How to measure mains waveform safely?

[JoshK]

I have a suspicion that my mains waveform is pretty ugly. Every transformer that gets hooked up buzzes to some extent.

My e-linear amp has some hum...ok, it could be my layout, but when you put your ear to the transformers they buzz mechanically, almost as loud as the speakers (you have to put your ear equally close). Other amps built produce the same mechanical buzz in the transformers.

Maybe it is just my substandard builds? Maybe, but every OneAC unit I've plugged in hums, and I have 5. I've built a few power filters using isolation transformers and they all hum.

So I am beginning to suspect my mains is pretty distorted. I live in the red headed step child community of my city so it wouldn't surprise me if we had subpar service.

I'd like to measure the waveform of my mains with my o'scope. Can any of you recommend a good way to do that?

I realize the liability. I am my own electrician and take full responsibility. I install my own circuits and often work on my breaker box, so I am not in the dark about the hazards and know how to work with live mains.

P.S. I have a stingray o'scope too...its a usb type oscope and you can use its software to do FFT. So maybe i can analyze the distortion.

[wg_ski]

Originally posted by JoshK
I'd like to measure the waveform of my mains with my o'scope. Can any of you recommend a good way to do that?

Any transformer that can be run far below saturation and very little load should have little distortion of its own. A 240v primary operated at 120V, a typical 200-400V tube mains trafo run backwards, even a large PP OPT will be running far enough down the BH curve that it won't affect a measurement of harmonics on the power line. Use a resistive voltage divider of a few k ohms to drop the secondary down to something the scope can handle. At light load (or no load, if the scope's input can handle it without a divider) leakage inductance will be negligible enough not to filter anything you want to measure.

[JoshK]

Good idea. I have some 500va step down transformers, that could work.

[aardvarkash10]

what happens if you turn off EVERY other appliance etc in the house? Still get the hum/buzz? Cos think about it - the waveform in the standard AC line is a pure function of the spinning of a generator somewhere... Does your TV hum or buzz? Any other appliance hum or buzz?

[tubelab.com]

For a different reason I took a standard 6600 ohm P-P OPT with the 8 ohm load, the scope, and the distortion analyzer connected across the secondary, and plugged the primary into the wall outlet. See the second post in this thread:

Screen drive and other P-P experiments

If you think about it the OPT should pass all of the audio frequencies with relatively equal loss, especially at this low power level (about 2 watts). It is also insulated for the plate voltage which can be near 1KV for a 60 watt amp, so the line voltage is not a problem.

What did I see? 10% distortion with a flattened top, not unlike an amp operated at the onset of clipping. Turning things on and off had little effect. The biggest contributors are the TV and the computer but the distortion improved 0.02% with both off. If you think about it there are 20 miles of other users between me and the power plant. Now if they ALL turned off the TV.... It would be interesting to try this at different times during the day, or night.

[JoshK]

I share a pole pig with about 5 neighbors (urban area) so the turning things off idea isn't very practical. About the only thing that was on this weekend (I was home alone) was the fridge was plugged in.

I had this problem before when I had an older fridge. I have a newer fridge now and I notice the same issue, so I wouldn't necessarily blame the fridge.

[zigzagflux]

Expect a flattened waveform. IEEE and PSC require the utility provide 5% THD or better distortion at the point of attachment. This means they really aren't going to go to great lengths to ensure you have a clean waveform. And over the past 15 years, I have noticed the THD actually getting worse in residential areas (3%), while in industrial areas it really doesn't look too bad (2%).

I suspect the typical 8320V residential distribution is poorly regulated with high impedance, and the increasing percentage of rectified loads in the home is making the problem worse. And you would be surprised to learn how many of those pole mount transformers are overloaded; at my previous house, we had like 8 homes fed off a tiny little 25 kVA transformer. Think of what temperature that ran at during cooling season! The rep came out and looked at the install (I called in an upgrade request), and he was shocked at how many homes they had on one unit. They replaced it with two 50 kVA transformers.

[AndrewT]

If I want to use a scope to measure/look at a mains waveform inside a piece of live equipment, or any signal that is not referenced to ground.

Do I insert an isolating transformer in the feed to the equipment being tested?
or
insert the isolating transformer in the feed to the scope?

[zigzagflux]

Originally posted by AndrewT
Do I insert an isolating transformer in the feed to the equipment being tested?]

That is an option on a temporary basis, but there is potential to energize other parts of the equipment that were not designed to be such. As soon as you connect your scope probe shield to the equipment, that node becomes effectively grounded.

insert the isolating transformer in the feed to the scope?

Not recommended, though I have done this before. Problem is two-fold; using the isolation transformer may not lift the ground of the scope. The third prong of your plug is typically connected to the probe shield, so unless that is lifted, you still end up grounding the device under test. Second problem is, if you do manage to lift the 3rd prong, you could potentially have an energized scope case. Not smart.

If you meant to apply the transformer in the signal path (between DUT and the scope probe) that would be a valid option.

Best option is either a differential probe (expensive) or use the math function of your scope to probe Ch.1 - Ch.2, effectively creating a differential probe. The probe shields remain at ground potential. Each probe (and scope input) still must be rated for the potential it is applied to, even if the differential signal is small.

[AndrewT]

thanks, I'll need to decide, but looks like isolating the equipment to be tested is the safer option.

I've tried the diff using chi-ch2 but with little that could be called success.