Flatened top sine wave, does it mean anything?

[Besoeker]

Yes, similar without all the ringing that the higher harmonics produce.

It isn't really ringing. It's just that it is, for practical reasons, a finite number of harmonics. I could add more and that would take out some more of the ripple. But is near enough for demonstration purposes.

Where is the 3rd harmonic??

There is no 3rd harmonic content in the line current of a three phase bridge with level DC.
The sequence is 6n?1 where n is the harmonic number.

 

[Besoeker]

I know I am not and I welcome any input from those that are.

Transients and harmonics are somewhat different phenomena.

A variable frequency inverter generally has IGBTs as the output switching device. The switching edge is very fast and usually produces an over voltage transient, and sometimes a considerable over voltage and a high rate of change of voltage (dv/dt).
That phenomenon would happen with just one switching. The transient is usually defined in the time domain.

Harmonics, on the other had, are defined in the frequency domain. For example 5th harmonic is, for you, 300Hz.
They are a means of mathematically describing periodic waveforms, particularly non-sinusoidal by breaking them down into constituents to represent them. A square wave (say ?10V) can be shown to contain a fundamental, 3rd, 5th 7th etc up to infinity in ever decreasing magnitude (reciprocal of the harmonic number)....
This is useful in determining how circuit elements will behave when there is a non-sinusoidal waveform present.

Fourier Series Analysis is the mathematical tool usually used analyse the content of the periodic waveform. In effect, it converts a time domain function to a frequency domain function with usually an infinite number of terms of decreasing magnitude. And significance. We generally don't consider anything above the 40th harmonic.

 

[wirenut1980]

This reminds me of a job in Ohio that one of my co-workers worked on last year. There was a small daycare center that was failing a lot of ballasts. The ballasts could accept a wide range of input voltage, and he did find something strange in the data. Attached is the voltage THD measured. The step changes are very unusual and we tied those step changes to more than one utility capacitor bank on the 12 KV distribution circuit. Waveforms were flat topped, and 3rd and 5th harmonics were the dominant components. This was studied and monitoring was done for about 6 months, but nothing ever conclusively tied the high voltage THD to the ballast failures. Other than the voltage THD nothing else looked abnormal in the data (although monitoring was done at the meter, so something happening inside may not have shown up there). Eventually, the day care center was moved to another circuit due to loading considerations. We have not heard from the day care center in over 6 months.

View attachment Voltage THD step change waveform FFT.pdf View attachment Voltage THD step change.pdf

 

[gar]

120116-1721 EST

iwire:

I will try to provide two definitions as a starting point. These being --- transient --- steady-state.

Transient analysis relates to viewing a circuit on an instantaneous basis vs time. This might be a one time event, a repetitive event that is consistent from one occurrence to another, or a varying event from one time to another.

A one time event might be a static discharge of a charged glass rod.

A repetitive event might be the charge curve of a timing capacitor in a multivibrator oscillator. It could be the capacitor charge, current, or voltage curve in power supply with a rectifier feeding a capacitor input filter, partly the subject of this thread.


Steady-state analysis is an approach to obtain numerical values that describe a circuit where the statistics are stationary. Stationary statistics mean the values of the statistical properties of the circuit do not vary with time. Some of these statistics are average, RMS, peak, probability distribution curve, standard deviation, etc. So in this case the parameter would not vary with time. For example the RMS voltage of a source would be the same now, 5 minutes later, two weeks later, or 50 years ago. The time frame of this consistency would be whatever you chose. A 6 V battery connected to a 6 ohm resistor is always going to have 1 A flow from the battery thru the resistor. A 6 V RMS AC voltage source connected to a 6 ohm resistor has a 1 A AC RMS circuit current.

I can view this AC circuit in a different light and call it a transient analysis. I can study the instantaneous circuit current from time 0 (a voltage zero crossing) to time Pi/2 (the first peak following the zero crossing). Assuming this AC waveform is consistent from one time to another, then if at each positive zero crossing I restart my transient time base, then I have a repetitive transient.


Where you have a steady-state system its possible to describe its operation based on simple single numbers and appropriate simple equations.


In a transient analysis you are interested many values that change with time.

The simple series RC circuit with a switch and constant voltage source is a classic circuit for transient analysis.

Draw a series circuit consisting of a 10 V battery, a switch, a 1 megohm resistor, and a 1 ufd capacitor. Assuming zero charge on the capacitor at time 0, then how does the voltage across the capacitor vary with time after the switch is closed?

I will come back to this later, but this is enough for now. Is the above discussion clear and are there any disagreements about what I have said, or needed clarifications?

.

 

[T.M.Haja Sahib]

Can anyone tell me why these voltage sine waves look odd?

It may be easy to ascertain if the distortion of voltage is caused by load.If possible,switch off the load or tranfer load to generator supply.Now measure the voltage again.If it is same,load is not causing the distortion.Otherwise it is problem with the load.

The voltage problem on the supply side is likely to disappear if the POCO is going to replace the existing transformer with a bigger one.Check up with them, if possible.But if this is not going to happen immediately,you need to install a harmonic filtering equipment.

 

[gar]

120117-0217

T.M.:

The flat top waveforms are not the fault of the local distribution transformer. This flat topping comes from further up in the grid system. In India you probably do not have the very heavy load of rectifiers with capacitor input filters that we have. Our loads heavily consist of computers, CFLs, and CNC machines.

In our local area we have a very reliable grid and power distribution system, but a very heavy industrial load.

.

.

 

[T.M.Haja Sahib]

gar

Your assertion

120117-0217
........This flat topping comes from further up in the grid system.

seems to contradict your another assertion

120117-0217
In our local area we have a very reliable grid and power distribution system,

If you measure welder current and voltage across the welder,you will find flat topped voltage wave across it.It does not mean that voltage wave is bound to be flat topped further upstream to the distribution system............

 

[Besoeker]

The voltage problem on the supply side is likely to disappear if the POCO is going to replace the existing transformer with a bigger one.

If the voltage distortion is on the supply side of the transformer, and I'm with gar that there is a fair probability of that being the case, changing that transformer will not change the distortion in the voltage supplied to it.

But if this is not going to happen immediately,you need to install a harmonic filtering equipment.

And, if the voltage distortion is on the supply side of the transformer, adding harmonics to the secondary of it isn't a fix either.

 

[T.M.Haja Sahib]

If the voltage distortion is on the supply side of the transformer, and I'm with gar that there is a fair probability of that being the case, changing that transformer will not change the distortion in the voltage supplied to it.

You be on my side.......

My logic is simple.If one goes further up the grid,fault level increases,source impedance decreases,the effect of voltage distortion causing loads decreases,the voltage wave resembles ideal sine wave more.Unless there is something wrong with the power generators in the grid,voltage wave is purely sinusoidal at that level.

And, if the voltage distortion is on the supply side of the transformer, adding harmonics to the secondary of it isn't a fix either.

A harmonic filtering equipment will prevent harmonics from damaging capacitors,chokes etc.,which is what the OP requires.........

 

[Besoeker]

My logic is simple.

But irrelevant to the claim you were made in post #45.
Changing the transformer fed by the distorted voltage will not that distorted voltage.

A harmonic filtering equipment will prevent harmonics from damaging capacitors,chokes etc.,which is what the OP requires.........

If the harmonics are on the supply to the transformer, harmonic filtering after it won't fix the problem.

 

[T.M.Haja Sahib]

But irrelevant to the claim you were made in post #45.
Changing the transformer fed by the distorted voltage will not that distorted voltage.

As I already stated in post #47

If you measure welder current and voltage across the welder,you will find flat topped voltage wave across it.It does not mean that voltage wave is bound to be flat topped further upstream to the distribution system............

Load on the consumer side can distort the supply voltage with a given transformer size.But if the transformer size is increased,this voltage distortion effect of the load diminishes.

If the harmonics are on the supply to the transformer, harmonic filtering after it won't fix the problem.

This is false,because for example,switching of HV capacitor bank can affect the downstream low voltage capacitor bank installed on the consumer side due to distortion of voltage wave. A harmonic filter can prevent this damage to the LV capacitors.

 

[Besoeker]

As I already stated in post #47

Stick to the point.
This your post that I was addressing:

The voltage problem on the supply side is likely to disappear if the POCO is going to replace the existing transformer with a bigger one

It won't if the distortion is on the supply side.

This is false,because for example,switching of HV capacitor bank can affect the downstream low voltage capacitor bank installed on the consumer side due to distortion of voltage wave. A harmonic filter can prevent this damage to the LV capacitors.

Clearly you don't understand the difference between a transient and harmonics.
Past experience has shown that I'd be wasting my time offering you an explanation so I won't. But you might care to read post #42.

 

[T.M.Haja Sahib]

As I already stated in post #47

Stick to the point.
This your post that I was addressing:


It won't if the distortion is on the supply side.

If the voltage distortion on the supply side is due to load side,something could be done about it,as I already stated.

Clearly you don't understand the difference between a transient and harmonics.
Past experience has shown that I'd be wasting my time offering you an explanation so I won't. But you might care to read post #42.

Flat topped harmonics are not transients,but peaked harmonics could be...........

 

[gar]

120117-0908 EST

T.M.:

I can turn everything in my house off, my neighbor is gone so he has no nonlinear loads on our common pole transformer, the pole transformer is 50 kVA, and I see a flat topped voltage waveform. This clearly comes from upstream in the grid.

When you take a large number of small loads that are nonlinear, produce large current spikes at the voltage peak, and add these up, then there is a big effect on the grid.

To analyze the electrical current waveshape to a nonlinear load on a sine wave voltage source one would use transient analysis within a half cycle period of the excitation signal. Further one can do a transient analysis of the effect of this current on the voltage source including the internal impedance of the source to see what the supplied terminal voltage to the load will look like.

.

 

[Besoeker]

If the voltage distortion on the supply side is due to load side,something could be done about it,as I already stated.

Nice attempt at a sideways shuffle.
This is what you stated:

The voltage problem on the supply side is likely to disappear if the POCO is going to replace the existing transformer with a bigger one.

It's not correct.
If you now want to introduce load side harmonics to bolster your incorrect case for resolving the distortion on the supply side by making the transformer bigger - it doesn't.
Still not correct.

Flat topped harmonics are not transients,but peaked harmonics could be...........

The harmonics are neither flat topped nor peaked. They are sinusoidal.

 

[Besoeker]

120117-0908 EST

T.M.:

I can turn everything in my house off, my neighbor is gone so he has no nonlinear loads on our common pole transformer, the pole transformer is 50 kVA, and I see a flat topped voltage waveform. This clearly comes from upstream in the grid.

When you take a large number of small loads that are nonlinear, produce large current spikes at the voltage peak, and add these up, then there is a big effect on the grid.

To analyze the electrical current waveshape to a nonlinear load on a sine wave voltage source one would use transient analysis within a half cycle period of the excitation signal. Further one can do a transient analysis of the effect of this current on the voltage source including the internal impedance of the source to see what the supplied terminal voltage to the load will look like.

Gar, as ever I mostly agree with what you say.
If I want to analyse a non-linear waveform I use Fourier series analysis.
If I want to estimate the effect of a transient I often do a bit by bit calculation rather than just the analytical approach. That would give me a single output value but if I change any component value, I have to do it again.

 

[Besoeker]

And back to the OP.

Here's a not entirely perfect reconstruction of iwire's waveform just using some mathematics:

But not bad for a first attempt,

 

[T.M.Haja Sahib]

Feedback from OP

Feedback from OP

The OP posed a problem and I offered solution(s) in post #45.I do not know whether the solution(s) of any use to the OP or not i.e there is no feedback from the OP.

 

[Joethemechanic]

I'm not saying a bigger (in KVA) transformer, but a transformer with more weight in laminations would move the current to the right hand side of that sine wave and tend to push the sign back into shape.

I really think part of the problem we are seeing has to do with less industry meaning less motors, and a large part of our motors are being run from VFDs

Not to mention more and more underground conductors and the capacitance between phases contributing to a leading PF problem distorting the sign on the left of the peak

 

[Besoeker]

The OP posed a problem

Not so. "Tell me why" requests an explanation, not a solution.
And explanations were provided in a number posts.

and I offered solution(s) in post #45.I do not know whether the solution(s) of any use to the OP or not.

Solutions they are not.