Measuring difference between current in live and neutral

[cupoftea]

Hi ,
We wish to use this CT for detecting if the current in live and neutral is the same....and if different, how different?
Supposing there was 30mA of difference between live and neutral, then do you know what the output would be of this?

ZCT409 torroid Current transformer

Zero Phase Current Transformer For Leakage Current Detection ZCT409 Manufacturer | Ctzentar.com

Xiamen ZTC Technology focus on leakage current detection since 1993,offer Zero phase current transformer for leakage current detection ZCT409,leakage current detection wholesale.Factory price.

www.ctzentar.com

 

[KlausST]

Hi,

it´s just physics. (I don´t call it "accuracy")

Current runs in a loop. Thus the current to a load is exaclty the same but in opposite direction.
Every current creates a magnetic field. Now if two wires feed the identical current but in opposite directions, then the magnetic field compensates to zero. Exactly zero.

But when the current finds a "new path" (second path, shunt path) around the RCD, then resulting magnetic field is not zero anymore. If the magnetic filed is higher than the RCD threshold then it trips.

***
If you try to use two independent current measurements - each with it´s own accuracy error - and subtract the values to get the difference...it´s much more difficult.
On 16A when one error is +0.1% and the other is -0.1% then you get a difference of 32mA.

But when an RCD has an accuarcy error of 10% (mind: this is hundred times 0.1%) then it just causes to shift the trip current from 30mA to 33mA.

Klaus

 

[crutschow]

Read the data sheet (below):
According to that, it outputs 125-150mV into a 150Ω load with 500mA difference current, thus with 30mA current you would get 7.5-9mV.

 

[cupoftea]

Thanks, it strikes me that we have chosen totally the wrong torroid CT for this job. Our current is only 16Arms.......so we are seriously sacrificing resolution by using this torroid (ZCT409), which is rated to 225A...would you agree with this?
Also, the inner diameter is 24mm......and our live and neutral conductors are each only 3mm diameter....as such, they barely fill this torroid. Would you agree that really the inner diameter should be just big enough to fit on the conductors that go through the torroid ring?

So do you agree we have chosen totally the wrong torroid.

I reckon we are going to struggle to read 7.5mV to 9mV. The noise could drown it out. Do you agree?

 

[crutschow]

Yes, I agree you should use a CT rated closer to the maximum current you will have.
It will give you more output voltage for a given current imbalance.

A 225A unit is definitely extreme overkill.
Why was such a large unit selected?

 

[KlausST]

Hi,

To measure difference between Live and Neutral you have to run both wires through the CT.
The resulting current is just the difference. No need for the CT to be rated for load current (16A).
If you safely want to measure 30mA then with a lot of headroom you may use a 100mA CT.

If you want to measure difference current up to 16A, then use a 16A CT.

Klaus

 

[cupoftea]

Thanks, i am confused why our contractor has selected a ZCT409 , which is 225A rated, and has an aperture way bigger then the two conductors going through it....do you know why he might have done this? For an EV wallbox application, surely as you say, its just the difference current we are interested in?

--- Updated Aug 13, 2021 ---

Also, in the ZCT409 "datasheet" of the top post, there are 4 wires coming out of the ZCT409, but it doesnt tell what each wire is for?

--- Updated Aug 13, 2021 ---

Hi,
Do you know a typical CT torroid for detecting earth current, and live/neutral imbalance current in 16Arms EV wallbox's? I am thinking they must be high volume, well known parts now?

 

[FvM]

I agree that the purpose of the four wire connection is unclear. A professional designer would ask for a complete datasheet and otherwise avoid the device. If your contractor selected the device, ask him for consistent data. May he can decipher the chinese text on the product drawing?

 

[cupoftea]

Thanks,
Hi,
I have just done the following test with a ZCT409 CT torroid.
I put conductors carrying differential , equal currents through it as shown, but i get an significant output voltage from the wires that come out of this torroid.
I should not get an output when the currents are equal. Why do i get an output?

ZCT409-4 hosted at ImgBB

Image ZCT409-4 in Cupoftea's images album

ibb.co

(Just to note, i get an even bigger output when i reverse the direction of one of the wires.)


Note:
In the test, the secondaries are shorted by the orange wires, and i raise the variac output volts till i get current flowing in the shorted secondaries.

 

[FvM]

The setup doesn't make sense. There's no guarantee that the current from two output windings is equal. A useful residual current mesurement setup would use forward and return wire of the same circuit.

How about the discussed four coil terminals?

 

[cupoftea]

I buzzed them out and found which were connected...then put the resistor on one pair, then the other......i just presumed that the wires were the ends of a "secondary" coil.....4 ends, so two coils.

Yes i realise they may not be equal currents in the above picture, but i would have thought reasonably equal, since its a dual secondary mains transformer?

--- Updated Aug 13, 2021 ---

A useful residual current mesurement setup would use forward and return wire of the same circuit.

..Thanks, yes, and we could easily do that if we load the device output...but with 240VAC mains, and a 16Arms load...thats wasting 3.8kW , and with multiple test jigs in the production facility...we are going to use too much power in the production facility.

--- Updated Aug 13, 2021 ---

Actually, sorry you're right, i just put the ZCT409 with go and return from one secondary going through it, and i get zero output from the ZCT409 CT torroid

 

[Easy peasy]

Thanks, yes, and we could easily do that if we load the device output...but with 240VAC mains, and a 16Arms load...that's wasting 3.8kW , and with multiple test jigs in the production facility...we are going too use too much power in the production facility.

Given that the test is only for ~ 15 - 30sec the ave power is very little indeed

If they are barring this test for real power reasons, but still want 230Vac @ 16A rms - then they have set an impossibility unless you choose for a load an inductor with XL = 14.375 ohm ( i.e. 45.76 mH @ 50 Hz ) = 16A ac for 230Vac applied, this will be a largish inductor of several kg

 

[cupoftea]

Thanks, thats a good idea about the big inductive load.Though i am not sure where we would stand with the actual high current still coming into the factory when we have multiple production test lines running in parallel.......this is the problem....the high number of parallel production lines all wanting the high power test.

 

[Easy peasy]

they are making ridiculous requests.

 

[cupoftea]

Hi,
We have live and neutral wires going through a torroid aperture, each carrying 16A maximum
The torroid is wound with a 100 turn coil.

Whenever live and neutral currents are not the same, there is obviously an output on the 100 turn coil.
We wish to tell when the live and neutral conductors are >30mA (RMS) different. (ie, its an RCD function)

Whats the best/smallest circuit to allow this measurement? (measuement will be by ADC of micro)?

We obviously cant diode rectify the coil output it as the diode drop would give inaccuracy.
Obviosuly when the live and neutral are 30mA in difference, we will get an output of 300uA out of the coil.

 

[betwixt]

I wouldn't bank on getting 300uA for 30mA difference!

However, the usual method of converting current to a voltage is to use a resistor! It's usually called the 'burden' when used in this application.

To linear convert it to a voltage an ADC can read, use a precision rectifier circuit and an amplifier if needed.

Brian.

 

[cupoftea]

I wouldn't bank on getting 300uA for 30mA difference!

..Thanks, ..even though its a 1:100 CT?

 

[betwixt]

Is it 1:100 though?
You have a wire passing through it and you have turns around it. To make it 1:100 you would need a complete turn on the primary side. Even then you would have losses to take into account.

Brian.

 

[cupoftea]

Is it 1:100 though?
You have a wire passing through it and you have turns around it. To make it 1:100 you would need a complete turn on the primary side. Even then you would have losses to take into account.

Thanks, you have put a very poignant point forward there....i am now wondering just what turns ratio we have with a 1:100 CT?
From knowledge of 1:100 CTs used as sensors in SMPS, you are right, they are an actual full single turn wound fully round the bobbin..not like just a wire going through a torroid aperture....

..you have opened up a whole wraft of thoughts in me...for example....RCDs work by sensing >30mA of difference current between line and neutral, -they work by passing line and neutral through a torroid aperture...so how do they manage to sense the difference current with such accuracy?........they cant possibly know what "real" turns ratio they have?

You are right, when you have a simple wire passing through a torroid aperture, the leakage flux must be all over the place....i am going to have to start by looking how RCDs work?

 

[FvM]

Measurement error of a CT doesn't depend much on the primary winding geometry, it's mostly caused by finite core permeability, secondary winding and burden resistance. You need to know the core and winding parameters for a reasonable calculation.