AC Line Waveform Distortion

[spiff72]

Hello all,

I used my Rigol 1054Z to measure the AC waveform in my house and I get a sort of shark-fix shape.  I attempted to use the FFT function on the waveform and ended up with the image shown.  Can someone confirm whether my conclusions are correct?  I used one of the other options under the FFT (Hanning?) when I took this image.

• The peak in the middle is the fundamental 60 Hz signal (I chose 60 Hz in the math menu)
• The following peaks are the 2nd harmonic (120Hz), 4th harmonic (240 Hz), and 6th harmonic (360 Hz - this peak is tiny)
• If the above is correct, then the graticules across are 100 Hz each RELATIVE to the fundamental, so the first graticle is at 100, not 100+60 (180 Hz)?
• The 2nd and 4th harmonic peaks are both about 27db down from the fundamental? 

What sorts of factors can distort the waveform like this?

[IanB]

Based on a similar question I asked in the past, I'd say that it is not unusual to see the mains wave form distorted like this.

Probably the mains is a nice sine wave when it comes out of the generator, and it gets more and more distorted the closer it gets to the point of use (basically because many consumer loads are non-linear and they influence it more the closer you get to them).

https://www.eevblog.com/forum/chat/when-is-a-sine-wave-not-a-sine-wave/

[spiff72]

Thanks.  I have a coworker with the same scope and he stated that when he does the same evaluation at home, he gets a very clean looking sine wave.  Obviously everyone's homes will be different, though.

Also, I think I need to repeat this test with the FFT again after seeing a few more tutorials.

[JustMeHere]

How does your probe look on the test point?  Have you compensated it?  You can actually pick up the 60hz wave by probing your finger and pressing auto.

[gf]

Looks like the mains line is not loaded with ohmic load, but heaviliy loaded with rectifiers (e.g. SMPS without or with poor power factor correction). One factor is likely that more and more people replace tungsten with led (and the power factor is certainly not the primary buying criterion for lamps or bulbs, but rather the price, or design of the lamp).

[soldar]

All mains power I have ever seen are clipped at the top due to rectifier-capacitor loads. This is normal.

[SG-1]

My mains waveform is also flat on the top.
Here is an example from an experiment with a mercury vapor lamp.

Voltage trace is red.

[rstofer]

Hello all,

I used my Rigol 1054Z to measure the AC waveform in my house and I get a sort of shark-fix shape.  I attempted to use the FFT function on the waveform and ended up with the image shown.  Can someone confirm whether my conclusions are correct?

I'm pretty sure the FFT horizontal scale is just what it says it is, 100 Hz/div.  The fundamental is centered on a division (according to the menu item) and the 3rd harmonic is 120 Hz above 60 Hz or 180 Hz.  I expect 3rd and 5th harmonics from loads that are essentially a square wave.

You opted to have the fundamental positioned 40 dB down from the top of the graticule.  I can't see the lines well enough to say anything about the two harmonics but just count them down from the top of the fundamental.

[MrAl]

Hello all,

I used my Rigol 1054Z to measure the AC waveform in my house and I get a sort of shark-fix shape.  I attempted to use the FFT function on the waveform and ended up with the image shown.  Can someone confirm whether my conclusions are correct?  I used one of the other options under the FFT (Hanning?) when I took this image.

• The peak in the middle is the fundamental 60 Hz signal (I chose 60 Hz in the math menu)
• The following peaks are the 2nd harmonic (120Hz), 4th harmonic (240 Hz), and 6th harmonic (360 Hz - this peak is tiny)
• If the above is correct, then the graticules across are 100 Hz each RELATIVE to the fundamental, so the first graticle is at 100, not 100+60 (180 Hz)?
• The 2nd and 4th harmonic peaks are both about 27db down from the fundamental? 

What sorts of factors can distort the waveform like this?

Hi,

That's what mine looks like too.
I can guess that the line is loaded more near the peaks that's why it seems so flat topped.
Rectifiers do that and a lot of things we use today use bridge rectifiers.

[rstofer]

Sure, rectifiers do that because of capacitors.  In every linear supply, there is a rectifier feeding a capacitor.  We like to use large capacitors so that the voltage doesn't change very much throughout the conduction cycle.  In fact, the rectifier only conducts for a very brief period of time when its voltage is higher than the drooping capacitor voltage.  The smaller this conduction period, the higher the current needs to be during conduction.  Big capacitors, small conduction angle, high current.

Attached plot of diode conduction shows that the diode is only conducting at the very top left edge of the sine wave.  Before the diode can conduct, the diode voltage (blue trace) has to get higher than the capacitor voltage (black trace) which is drooping.  The diode conducts until the capacitor is recharged (at the crest of the sine wave) and then it  stops conducting.  You can see that the conduction angle is quite small yet all of the load current for the entire cycle needs to pass through the diode in that small amount of time in order to charge the capacitor.

[spiff72]

Thanks for all the responses.  I do have a lot of LED bulbs scattered around the house.  I might have to repeat the experiment with all the lights in the house turned off and see if that makes a difference.  I did actually try turning lights off in just my office (where the scope is located) and it didn't make a difference.

Also, thanks to rstofer for the clarification of the FFT numbers.  It does make more sense that the peaks are the 3rd and 5th and my logic was flawed on the counting of the divisions.

Thanks!

[joeqsmith]

If I want to run a test with a clean wave, I have to generate it.   Link showing the incoming voltage at our house.   

https://youtu.be/04I7nHA_HxM?t=731

[radiolistener]

The same waveform on my mains  
I think this is because of switching power supplies in the computer and other home electronics.

[bdunham7]

How is the scope connected to the mains?  Direct, voltage divider, transformer?

3rd, 5th and 7th harmonics are common.  3rd would be mostly generated locally, typically an unloaded transformer will generate some 3rd due to hysteresis.  However, even though industrial loads and distribution systems do generate a lot of 3rd, you don't see them much because in a 3-phase system, they cancel out.

5th is generated by induction motors and certain generator configurations.

What you have looks like an unloaded transformer to me, although rectifier/capacitor systems could be contributing as well.  THD is supposed to be under 5% and it looks like you may be approaching that.

B/T/W, that is a nice clean FFT, especially as a freebie add-on to an entry level scope.  What exact settings (window type, etc) did you use?

[radiolistener]

In my case, I connected mains to oscilloscope through toroidal transformer. Secondary coil is unloaded.
But when I load it with resistor, the waveform shape is the same, just a little smaller amplitude.

[soldar]

I might have to repeat the experiment with all the lights in the house turned off and see if that makes a difference.

Probably not. You'd have to turn off a lot more than just lights and just your house. There are many electronic loads like that and they are all over the power grid. You might see a bit of difference but I doubt it. You would probably see the same shape even at the distribution transformer.

[bdunham7]

In my case, I connected mains to oscilloscope through toroidal transformer. Secondary coil is unloaded.
But when I load it with resistor, the waveform shape is the same, just a little smaller amplitude.

I'm guessing your 3rd harmonic is mostly coming from the toroid.  I know mains voltage freaks people out, but there's no reason you can't just poke your mains with the probes that are included with the Rigol as long as you set it for 10X and DON'T USE THE GROUND LEAD!  They're 300V CATII.  If you're nervous, just open up a device that has a polarized plug and uses a fairly small protection fuse, then probe downstream of the fuse. 

[soldar]

Another really easy thing to do is to use a voltage divider. You can use 100K + 10K, connect the 10 K to neutral and the 100K to live and you divide the voltage but also have a current limiter so that even if you short the leads you are safe.

[BravoV]

Never seen a perfect Sine wave shape from mains, mine is similar to OP too, using Fluke DP-120 HV Diff.Probe straight at mains that powered the scope it self.

[Gyro]

I'm guessing your 3rd harmonic is mostly coming from the toroid.  I know mains voltage freaks people out, but there's no reason you can't just poke your mains with the probes that are included with the Rigol as long as you set it for 10X and DON'T USE THE GROUND LEAD!  They're 300V CATII.  If you're nervous, just open up a device that has a polarized plug and uses a fairly small protection fuse, then probe downstream of the fuse.

Another really easy thing to do is to use a voltage divider. You can use 100K + 10K, connect the 10 K to neutral and the 100K to live and you divide the voltage but also have a current limiter so that even if you short the leads you are safe.

Guys, this is the Beginners section. We never suggest connecting scope probes to, or directly tapping the the mains, especially when the experience level of OP is unknown! 

Also, many Chinese scope probes are Cat I rated at 300V (even if they claim otherwise).

[bdunham7]

Also, many Chinese scope probes are Cat I rated at 300V (even if they claim otherwise).

Is there evidence of this?  I seriously doubt Rigol has faked their CAT rating, and even if they have, then they aren't CAT rated at all, not CAT I.

I'll respect the rules, if that's what they are, but I think you are doing even beginners a big disservice in telling them to avoid a very simple, common measurement technique instead of telling them how to do it safely. 

[IanB]

I'll respect the rules, if that's what they are, but I think you are doing even beginners a big disservice in telling them to avoid a very simple, common measurement technique instead of telling them how to do it safely.

I think it's important to point out the risks of trying to measure mains with oscilloscopes. Bench scopes generally do not have isolated and floating inputs. I have created unfortunate sparks and overloads when working with low voltages when I was careless. Doing the same with the mains could have had much more severe consequences.

[Gyro]

Also, many Chinese scope probes are Cat I rated at 300V (even if they claim otherwise).

Is there evidence of this?  I seriously doubt Rigol has faked their CAT rating, and even if they have, then they aren't CAT rated at all, not CAT I.

I'll respect the rules, if that's what they are, but I think you are doing even beginners a big disservice in telling them to avoid a very simple, common measurement technique instead of telling them how to do it safely.

Ok, I'll concede that original Rigol probes do most likely meet their CAT rating, although as I indicated, low cost Chinese probes are not all created equal. If the probe is accidentally set to it's X1 position, the safety rating immediately reverts to the CAT I rating of the scope input. Stated CAT rating of scope probes is therefore 'suspect'.

You do still need to find a way of connecting the probe to an outlet (without exposing the ground skirt*). Mains outlets don't naturally lend themselves to direct probing with scope probes, this inevitably leads to dodgy hookups of one sort or another, either objects inserted, or leads with a plug on one end. You also can't assume that the user has suitably tied back and insulated, or removed the dangling ground clip. There are simply too many variables to assure inexperienced user safety.

In the Beginners section we do not assume a level of knowledge. There have certainly been cases of beginners (of all ages) receiving their shiny new scopes and having exhausted the Probe cal output and in the absence of other signal sources, go straight to probing the mains (not suggesting this is the case with the OP - in fact he was playing safe and using a transformer). This is the reason Dave produced his comprehensive and oft quoted 'How not to blow up your scope' video. It is too common an occurrence. As Ian says getting it wrong can have nasty consequences.

Not criticizing, just explaining.

* Edit: Image captured from an ebay listing ( ) a few years back: https://www.ebay.co.uk/itm/Professional-BNC-60MHz-Oscilloscope-Probe-Clip-P6060-X1-X10-600V-/161805210779?hash=item25ac57949b:g:An8AAOSwgQ9V3JuV

[The Electrician]

The OP did not say that he used a transformer; it was radiolistener that said he used a transformer.

Tell us, spiff72, did you use a transformer, or did you probe the mains directly with a standard oscilloscope probe?

[spiff72]

Sorry for the lengthy delay in response.  I didn't have notifications enabled until just now, and didn't think to come back and check this thread.

My method to measure in my OP was to remove the ground lead from the probe (not the scope power cord) and then measured directly.  I know that the outlets are wired correctly, and i have a plug and short wire length with the correct indicators for hot, neutral, and ground. The scope was set to use AC as the trigger.

I am aware of the risks, and was comfortable with them.

How is the scope connected to the mains?  Direct, voltage divider, transformer?

3rd, 5th and 7th harmonics are common.  3rd would be mostly generated locally, typically an unloaded transformer will generate some 3rd due to hysteresis.  However, even though industrial loads and distribution systems do generate a lot of 3rd, you don't see them much because in a 3-phase system, they cancel out.

5th is generated by induction motors and certain generator configurations.

What you have looks like an unloaded transformer to me, although rectifier/capacitor systems could be contributing as well.  THD is supposed to be under 5% and it looks like you may be approaching that.

B/T/W, that is a nice clean FFT, especially as a freebie add-on to an entry level scope.  What exact settings (window type, etc) did you use?

Thanks for that feedback.  I will have to redo the measurement re-check the settings I was using on the FFT - this was over a month ago so I don't remember them.  I also remember doing a fair amount of fiddling with the settings trying to make sense of it.

I might have to go on a search for unloaded supplies.  Would this be only for linear supplies (both the AC/DC and AC stepdown), and would unloaded SMPS be included too?  I am sure that there are a TON of scattered SMPS (wall warts, phone chargers, etc), but probably not as many linear supplies, and even fewer purely stepdown transformers.  The only stepdown transformer I am aware of would be for the furnace/thermostat - and maybe the doorbell.

There are probably enough options to just shut off breakers until it is eliminated (assuming the source is in the house).

EDIT:  I now realize that maybe the question about FFT settings was directed at Radiolistener (I would like to know how he got that plot as well, and what scope he used).