Production test of a 16Arms mains distribution box?

[cupoftea]

Hi,
Please may i post fresh about this problem?, as i now am coming at this from an entirely different point of view.......

We need to production test a 16Arms mains distribution box (as attached). How would you do this?

It must be tested with 240VAC mains input.
We must test that the live and neutral conductors can pass 16Arms.
We must test that the Live current measurement device measures 16Arms live current fairly accurately.
We have been told that we cannot dummy load the device with 16Arms, as that would mean dissipating 3.8kW in a dummy load.

How would you do this?

Note the Live/Neutral current imbalance measurement device trips if there is >30mA of difference between live and neutral current........so >30mA of difference must be avoided

 

[Easy peasy]

A variac and a resistor would also perform the 0 - 40mA function

 

[cupoftea]

Thanks,
Thanks for replies.
Please may i ask the question in a different way?

Is the attached actually possible without incurring a budget of
many hundreds of thousands, and employing a large team
from the ivy League of universities?

The current difference of the injected currents must never be outside 30mA unless for <10ms at a time, or else the RCD trips.
The injected currents are AC sines.

 

[FvM]

As previously mentioned, I believe it's possible with an arrangement of (series connected) current transformers. Current imbalance depends on CT performance and DUT impedances.

 

[cupoftea]

I believe it's possible with an arrangement of (series connected) current transformers.

Thanks, i must admit, its the inaccuracy of current transformers which makes me think this is not possible.
...two injected currents, AC, 50Hz, and must always be within 30mA of each other as they are varied from zero to 32Arms.......if they differ by 30mA for more than 10ms, then its not acceptable. (because the RCD trips)

 

[cupoftea]

Basically, We have to make an RCD first of all....ie, be able to accurately detect the imbalance current between line and neutral currents........

We have to detect >30mA current imbalance between Line and Neutral conductors which carry from 0-32Arms, 50Hz. These conductors are 4mm in diameter.

We wish to be able to detect the 30mA imbalance as accurately as possible, and with as much resolution as possible.

Please assist us in selecting the correct CT torroid for this job.?

If not, then please inform us whether it should be constructed from iron, or laminated iron, or powdered iron, or ferrite?


The ZCT409 torroid could be used, but will give us an output of just 8mV peak when the current imbalance is 30mA. (Also, there is a tolerance here, so its actually 6.5 to 9mV peak.)

ZCT409 CT torroid:
https://www.ctzentar.com/zero-phase...for-leakage-current-detection-zct409_p28.html

 

[KlausST]

Hi,

We wish to be able to detect the 30mA imbalance as accurately as possible, and with as much resolution as possible.

No that is no specification.
You won´t spend millions of euros to get the accuracy down to picoamperes.
--> give useful requirements

The CT just needs to be able to process the 30mA @ 50Hz. The 32A don´t count.

There are active ones that even can go down to DC. But I guess the most accurate ones for AC (not DC) are passive CTs with 10...100 secondary side winding count, active current to voltage converters (TIA. For best linearity and low distortion) and digital RMS converters.

***

A TIA in best case acts like zero ohm input. To further improve on linearity and distortion you could try to generate "negative resistance". The limit is the ohmic secondary coil resistance.

***

RCDs have a difference_current to trip_time dependency. So maybe you need a fast RMS converter algorithm.

Klaus

 

[FvM]

Basically, We have to make an RCD first of all...

So, if you build a car, you have to start with re-inventing the wheel? There are wheels available on the market, but they are of course not good enough for your innovative product...

--- Updated Aug 20, 2021 ---

How did the discussion shift from testing the RCD to designing it? Is it still about "production test" or did we silently arrive at designing a wall box?

 

[cupoftea]

the design is for the test kit for the distribution box.
We need to sense imbalance current in order to zero it out.
This is the test kit that injects matched currents into the disty box.

Do you know why ZCT409 is rated to 225A?

Zero phase current transformer - Xiamen ZTC Technology Co.,Ltd

Differential current sensor, zero phase current transformer made by Xiamen ZTC Technology Co.,Ltd is widely used in EV charging Module, GFCI, RCD.

www.cnzentar.com

 

[KlausST]

Hi,

the design is for the test kit for the distribution box.
We need to sense imbalance current in order to zero it out.
This is the test kit that injects matched currents into the disty box.

I think you misunderstood the whole thing about injecting imbalance.
My guess is that the test kit never injects imbalance in a way that the imbalance exists at the RCD.
Again my gess is that what you call "imbalance" either goes to both wires of the RCD (no imbalance on RCD) or the current does not pass the RCD at all (no imbalance on RCD).

I recommend to draw a sketch (for us, but also for yourself) with
* grid input
* RCD
* test box
* distribution box
....
and show the loop for the usual load current as well as the test current that potentially may cause imbalance on the RCD.

Klaus

 

[Easy peasy]

You appear to be really over-thinking this whole business - really there are only a few tests of importance:

• Does it handle 265Vac without blowing up ?
• will it handle 1000V megger phase and neutral to earth ?
• are the earth connections sound, < 200mV @ 10A
• on 230Vac applied if you slip a small wire thru the RCD and put >=30mA through it - will it trip ?
• if you short the output P & N, and put 30A into the input for 3 mins - does anything over heat ( IR handheld inspection )

job done - you're welcome ....