Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

[SOLVED] current sensors vs current transformers

Status
Not open for further replies.

gary36

Full Member level 4
Joined
Mar 31, 2018
Messages
208
Helped
0
Reputation
0
Reaction score
0
Trophy points
16
Activity points
1,806
I had been googling on this topic to decide the choice of CT vs current sensors. Could anybody explain the difference? and when to choose one?
 

Current sensor is a generic term covering different measurement principles including current transformers.

A basic distinction ist between pure AC (CT, Rogowski sensor) and DC capable (shunt, open and closed path magnetic) current sensors.
 

Hi,

Please give a link to a (good) search result, so we can discuss in detail.

Klaus
 

Last edited:

CT has the benefit of intrinsic isolation and are pretty fast. Scale factors exist but choices may be limited (off the shelf). CTs are fairly fast because they are bare.

Hall sensors tend to have an amplifier which slows response. A speed/accuracy trade is inherent if you can't use a bare Hall element (usually the case). I've found some fast enough for load step testing but not for per-pulse current sense at >1Mhz
 

Hi
For harmonic measurements, which device is suitable?
so DC produces no harmonics. Thus there is no need for a Hall sensor.

And for AC: the current measurement device just needs to have a suitable bandwidth to measure the harmonics.
.. and it should not add/generate significant harmonis.

For details I´d need to read datasheets.

Klaus
 

DC does not mean constant current. The current measuring device much be capable of the desired measurment bandwidth.
 

Hi,

When we talk about frequencies or frequency analysis (like with Fourier analysis): DC means 0 Hz, no fluctuation.

When "your DC" fluctuates, then this is caused by AC.

Klaus
 

Hi,

When we talk about frequencies or frequency analysis (like with Fourier analysis): DC means 0 Hz, no fluctuation.

When "your DC" fluctuates, then this is caused by AC.

Klaus

Pretty narrow definition of DC current solely based on Fourier analysis.

Turning on a DC device changes current so DC device that draws current must not exist.

A DC to AC sinewave power inverter has a fullwave rectified sinewave battery current profile so it should be called an AC to AC inverter by your definition.

I would be okay calling DC current as a current that has some net average one way current flow.
 
Last edited:

I would be okay calling DC current as a current that has some net average one way current flow.

A series of pulses (a pulse train) will be called DC according to you?

It is important to realize that there are mixed states; you can have DC + AC.

Whereas DC can be called AC with zero frequency, it is not accurate to call AC as fluctuating DC.

When the AC frequency is specified, it implies sinusoidal waveform. Other waveform(s) add extraneous frequencies.

When we mention bandwidth, it is always implicit a sinusoidal waveform (why?)
 

Hi,
Turning on a DC device changes current so DC device that draws current must not exist.
Multiple frequencies can coexist...and DC, too. No problem with this.

You may have a sinewave.
You may add overtones to get a square wave.
You may DC to get what you call a "pulse train"
All fine.

And yes, I agree that pure DC can not exist. It means "eternally".
But similar to a fourier analysis .. you focus on a limited piece of time, a window, and imagine the same window will repeat again and again.

So if you focus on a window without a switching edge you may focus on the DC.
But as soon as there is a switching edge in this window you will see all the "switching noise".

A DC to AC sinewave power inverter has a fullwave rectified sinewave battery current profile so it should be called an AC to AC inverter by your definition.
Surely not. Because the very very most of the energy is drawn by the DC component. And the AC ripple is not much of interest. But it does exist.
On the other hand if you focus on the SPWM ... you can't ignore the frequencies. It's a main part of the inverter.
Then you may focus on the AC output. Low frequency. You clearly should avoid DC here, but you may suppress the switching frequency.

I would be okay calling DC current as a current that has some net average one way current flow.
No, you would not. The output of an audio amplifier will have some "average one way current flow", almost unavoidable.
It is called DC offset. Unwanted. A bad one may have 100mV, a good one below 1mV.
You may say this is ignorable. But what's the AC voltage level at quiet room loudness!?

I'd say there is:
* DC with a little AC
* AC with a little DC
* and there is AC and DC mixed at about identical RMS levels.

Klaus
 

I would be okay calling DC current as a current that has some net average one way current flow.
Related to current sensors, the important point is if you want to measure this "net average one way current flow", simply can ignore it or need to assure that it doesn't disturb the AC component measurement.

In the former case you'll use a DC-capable sensor, e.g. hall sensor. If the DC component isn't of interest for your measurement, it may still cause problems e.g. by saturating a CT core. You can use a coreless Rogowski sensor or use a DC-capable sensor although you don't need the DC information.
 

DC is current that flows in one direction only. There’s nothing in the definition of DC that says it has to be CONSTANT current.
 
Related to current sensors, the important point is if you want to measure this "net average one way current flow", simply can ignore it or need to assure that it doesn't disturb the AC component measurement.
If you consider "net average one way current flow" over a complete period (cycles) then the AC component automatically gets canceled.

Consider an example: a square pulse train (1V peak; 50% duty cycle) has a DC component of 0.5V. All the AC components gets averaged to zero if average is taken over a complete period.
 

For harmonic measurements, which device is suitable?
Bandwidth, bandwidth, bandwidth.

Nth harmonic times fundamental freq is maybe half
your desired measurement BW. What's N?

CT-1 current xfmr is good to past 100MHz.
A LEM-series (post-amplified Hall) is about 200kHz.
 

Attachments

  • LEM_CAS_Series.pdf
    1.2 MB · Views: 91
  • Making Accurate Current Measurements on Power Supplies with Oscilloscopes.pdf
    4.1 MB · Views: 125
  • Tek CT-2 CT-1 CT-6 AC_Current_Probes.pdf
    1.3 MB · Views: 87

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top