Distorted residential power, only at night? 3

[st50maint]

For many weeks I have had this problem with my power. I hear a slight buzz from my audio equipment, only at night. Hooking up an oscilloscope shows this waveform.

?
At dusk the waveform becomes dirty. At dawn it becomes clean. Everyday like clockwork.
Some more clues:
1. The problem is only on one of the 120 volt legs. The other leg is always relatively clean.
2. I thought that street lights might be a problem because of the timing. However I observed them on a few occasions and they did not come on or off exactly with the start or end of the distortion.
3. This is a residential only neighborhood with 14 houses and 3 streetlights on one transformer as far as I can tell.

Any thoughts on what is causing this and a possible solution?

 

[jghrist]

The flat top voltage looks like there are a lot of switched-mode power supplies on line. Switched-mode power supplies power computers and similar equipment and conduct current only at voltage peaks. This causes voltage drop at voltage peaks, flattening the waveform. Squiggley CFL bulbs also use switched-mode power supplies. Maybe your neighbors are environmentally conscious and have quit using the old nasty resistive type incandescent lamps that don't cause powerline distortion.

 

[wcarter]

st50maint,

That appears to be harmonic distortion. Have you tried using your osilloscope to check for which odd harmonic it being applied into your local line?

 

[st50maint]

The flattening of the waveform is not what comes and goes. It is the spikes just after the peaks that I am concerned with.

 

[dpc]

The first step would be to start switching off circuit breakers in your house to determine how much of this distortion is being created by your loads.

14 houses on one transformer is a lot of houses. Are you sure about this?

Also, where are you measuring this voltage?

One of your neighbors have an indoor plant growing operation?

David Castor

http://www.cvoes.com

 

[Compositepro]

Did any of your neighbors get an electric car recently?

 

[Skogsgurra]

That's a very intelligent observation, Compositepro.

This is not a combination of several SMPS-es, but looks more like one hefty rectifier taking a lot of current over a conduction angle that is close to 45 degrees, which is much more than one can see in a typical SMPS. The fact that current snaps off distintly and causes significant ringing at the end of the conduction angle also tells that this is one consumer and not the typical 'co-operation' of many small loads.

The ringing is probably the origin of the buzz you hear.

Charging an electric car could very well be such a load.

Interesting.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[st50maint]

I have switched off all breakers except the one powering the scope. Then I put the scope on a different line and shut off the original breaker. This was done at the outlets closest to the circuit breakers.

I am pretty sure that there are 14 houses on the transformer.

I am not aware of any electric cars or indoor plant growing but I will ask around.

 

[jraef]

... indoor plant growing...

Careful, those guys tend to not appreciate being found!

I would start by finding a friendly neighbor and ask to connect your scope to his system too and see if the problem exists outside of your home. If they don't have it, the problem is closer to home. If they do, and if you can get to enough other friendly neighbors, you may find a variation in the level of attenuation going on by distance and that may allow some gross triangulation toward the culprit.

"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[stevencress]

You might be able to take a small portable radio tuned to the hi end or low end of the AM dial and walk down the street if front of the 14 houses and see if the noise lever increases in front of one house. Even better is to check at the utility meters. Sounds like a nearby neighbor may have a light dimmer, old electric blanket, a lamp that goes on/off by touching it. May not even be on the same utility transformer.

 

[Skogsgurra]

Not much HF in that. Will probably not be picked up by a radio. Not a low wattage load like a lamp or blanket either. This is substantial.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[PHovnanian]

If you 'scope the neighbors services, you'll have to take a look at their current waveforms as well as voltage. Everyone will be seeing more or less the same voltage. Particularly if this waveform is caused by the interaction of a non-linear load with the common utility transformer.

You'll be looking for a step in the load current that coincides with the notch in the voltage waveform.

Searching for radiated RF can fool you. At times, the signal can be conducted a ways along a service line and then radiated from a point some distance away from the culprit.

 

[ScottyUK]

skogs,

Don't forget that the US seem to use small pole-mount distribution transformers of a hundred kVA or so serving a small number of properties. That certainly contrasts with British practice where a suburban distribution transformer is more typically in the 0.5 - 1.5 MVA range and serving a much larger number of properties. The 'single big load' might not have to be all that big to cause a problem: maybe someone with a home workshop running a lathe or milling machine on an inverter? I have never looked at my domestic mains when my lathe is running, but it certainly upsets the ADSL broadband modem connection.

Another thought: how would a high primary voltage pushing the transformer core in to moderate saturation manifest itself in terms of secondary voltage waveform? Light loads on the system in the evening would tally with a high primary voltage.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[itsmoked]

st50maint; The flat top is what you should be interested in not the little squiggle after it. It's the flat top that's causing the squiggle.

Is the time this starts in the evening and ends in the morning accurate and repeatable? I ask because if it is then it's probably not a consumer but rather a power company caused function like a switched capacitor voltage regulator. If it were, say, an electric car being charged it would end at different times and start with a multi minute variation.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

Yes, I have had one of them just outside my window in Elizabeth, NJ. Not exactly beautiful. And that arrangement with two transformers in one pole to get that open delta does not make it any more beautiful. 'My' transformer kept buzzing all day and all night. I couldn't sleep - waiting for it to explode any minute.

Yes, but even 100 kVA is big compared normal switch-mode power supplies. A conduction angle close to 45 degrees is not usual. I am very excited to learn what it actually turns out to be.

BTW, many PWM inverters take out communication. But they do it with HF from the PWM switchiing. Not with low-frequency distortion from the rectifier.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[st50maint]

It is always good when I get up about 8 AM. It is always bad when I get home after dark. The few times I have been able to observe the transition it has been fairly accurate and repeatable, within minutes of my street lights going on or off.

"Is the time this starts in the evening and ends in the morning accurate and repeatable? I ask because if it is then it's probably not a consumer but rather a power company caused function like a switched capacitor voltage regulator. If it were, say, an electric car being charged it would end at different times and start with a multi minute variation."

 

[davidbeach]

within minutes of my street lights going on or off

A bad street lighting ballast. The problem may be thermally induced and therefore not present at turn on, but only shortly thereafter.

 

[ghostbuster7]

We had a case recently,in an industrial area where the voltage on the step down side of the dedicated factory transformer would increase by approx. 20% with lots of harmonic distortion.It would stay at this level for 10-15 minutes each time.It was later found down the street,the scrap metal yard(supplied from another step down tx.) was turning on their large electromagnet to pick-up the cars.

The point is :harmonic/waveform distortion on the secondary/primary tx. of offending client can be transferred to the tx.secondary side of an unsuspecting client.

I would lodge a complaint with the utility and let them investigate.It could be a large load miles away that is turning on.

 

[magoo2]

a switched capacitor voltage regulator? What is that?

A switched capacitor is a shunt capacitor with an on/off switch.

A voltage regulator is an autotransformer with a tap changer.

I'm just curious what you mean by a switched capacitor voltage regulator.

 

[jghrist]

I agree with Gunnar that it looks like a single load. A bunch of small loads might cause the flat spot, but they all would not fire at exactly the same angle and the transient ringing would cancel out. It isn't switched utility capacitors. These could cause a high frequency ringing, but it would be a single transient when the capacitor is switched, not something that happens every cycle and shows up on an oscillograph.

How big a load? It looks like about 30 volts is chopped off the voltage waveform. Say the transformer is 25 kVA with a 2% impedance. This is about 0.012 ohms at 120 volts. Say the load is 200 ft from the transformer and the secondary is #1/0 aluminum. The two-way impedance would be 0.076 ohms. These are at 60 Hz. The impedance for the peak part of the current would be probably be 4 times this because of the increased frequency. So, we have 30 volts divided by 0.088 ohms, or 340 A peak, but only lasting . That's a very large load. If it was an electric car charging overnight, I would expect the load to be very small by morning when the batteries are fully charged.