• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Testing AC Line Quality with an Oscilloscope - Generator power quality

Without reading through the entire thread, this looks like the lack of a ground. If you are using differential input to the scope, the source still has to have some kind of a ground reference or the noise will get though.

Tie the safety ground and the neutral together and connect the scope probe ground clip to that point. Then connect the scope probe tip the the hot output. Look at it singled ended. Worst that could happen is that you could blow a fuse/pop a circuit breaker in the generator/inverter/wall outlet.
Thanks.
You and reloadRon both suggest the same thing so I'm sure that's it.
Your patience is appreciated.
 
Finally got a decent SineWave thanks to ReloadRon and many of you kind enough and patient enough to guide me.
Thank you!

Probe Ground directly connected to Neutral
Probe Tip directly connected to Hot

Here is the much better sine wave I finally got. The settings puzzle me a bit however....
VARIABLE Volts/Div setting at 2
TIME/DIV setting at 5ms
PROBE set at 10X

OScope_Sinewave_1A_20190901_140437.jpg
 

Reloadron

Well-Known Member
Most Helpful Member
2V/div with a 1:1 probe. You have a 10:1 probe so it is 20V/div. Next is the red know turned to "cal". I think not.
You may want to make sure of that.

Also you want to make sure both your A and B Time/Div knobs are set the same as it looks like you have an intensified trace. The inner and outer knobs should be aligned so they are indexed together.

Ron
 

JimB

Super Moderator
Most Helpful Member
it looks like you have an intensified trace.
Are you sure that is not a camera shutter effect?
You can get some odd things when taking pictures of a CRT screen.
Look at this:
 

Reloadron

Well-Known Member
Most Helpful Member
Are you sure that is not a camera shutter effect?
You can get some odd things when taking pictures of a CRT screen.
Look at this:
While I would not rule that out the thread starter has posted several images and if I see things correctly they all look to have an intensified portion of the sweep. With that in mind and how pronounced the intensified portion of the sweep looks I just tended to see it as a result of the A & B sweep speed settings. Purely a guess on my part. Also the A & B sweep time per division do not seem to be indexed the same. Beats me? :)


I also can't seem to get the link you were kind enough to post to work.


Ron
 

JimB

Super Moderator
Most Helpful Member
Also you want to make sure both your A and B Time/Div knobs are set the same
The selector switch below the TIME/DIV controls is clearly set to A,
so the OP is using just the A timebase which would be the correct thing to do.

it looks like you have an intensified trace.
You can get some odd things when taking pictures of a CRT screen.
Look at this:
View attachment Marconi Spectrum Analyser.JPG

The multiplexed frequency display as well, yuk!

JimB
 

JimB

Super Moderator
Most Helpful Member
I also can't seem to get the link you were kind enough to post to work.
I seem to be having a bad day with attachments, I have repeated the whole post.

JimB
 

Reloadron

Well-Known Member
Most Helpful Member
I seem to be having a bad day with attachments, I have repeated the whole post.

JimB

:) There will be days like that, been there and done that.

My thinking and I could well be wrong is while the sweep setting is set to "A" I am not sure with each knob set differently what the effect would be. I have done some fast shutter speeds and caught retrace between sweeps on my old analog scopes. What drew my attention is the effect shows up in several images posted so I figured the odds were against it being an effect in the image frame. The intensity difference in the sweep was just so well pronounced. Really don't know Jim but if the original poster returns maybe we will know? Beats me? :)

Yeah, the link now works on the bright side. :)

Ron
 

JimB

Super Moderator
Most Helpful Member
Earlier in this thread The Electrician did some tests to ascertain the usefulness of a "filament transformer" as a means of making isolated measurements of mains supply waveforms, with particular reference to harmonic distortion.

His tests are here:

His results were good, this lead to the conclusion that a simple mains transformer had a better frequency response than one may expect.

As a follow up to that, I did a few experiments on two transformers that I had in my "junk box".
One transformer was a 240v to 12v 3VA type, the other was a 240v to 9v 12VA type.
Both transformers are of the split bobbin type with the primary and secondary isolated from each other, rather than the secondary being wound over the primary as was usually the case with older transformers.

I used my function generator to inject 7.5v into the primary of each transformer, and used the oscilloscope to measure the voltages at the primary and secondary terminals of the transformer as I varied the frequency of the function generator.

I put the results into a spreadsheet and created a graph of the voltage transformation ratio of the two transformers.

The graph:
120292

Note that a higher transformation ratio indicates a lower voltage at the secondary of the transformer.

The 12VA transformer shows a reasonably flat response up to about 2kHz, the voltage then drops off slowly as the frequency is increase to 100kHz.
Above 200kHz the voltage drops off quire rapidly.

The smaller 3VA transformer has a reasonably flat response up to 40 to 50 kHz and then rises, falls and rises again as the frequency is raised to 500kHz, which was the limit of my test.
Further investigation showed that the 3VA transformer exhibited resonances at 120 kHz and 533kHz.

JimB

(Please excuse the thread drift).
 

ronsimpson

Well-Known Member
Most Helpful Member
Good work. The blue 12VA transformer is easy to flatten out with simple RCs. I think the 100khz point can be brought back to flat. At 100 to 200khz I might give up and attenuate hard. I need to make one or two.
 
Are you sure that is not a camera shutter effect?
You can get some odd things when taking pictures of a CRT screen.
Look at this:
Interesting that you mentioned this.
I've been struggling with getting photos of the waveforms.
I've tried 1/2 second shutter speeds and it's better but still not perfect. I have to use a tripod to hold the camera steady.
Sometimes I have to take 6 or 7 shots to get a good one.

I've even asked for help on this as some photography specific forums. (no answers yet)

Is there a "trick" to taking good snapshots of the waveforms?

Here's one I took of the home Utility AC today using the slower shutter speed. Better but not great.
Needs to be sharper and better contrast. The background is too bright imo. The waveform needs to stand out better.
May try a lower ISO number tomorrow but it's harder to get clarity at lower ISO settings.
HOME-AC_Waveform_09-02-2019_2_IMG_20190902_154727.jpg

The waveform from my Inverter Generators (both) produce a MUCH closer to pure sine waveform than the House AC above. The House AC seems to have slightly flattened and clipped peaks
 
Last edited:
Earlier in this thread The Electrician did some tests to ascertain the usefulness of a "filament transformer" as a means of making isolated measurements of mains supply waveforms, with particular reference to harmonic distortion.

His tests are here:

His results were good, this lead to the conclusion that a simple mains transformer had a better frequency response than one may expect.

As a follow up to that, I did a few experiments on two transformers that I had in my "junk box".
One transformer was a 240v to 12v 3VA type, the other was a 240v to 9v 12VA type.
Both transformers are of the split bobbin type with the primary and secondary isolated from each other, rather than the secondary being wound over the primary as was usually the case with older transformers.

I used my function generator to inject 7.5v into the primary of each transformer, and used the oscilloscope to measure the voltages at the primary and secondary terminals of the transformer as I varied the frequency of the function generator.

I put the results into a spreadsheet and created a graph of the voltage transformation ratio of the two transformers.

The graph:
View attachment 120292

Note that a higher transformation ratio indicates a lower voltage at the secondary of the transformer.

The 12VA transformer shows a reasonably flat response up to about 2kHz, the voltage then drops off slowly as the frequency is increase to 100kHz.
Above 200kHz the voltage drops off quire rapidly.

The smaller 3VA transformer has a reasonably flat response up to 40 to 50 kHz and then rises, falls and rises again as the frequency is raised to 500kHz, which was the limit of my test.
Further investigation showed that the 3VA transformer exhibited resonances at 120 kHz and 533kHz.

JimB

(Please excuse the thread drift).
I'm again a bit confused......
Is this to say that you will get somehow, "better" readings or more accurate waveforms by using a transformer than connecting the Probe's ground to Neutral and Probe tip to Hot ?

I have a transformer (117v to 25v) but I haven't used it yet.
If safety is the only concern, I've built a rig that has a nice AC wall plug on one end and the ground and probe tip safely connected and insulagted to the other end so electric shock is highly unlikely.
 

ronsimpson

Well-Known Member
Most Helpful Member
I'm again a bit confused......
We (more or less (+/-)) have more than one thing going. I think for your job, do what is working. Some of us work on power line applications where we could use a poor man's dif-probe where the transformer will work. My applications often have me looking at a signal where neither end is connected to neutral or ground. I have put micro computers on the hot line. You can't connect a scope ground to that type of computer. But the transformer will give isolation and differential.
 
Just for the benefit of anyone who finds this thread later.

The method suggested by ReloadRon, RonSimpson and others allowed me to get the sine waveforms I needed.

I used the same setup for all my generators AND the House AC current.

ONE Channel
PROBE set to 10x
GROUND clip from probe connected to Neutral
PROBE TIP connected to Hot

Don't zap yourself. :)
 

Nigel Goodwin

Super Moderator
Most Helpful Member
Just for the benefit of anyone who finds this thread later.

The method suggested by ReloadRon, RonSimpson and others allowed me to get the sine waveforms I needed.

I used the same setup for all my generators AND the House AC current.

ONE Channel
PROBE set to 10x
GROUND clip from probe connected to Neutral
PROBE TIP connected to Hot
Also for the benefit of anyone reading this thread, that's a seriously BAD idea.

Assuming the scope is earthed, as it should be, such a connection would immediately blow the GFCI/RCD (earth leakage trip) due to the voltage difference between Neutral and Earth.

It 'should' work OK in the scope of this thread, with a non-earthed generator, but not with a normally wired house.
 

Reloadron

Well-Known Member
Most Helpful Member
Are you sure that is not a camera shutter effect?
You can get some odd things when taking pictures of a CRT screen.
Look at this:
Jim
I have taken dozens of screen shots using my camera and never saw the effect of shutter speed. So today I decided to actually use camera settings other than Auto. Oh well and just plain wow.
Scope Intensified.png

The waveform is 60 Hz and I am not sure what my shutter speed was but apparently a lot faster than the 50 mSec it takes the beam (and then a little) to cross the CRT. I never would have thought the image would be so well defined. I shot subsequent images using slow shutter speeds but not so slow I needed to support the camera.

Ron
 

Reloadron

Well-Known Member
Most Helpful Member
Moving on and in the interest of safety. I suggested two methods of observing the waveform(s) in question. I did so in this as well as another forum. Unless someone is in their comfort zone working with AC mains, be it 120 or 240 VAC RMS at 50 or 60 Hertz Then I suggest against it. I also mentioned a 120 VAC 60 Hz waveform RMS will be 169.68 Volts peak and as displayed on the scope 339.36 volts peak to peak. Using a 10:1 scope probe within its limits and the vertical gain set for 5 V/Div that allows 50 V/Div. The typical scope and the thread starter's scope allows 8 vertical divisions on the display.

The thread began as measuring the quality of a generator's power. I believe back there somewhere I pointed out that the scope, in this case, should not be plugged into the power source (generator) it is measuring. I also pointed out how scope ground works and that on household AC mains Since on US Power mains supplies the earth ground and neutral are bonded at the service entry. Needless to say that the outer shells of the BNC vertical input channels are at ground potential so if the vertical input probes grounds are connected to the 120 VAC hot side there will be a large flash and a circuit breaker will likely trip. All of this has been previously covered and if while using the scope the user is in anyway grounded and makes bodily contact with AC mains hot a shock will result.

Using a differential method was covered several times and the reasons were stressed.

The thread began with measuring the quality of a generator output and in such a case the generator would have been an isolated source. Additionally "quality" was never defined as to what was expected to be seen? Distortion? Distortion as Harmonics, Hum and Noise? If THD (Total Harmonic Distortion) is what is wanted to be measured then a scope is a poor choice but usable.

How to measure powerline quality with an oscilloscope is an OK read on the subject. A good newer scope which offers FFT (Fast Fourier Transform) as aprt of the math functions are a good start but everyone has that. About all that matters in the case of a generator is any deficiency in power quality doews not come in light of the power's intended use.

The reason I would choose not to use a transformer is in my own opinion the generator under test should also be tested under a load and I also feel adding a transformer may not afford a true look at what the actual generator output is. That being just my thinking. Now if someone would like to explain in detail how to measure a generator's power quality please do explain it.

Ron
 
Also for the benefit of anyone reading this thread, that's a seriously BAD idea.

Assuming the scope is earthed, as it should be, such a connection would immediately blow the GFCI/RCD (earth leakage trip) due to the voltage difference between Neutral and Earth.

It 'should' work OK in the scope of this thread, with a non-earthed generator, but not with a normally wired house.
Ok thanks for that Nigel.
But tell us, what exactly does it mean if it works just as well for the House AC as it does for the generators (since it did, no problem)?
why did I not have the problem you mention?

I understand you are saying that the house AC should have Earth bound to Neutral? (And yes, the Oscilloscoipe has a three prong, grounded plug)
Perhaps my 1982 home does not due to older electrical code?

Or worse, could there be a problem with my homes wiring?
 
Last edited:
Ron,
for my purposes I found exactly what I was hoping to find.

The waveform from two of my Inverter generators appeared on the oscilloscope as perfect sine waves. No distortions.
The waveform of the House AC was nearly perfect, but demonstrated slightly flattened peaks and a small distortion.
The waveform from the NON inverter generator was also as expected. Noisy, but still a fairly good sine wave otherwise. it was not a squared or blocky waveform at all as I expected.

I was not trying to get scientific, exact measurements of THD for example. That's above my pay grade.
Just a general visual idea of what the waveforms looked like which I tend to believe is a fairly good indicator of the general quality of the power.
I could be wrong about that.
 

Latest threads

EE World Online Articles

Loading

 
Top