few visits and many replaced parts then the almighty lent a hand by striking the cell mast with lightning, raising it to the ground. At this point, the problem went away no matter the state of the air con.

Some months later, the problem came back - at exactly the moment the replacement mast was turned on. Clearly a combination of the air con and the cell so we got everyone in to sort it out.

2 years of visits from air con, electric and phone people (who all blamed each other) they all gave up. So did we