High current on neutral with a balanced load

[synchro]

I assume the voltage waveform on the neutral Vn is more noisy because the scale has been made more sensitive for it on the analyzer? Also, what was the reference point for these voltage measurements (perhaps the GES or EGC)?

 

[jhardy13]

If I'm reading your analyzer shots correctly, there is roughly 20A of DC current on Ia.
Ib has little or no DC, but the waveform looks like the input current of a transformer would if it was saturating near the peaks of the applied voltage (the current will rise up because the inductance drops significantly during core saturation).
In does not appear to have a DC component. And so the DC current on Ia must be returning on some other conductor, most likely the GES and/or EGC.

Why do you think there is DC current? I don't fully understand what you mean.

 

[synchro]

Why do you think there is DC current? I don't fully understand what you mean.

Your second picture (20200825_121849.jpg) shows the waveform of Ia going between a positive peak value of about +11A and a negative peak value of about -48V. The waveform is reasonably symmetrical in the positive and negative directions, and so the average value of the waveform over one cycle (and over the long term) is about (11A - 48A) / 2 = -19A. This is the DC component of the waveform. A waveform that only has an AC component and no DC component will average to zero.

Your last picture (20200825_121701.jpg) shows the waveforms of both Ia and Ib in the same plot. You can see that once again the average of Ia is about -19A DC.
Now maybe a -19A DC offset has been applied only to the channel with the Ia waveform by using an adjustment within the analyzer itself, but I think that's unlikely (and maybe not even possible).

 

[jhardy13]

Your second picture (20200825_121849.jpg) shows the waveform of Ia going between a positive peak value of about +11A and a negative peak value of about -48V. The waveform is reasonably symmetrical in the positive and negative directions, and so the average value of the waveform over one cycle (and over the long term) is about (11A - 48A) / 2 = -19A. This is the DC component of the waveform. A waveform that only has an AC component and no DC component will average to zero.

Your last picture (20200825_121701.jpg) shows the waveforms of both Ia and Ib in the same plot. You can see that once again the average of Ia is about -19A DC.
Now maybe a -19A DC offset has been applied only to the channel with the Ia waveform by using an adjustment within the analyzer itself, but I think that's unlikely (and maybe not even possible).

I see what your saying. I am not knowledgeable enough on this to understand what that means or how that could affect our electrical system. Is it something that should concern me and should I bring it up with the engineer that is helping me? Also does it point to a certain specific cause? The engineer was adamant that current harmonics can only come from devices within my home. Is this true or could it be caused from an external source such as our neighbor that shares our transformer?

 

[GoldDigger]

If there is no harmonic content to the voltage waveform coming from POCO, then any current harmonics on your side of the service point would have to come from your loads. With the caveat that we are measuring neutral current, while the true return current is neutral current plus earth current in the grounding system to the POCO ground at the transformer pole.
'
It would be very interesting (though impractical) to measure the neutral current right at the POCO transformer secondary.

 

[synchro]

I second what GoldDigger says that your loads must be causing the nonsinusoidal current waveforms that have significant harmonic content.

The nonlinear loading is present between one of the phases and neutral based on your first and third pictures, while the other phase has a much more sinusoidal current in the second picture and therefore its loads must be relatively linear.

If it's really there, the large amount of DC current that's evident on one of the phases is highly unusual. I think it would be desirable if this could be verified. Unfortunately, from a quick glance at their specs the meters shown in your pictures on this thread do not measure DC current when using the clamp. But there are clamp meters that measure DC as well as AC current.

 

[jhardy13]

I second what GoldDigger says that your loads must be causing the nonsinusoidal current waveforms that have significant harmonic content.

The nonlinear loading is present between one of the phases and neutral based on your first and third pictures, while the other phase has a much more sinusoidal current in the second picture and therefore its loads must be relatively linear.

If it's really there, the large amount of DC current that's evident on one of the phases is highly unusual. I think it would be desirable if this could be verified. Unfortunately, from a quick glance at their specs the meters shown in your pictures on this thread do not measure DC current when using the clamp. But there are clamp meters that measure DC as well as AC current.

I will find someone who has a clamp that can measure DC current and update.

 

[ramsy]

Someone's running a bitcoin processor with UPS back up, which kicks in on shutdows.

 

[jhardy13]

Someone's running a bitcoin processor with UPS back up, which kicks in on shutdows.

Are you implying that these current harmonics could be caused by someone that shares our transformer? Because we do not have a UPS system or bitcoin farm, but it is possible that our neighbor does. The cyclical nature of the harmonics we recorded are very strange. The THD% seems to spike up dramatically at night while we are sleeping and almost everything in our home is turned off. See picture #5. It shows time on the X axis and THD% on the Y axis.

 

[Electricmo]

Take a look at this article. https://electricalacademia.com/electric-power/harmonics-come-causes-harmonics-power-system/

 

[jhardy13]

I apologize for my ignorance on the topic of harmonics. I've done a lot reading since my last few posts and have a much better understanding now. I went ahead and purchased an ammeter probe for my oscilloscope and these are some current waveforms that I have measured on our mains. The black waveform is phase A and the red waveform is phase B. I have been switching breakers on and off in search of what could be causing the significant distortion to phase B. I started by turning off all of our two pole breakers and then proceeded to turn all of phase A off to isolate phase B. What I found is that our kitchen circuit is causing the most distortion but even with it turned off, there is still a considerable amount of distortion no matter which breaker on phase B is turned on. I have a few questions for the experts here. Should the waveforms be more sinusoidal? What level of current distortion is acceptable? What negative affects could I expect from these waveforms (if any)? It may be unrelated but I should mention I have had to replace several high end power supplys for my pc and we have had to buy three new vacuum cleaners in the last 6 months as well as a few smaller appliances such as a keurig coffee maker. We also used to have a serious problem with lights burning out constantly. Our solution was to replace all of our light fixtures with LED fixtures because they last much longer. And lastly, are there ways to mitigate the distortion? The engineer will be back tomorrow morning because we have been continuing to monitor the power with the PA9 Plus. Is there anything specific that I should ask him that may help find the issue?

 

[jhardy13]

Waveforms

 

[jhardy13]

Waveforms

I should also note, at the top of the photos you may notice it shows DC. That is the coupling mode. This is the description: DC coupling allows you to see all signals from 0 Hz up to the max bandwidth of your scope. AC coupling filters out DC components. When you enable AC coupling on an oscilloscope channel, you're switching in a high-pass filter on the channel's input signal path. This filters out all the DC components.

 

[GoldDigger]

For what it is worth, problems with lights burning out and appliances failing are often caused by large voltage surges (visible high spikes on the oscilloscope) or by excess voltage caused by a compromised neutral. Harmonics alone are not a problem unless they result in a sharp edged (square) waveform.

Sent from my Pixel 4a using Tapatalk

 

[synchro]

The engineer will be back tomorrow morning because we have been continuing to monitor the power with the PA9 Plus. Is there anything specific that I should ask him that may help find the issue?

Your previous waveforms from the PA9 Plus showed a large DC component on what is apparently Phase A in your latest measurements, but yours don't have any significant amount of DC. I would ask the engineer to check for DC anyway if he can measure it with his equipment.

Do you know if the current probe you got for your scope will measure down to DC? If it doesn't explicitly say that it can, then I suggest testing it at DC. For example you could put the clamp onto one cable to your car battery and then measure it with the lights turned on but with the engine not running.

I also suggest measuring the current waveforms on the neutral conductor, preferably at multiple places if accessible. For example on a wire jumper between neutral bars if there's one present. Of course be very careful of exposed live conductors. Also put your current probe on any conductors of the grounding electrode system that the probe can fit over and see what, if any, significant current waveforms are present.

As GoldDigger mentioned it's almost always excess voltage that will kill appliances. It's very unlikely that harmonic currents from other loads would damage an appliance as long as they do not excessively distort the voltage waveform, such as causing high level ringing . But tracking down unusually distorted current waveforms still might lead to clues about what it is about your system that may be increasing the vulnerability to appliance damage, if not outright causing it.





.

 

[jhardy13]

Your previous waveforms from the PA9 Plus showed a large DC component on what is apparently Phase A in your latest measurements, but yours don't have any significant amount of DC. I would ask the engineer to check for DC anyway if he can measure it with his equipment.

Do you know if the current probe you got for your scope will measure down to DC? If it doesn't explicitly say that it can, then I suggest testing it at DC. For example you could put the clamp onto one cable to your car battery and then measure it with the lights turned on but with the engine not running.

I also suggest measuring the current waveforms on the neutral conductor, preferably at multiple places if accessible. For example on a wire jumper between neutral bars if there's one present. Of course be very careful of exposed live conductors. Also put your current probe on any conductors of the grounding electrode system that the probe can fit over and see what, if any, significant current waveforms are present.

As GoldDigger mentioned it's almost always excess voltage that will kill appliances. It's very unlikely that harmonic currents from other loads would damage an appliance as long as they do not excessively distort the voltage waveform, such as causing high level ringing . But tracking down unusually distorted current waveforms still might lead to clues about what it is about your system that may be increasing the vulnerability to appliance damage, if not outright causing it.





.

The clamp I got does measure DC as well. It is a TA019 from the same company as my scope. Yesterday I did in fact see some DC components on my scope, but it must be intermittent because it was not present while I measured today. I was specifically looking for it since you mentioned it the other day. I'll keep monitoring it to see if it shows up again and I will ask the engineer about it as well. What could cause that? Seems really strange. Also, should the current frequency be 60hz just like the voltage frequency? I was seeing some fluctuation. It spiked a few times, the highest being 72hz for a split moment then returned to bouncing up and down between 59hz-61hz. It may not matter but figure I should ask anyways.

 

[GoldDigger]

The current frequency fundamental must be exactly the same as the voltage fundamental frequency.
But a distorted waveform, especially with visible ringing, can cause an incorrect reading on a frequency counter that simply counts zero crossings.

Sent from my Pixel 4a using Tapatalk

 

[jhardy13]

The current frequency fundamental must be exactly the same as the voltage fundamental frequency.
But a distorted waveform, especially with visible ringing, can cause an incorrect reading on a frequency counter that simply counts zero crossings.

Sent from my Pixel 4a using Tapatalk

What do you mean by ringing?

 

[GoldDigger]

What do you mean by ringing?

What appears to be a higher frequency waveform visibly riding on the fundamental waveform. If the amplitude is high enough the composite waveform can cross the zero line several times at the point that the fundamental goes through zero.
More properly ringing is used to refer to a decaying AC oscillation caused by an applied voltage or current transient.

Sent from my Pixel 4a using Tapatalk

 

[LarryFine]

The only sources for DC I can think of are cable-TV equipment and severe electrolytic action.