# Why do Current Transformer cores have different knee point voltages?

#### Manoj Sahu

I have to test a Current transformer. I read knee point is a point at which 10% increase in voltage leads to 50% increase in current. As there are different classes of current transformer (cores), how it is decided what/how much will be the knee point voltage of that core.

Related Electrical Engineering News on Phys.org

#### Baluncore

Science Advisor
I have to test a Current transformer.
Then you will need to know the specifications of that current transformer. What accuracy must it meet, over what range of line current and frequency. The secondary load specification will be an important parameter. You must test those characteristics rather than design a new transformer core.

I read knee point is a point at which 10% increase in voltage leads to 50% increase in current.
That knee specification of the core is applicable to inductors and voltage transformers, but it is not directly applicable to current transformers.

A current transformer operates with an independent primary current. Part of the primary current produces the magnetising flux for the core while the remaining primary current must be balanced by the secondary ampere turns. It is therefore critical that the secondary load be specified and controlled.

The selection of core material for a current transformer requires the magnetising current that generates the flux be kept small compared to the current being measured. See; Chap 16, Transformer and Inductor Design Handbook, 2011, Colonel Wm T McLyman.
https://coefs.uncc.edu/mnoras/files/2013/03/Transformer-and-Inductor-Design-Handbook_Chapter_16.pdf

See also; https://www.eiseverywhere.com/file_uploads/1b3d32ecaf287ed7ce68ae5799836a96_CTTestingTheory.pdf

Last edited:

#### Manoj Sahu

Then you will need to know the specifications of that current transformer. What accuracy must it meet, over what range of line current and frequency. The secondary load specification will be an important parameter. You must test those characteristics rather than design a new transformer core.

That knee specification of the core is applicable to inductors and voltage transformers, but it is not directly applicable to current transformers.

A current transformer operates with an independent primary current. Part of the primary current produces the magnetising flux for the core while the remaining primary current must be balanced by the secondary ampere turns. It is therefore critical that the secondary load be specified and controlled.

The selection of core material for a current transformer requires the magnetising current that generates the flux be kept small compared to the current being measured. See; Chap 16, Transformer and Inductor Design Handbook, 2011, Colonel Wm T McLyman.
https://coefs.uncc.edu/mnoras/files/2013/03/Transformer-and-Inductor-Design-Handbook_Chapter_16.pdf

See also; https://www.eiseverywhere.com/file_uploads/1b3d32ecaf287ed7ce68ae5799836a96_CTTestingTheory.pdf
Thank you

#### jim hardy

Science Advisor
Gold Member
2018 Award
Dearly Missed
how it is decided what/how much will be the knee point voltage of that core.
I sense possible confusion on my part as to the wording of your question.
Did you really mean "decided" ? as by the transformer designer ?
or
Measured ? as by a student in an EE lab class ?

That's two different questions.

The student in a lab class should think for a minute about the current transformer .
It's just two windings on a core, and that's no different from a voltage transformer
so the current transformer will transform voltage just as happily as it transforms current.

That said,
The fundamental difference between a Voltage transformer and a Current transformer is how we use it.
With a current transformer we control primary current not voltage,
whereas with a voltage transformer we control primary voltage and let primary current fall where it pleases. Unloaded, that's its magnetizing current.

So to me, a transformer is a transformer
A current transformer is just a transformer with an excellent core that requires very little magnetizing current.

So, a student can apply small voltage, measure current that results, and plot current versus volts on graph paper.
That's called an "Excitation Curve" for the transformer irrespective of how it will be used.
The knee will be readily apparent on the excitation curve. .
The student should be very careful in applying voltage to not go very much beyond the knee
because if he "saturates" the core he can ruin the insulation and wreck the transformer.

A designer would decide what accuracy he needs and follow the process in that most excellent link provided by @Baluncore .

old jim

Last edited:

#### Manoj Sahu

Thanks Jim
I meant calculated/decided by the designer.
If a current transformer is saturated, does it act like permanent magnet.
I sense possible confusion on my part as to the wording of your question.
Did you really mean "decided" ? as by the transformer designer ?
or
Measured ? as by a student in an EE lab class ?

That's two different questions.

The student in a lab class should think for a minute about the current transformer .
It's just two windings on a core, and that's no different from a voltage transformer
so the current transformer will transform voltage just as happily as it transforms current.

That said,
The fundamental difference between a Voltage transformer and a Current transformer is how we use it.
With a current transformer we control primary current not voltage,
whereas with a voltage transformer we control primary voltage and let primary current fall where it pleases. Unloaded, that's its magnetizing current.

So to me, a transformer is a transformer
A current transformer is just a transformer with an excellent core that requires very little magnetizing current.

So, a student can apply small voltage, measure current that results, and plot current versus volts on graph paper.
That's called an "Excitation Curve" for the transformer irrespective of how it will be used.
The knee will be readily apparent on the excitation curve. .
The student should be very careful in applying voltage to not go very much beyond the knee
because if he "saturates" the core he can ruin the insulation and wreck the transformer.

A designer would decide what accuracy he needs and follow the process in that most excellent link provided by @Baluncore .

old jim

#### jim hardy

Science Advisor
Gold Member
2018 Award
Dearly Missed
I meant calculated/decided by the designer.
Thanks for clarifying ! As i suspected the misunderstanding was mine .

If a current transformer is saturated, does it act like permanent magnet.
Not necessarily, though it can become magnetized (look up remanence) which leaves it some distance up the B-H curve and will make it inaccurate.
If that happens, one should demagnetize it by applying gradually decreasing AC current to it.

The danger of driving a current transformer into saturation with AC current is that flux will change from maximum positive to maximum negative over just a fraction of a line cycle;
meaning $\frac{Δflux}{Δtime}$ is very large
and such a rapid rate of change of flux induces far larger than normal voltage in the windings;
which might pierce the insulation,
in addition to leaving it magnetized

old jim

#### Manoj Sahu

Thanks for clarifying ! As i suspected the misunderstanding was mine .

Not necessarily, though it can become magnetized (look up remanence) which leaves it some distance up the B-H curve and will make it inaccurate.
If that happens, one should demagnetize it by applying gradually decreasing AC current to it.

The danger of driving a current transformer into saturation with AC current is that flux will change from maximum positive to maximum negative over just a fraction of a line cycle;
meaning $\frac{Δflux}{Δtime}$ is very large
and such a rapid rate of change of flux induces far larger than normal voltage in the windings;
which might pierce the insulation,
in addition to leaving it magnetized

old jim
While doing the knee point test, I injected voltage upto 1500 V in a current transformer having knee point voltage 750V @ 30mA. The test was performed using Omicron's CPC 100 and it also had option to demagnetize.

#### Attachments

• 51 KB Views: 219

#### jim hardy

Science Advisor
Gold Member
2018 Award
Dearly Missed
I injected voltage upto 1500 V in a current transformer having knee point voltage 750V @ 30mA.

Interesting - i never saw such an instrument...
in my day we'd have used analog meters and graph paper .

i can't make out the display
milliamps on horizontal, volts vertical ? Knee where i circled in red?

#### Attachments

• 53.5 KB Views: 270

#### Manoj Sahu

Interesting - i never saw such an instrument...
in my day we'd have used analog meters and graph paper .

i can't make out the display
milliamps on horizontal, volts vertical ? Knee where i circled in red?

View attachment 236920
Yes. Exactly. Voltage on y -axis and Current on x -axis. You have marked at the right place. The result was as follows
Knee point voltage -804V
Current - 5.508 mA

#### Windadct

The Knee is just the Saturation point, viewed by exciting the Secondary ( instrument side) - so really the question is how does the designer determine the saturation limit of a transformer. So different objectives ( Price, weight, size, cost, window,) are balanced against the technical limitations. Since the window types only have One primary turn, the secondary turns is set, limiting the options of the designer. Core type ( material) and size ( think cross sectional area) - are the biggest factors.

The different classes have different objectives - a metering CT needs to be very accurate at all of the operating point up to the rated current - usually this is with a relatively low burden load ( the meters). A protection CT really has a much bigger ask - as the accuracy must at least be known and understood at 10, 20 or more times nominal current - this may be through a number or variety of protective relays ( although the multi function digital protective relay helps greatly with this issue). In basic applications, like a feeder with only Over-current, GF protection the same CT may be used for basic metering ( usually non-revenue).

#### Manoj Sahu

The Knee is just the Saturation point, viewed by exciting the Secondary ( instrument side) - so really the question is how does the designer determine the saturation limit of a transformer. So different objectives ( Price, weight, size, cost, window,) are balanced against the technical limitations. Since the window types only have One primary turn, the secondary turns is set, limiting the options of the designer. Core type ( material) and size ( think cross sectional area) - are the biggest factors.

The different classes have different objectives - a metering CT needs to be very accurate at all of the operating point up to the rated current - usually this is with a relatively low burden load ( the meters). A protection CT really has a much bigger ask - as the accuracy must at least be known and understood at 10, 20 or more times nominal current - this may be through a number or variety of protective relays ( although the multi function digital protective relay helps greatly with this issue). In basic applications, like a feeder with only Over-current, GF protection the same CT may be used for basic metering ( usually non-revenue).
As you said metering core needs to be accurate all the time and protection core accuracy must be at least known or 10- 20 nominal current, is this the reason protection core is having larger saturation point.
I have a CT in which there are 4 core.
The first core is used for metering whereas the other cores are used for protection and differential protection.

#### Windadct

You have a CT with 4 core - or 4 windings? Ether way that seems odd ... - do you have a link to a Datasheet?

But - yes the protection one I would expect to have higher saturation

#### Manoj Sahu

You have a CT with 4 core - or 4 windings? Ether way that seems odd ... - do you have a link to a Datasheet?

But - yes the protection one I would expect to have higher saturation
It has 4 core marked as
1S1-1S2-1S3 400-300/1A
2S1-2S2
3S1-3S2
4S1-4S2

#### Manoj Sahu

Exactly. It's installed in a 66kV Gas Insulated Substation

#### Windadct

Here - the cost of adding the CT is high so it is worth using a more complex one...

#### Rajesh52

Knee point of the CT at BH curve depends on the following parameters.

1. The Fault current of the electrical network
2. The burden of the CT (Rct)
The designer ask the above information before designing the PS(Special protection) class CT.The information is sought to design the CT accordingly to system requirement in order to avoid the saturation of the CT in case of the fault. The CT saturation will cause no tripping of the circuit breaker and severe damage may takes place.

The Knee point voltage of the CTs installed for particular installation will differ from other electrical installation because different electrical installations have different fault current. The fault current depends on the rating of the transformer installed, number of HT and LT motor installed etc. The protection class and PS class Cts have less knee point voltage as compared to the PS class CT. The metering class CT saturate at lower knee point voltage. The protection class CT has more KPV and it does not saturate in case of the fault.

Last edited:

### Want to reply to this thread?

"Why do Current Transformer cores have different knee point voltages?"

### Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving