AC current sensing 2

[checonbill]

Hi
I am looking for a current sensing device with a response time of 100 ms or less that will be able to sense up to 600 amps this will be interfacing with an automation controller.Any input will be helpful

Thanks

 

[sreid]

Check out LEM current sensors. Fast, reasonably priced.

http://www.lem.com/

 

[ScottyUK]

If it's AC then consider a Rogowski coil. Somewhere like

http://www.rocoil.co.uk

would be a good place to start.

Use the LEM transducers with care on fast edges, or install it at the earthy end of the circuit. Some of their products suffer badly from dv/dt coupling from fast edges which gives false output signals. It's fairly easy to screen the transducer using an electrostatic shield made from self-adhesive copper foil with a flying lead to earth.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Skogsgurra]

Why reject the standard CT?

At 600 A - that seems to be a natural choice.

Are high frequencies involved? "Automation controller" - sounds like 50 or 60 Hz to me.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ScottyUK]

Very true skogs. It is the "100ms or less" requirement which seems odd for a 50/60Hz application: it would be more normal to see a bandwidth requirement for an AC transducer. Perhaps the OP can expand a little what he is trying to measure.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[checonbill]

It is 60 Hz what is the response time of a regular CT as the output from that would be input into the controller which has a 75 ms response time using a 0-5 a signal if I can get a 0-5 vdc signal my controller response time is at 10 ms but I am sure I will lose with a conversion but not sure how much

 

[checonbill]

I am trying to measure a current of about 550 amps 60 Hz in a 3 phase circuit this signal input to an automation controller.The load is only on for 1 sec and during this time I need to determine what the sequence of phase operation is and what the load is.The other information I am looking for is if my load gets to 600 amps on any phase I need to dump the load very quickly so my response time is very important so the controller will turn off the load.The controller as I said has a response time of 10 ms using a 0-5vdc input and an output turn off time of 100 micro seconds the controller has the ability to make this decesion of turn off in microseconds.So I am looking for the fastes response time I can get with the current sensing. Thanks for any input on this project

 

[roydm]

I would go with the LEM also, they have 4-20 mA or Voltage output models which makes it easy to interface with controller.
I think they are Hall Effect 20 kHz seems to be mid range for the transducer. Why do you need so high?
Roy

 

[davidbeach]

Hang on a moment. First off, the times in milliseconds (ms) are well within the capabilities of conventional CTs. The one reference to microseconds ([μ]s) seems a bit odd when dealing with 60Hz AC power.

What does "load gets to 600 amps on any phase I need to dump the load very quickly" mean in quantifiable terms? If that 600A is rms, then you need at least a full cycle to know the rms magnitude. If you need something faster than that, you will have to use a level detector set at 848A, but then you will be subject to tripping on transient currents, such as inrush.

 

[LaffeyDD724]

Checonbill, you may find what you need at this website try this URL.

http://www.loadcontrols.com/products/products.html

Tom

None of us is as smart as all of us.

 

[VladimirM]

Hello everyone,

Would You please explain the following: we have a conventional current transformer (CT) and we have a Rogowski Coil Current Sensor (RCCS). Both use the same principle - Faradey's Law, right?

Then why do we have such a narrow frequency range for CT - just 50/60 Hz, while RCCS has a very broad one? For example AFAIK a sensor from Phoenix Contact has a (30...6000)Hz range with 4-20mA output - very good comparing to conventional CT.

Is it somehow related to the fact that CT has iron core (saturation & hysteresis) while RCCS has a non-magnetic core?

 

[Skogsgurra]

Yes. A Rogowski coil never saturates. The electronics does, but not the coil.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Skogsgurra]

Also, the high-frequency response of a Rogowski coil is a lot better because there is no leakage reactance in it.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[VladimirM]

Yes, but why does CT has only 50/60Hz while RCCS may have (30...6000)Hz or even wider frequency range?

 

[VladimirM]

No leakage reactance? Wow! I guess I should the operating principle of this Coil more properly.

And what can You say about CT?

 

[ScottyUK]

The CT will struggle below nominal frequency at rated current because of core saturation. A CT will work at higher frequencies but accuracy will degrade as the leakage reactance becomes increasingly dominant. Hopefully scottf will chip in - he works for a CT manufacturer.

A Rogowski coil bandwidth is essentially limited by the integrating amplifier and the interconnecting cable rather than the coil. BW's into the MHz are achieveable with enough care to the design.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[VladimirM]

"The CT will struggle below nominal frequency at rated current because of core saturation"

That has made a lot of sense to me. I thought through this idea and I had a question to it: does the degree of saturation depends on the secondary circuit ohmic resistance? I mean I came to a conclusion that:
1. If there is an ideal case when total ohmic resistance of the secondary circuit equals to zero (i.e. R2 = 0 [ohm]) then there is no saturation at rated current below nominal frequency. The magnitude of the flux in the core will be the same at 20Hz as at 50Hz (let's forget about 0Hz).
2. If there is a real case when total ohmic resistance of the secondary circuit is non-zero then we have a situation where the greater the R2 is the more the degree of saturation which we will face at lower frequencies at rated current. In other words the greater R2 is the more the magnitude of the core flux at lower frequences. And, BTW, this sounds good at any frequency - I mean that it is true for 50Hz: the more I burden my CT the more it saturates! And at lower frequences we just have a situation where the same ohmic resistance becomes, so to say, "heavier" for CT to carry than is was at 50Hz.

ScottyUK, did I get Your idea correctly?

 

[ScottyUK]

Yes, I think you have the principle. A CT is rated for a specified burden at a specified frequency. If you halve the frequency you could halve the burden, or halve the rating. Bear in mind that the burden includes the internal resistance of the CT. If you increase the frequency then you would assume that you could increase the rating or the burden in proportion, but the leakage reactance starts to become significant.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[slavag]

Hello Vladimir.
Please find attached link on the issue.
Guys sorry, Russian lang.

http://www.electrik.org/forum/index.php?showtopic=11172

Best Regards.
Slava

 

[VladimirM]

ScottyUK, thanks a lot.

Skogsgurra, You mentioned that Rogowski Coil has no leakage reactance.
I looked through the operating principle of this device again and I came to conclusions that:
1. Actually there is leakage reactance in the secondary of the Rogowski Coil but since secondary current is negligible (almost zero) so the influence of this reactance is naught.
2. Rogowski Coil is smth very similar to conventional current transformer if leave CT secondary open and change an iron core to a non-magnetic one. And add electronic integrator.

What do You think about it?