Amecal Earth Leak Detector Question

[nellymcl]

So I purchased the Amecal ST-9810 to see if I could fathom if I had any apparent leakage which was intermittently tripping by RCCB.

Nothing was showing apart from items which heat up i.e. kettle, toaster, tumble dryer, hair dryer. I was getting readings of up to 20mA from the kettle which doesnt seem right. In the teens with the tumble drier, although none of them sit still, they jump around.

Attached is the lead I am using to test with. I am testing around the L and N

Am I doing something wrong that I can't get a good reading?

Thanks.

 

[EFLImpudence]

Clamping L&N is the correct way of doing it. Perhaps it is jumping around.

It could be a faulty meter, though, 20mA alone could be enough to trip an RCD.


Does it read zero if you clamp L, N & E?

 

[nellymcl]

Thanks
No, even holding the meter close to the cable without even clamping it, I get a reading of high teens??
This only happens on the high load items.

 

[EFLImpudence]

Mmmm. Mine doesn't do that - maybe 1, 2 or 3mA.

If yours is reading the same whether the cable is clamped or not, it is presumably not working properly.

 

[nellymcl]

Thanks. Looks like I will be returning it.

 

[endecotp]

For a device like a kettle which you can easily insulate from the ground etc., just measure the current in the earth wire. Preferably not with a clamp meter.

Your "devices that heat up" are just devices that take a substantial current.

20mA is 0.2% of the 10A that is flowing in L and N when your kettle is on. The clamp is somewhat sensitive to the position of the conductors within it, and your L and N conductors are not exactly coincident. It would only need one to be +0.1% and the other to be -0.1% to get that result. What is the spec of the meter? It will surely be worse than 0.2%. What you are seeing is normal.

 

[ban-all-sheds]

Clamping L&N is the correct way of doing it.

I've never looked into doing this.

What "Doh! " am I missing, 'cos I would have thought that if you wanted to measure earth leakage you'd measure the current in the cpc...

 

[nellymcl]

If its just the accuracy of the meter why am I still getting a reading being close to the cable wothout clamping?

 

[endecotp]

If its just the accuracy of the meter why am I still getting a reading being close to the cable wothout clamping?

Because the clamp sensor is not perfect.

 

[EFLImpudence]

I've never looked into doing this.

What "Doh! " am I missing, 'cos I would have thought that if you wanted to measure earth leakage you'd measure the current in the cpc...

It wouldn't make any difference in this case with a kettle but around L & N is the proper way to do it.

 

[EFLImpudence]

20mA is 0.2% of the 10A that is flowing in L and N when your kettle is on. The clamp is somewhat sensitive to the position of the conductors within it, and your L and N conductors are not exactly coincident. It would only need one to be +0.1% and the other to be -0.1% to get that result. What is the spec of the meter? It will surely be worse than 0.2%. What you are seeing is normal.

From an ebay seller who quotes the specs.

Would the error be 1.2% of the 10A or 1.2% of the residual current?

If what you are saying is true then the device surely is not fit for purpose and measuring the cpc would be the only accurate way, although that would be no good for circuits where the actual earth leakage, not cpc current, is needed.

 

[endecotp]

Here is a picture from Wikipedia of how a current transformer works:

The assumption is that all of the magnetic flux from the primary is captured by the iron or ferrite core and so also passes through the secondary.

This would be the case if the core had infinite magnetic permeabillity and air zero. Actually the ratio is about 1000 for iron or 10,000 for ferrite, and the imperfections in the core due to the clamp jaws will make it worse than that. So some of the flux is not captured, and hence some of the primary current is not detected. How much that is will depend on the precise position and alignment of the primary conductor within the core, with a conductor closer to the core probably having more of its flux captured than one further away.

(The meter may use a hall-effect sensor rather than a secondary coil, but the same issue arrises.)

(BTW, I've wondered what lengths RCDs have to go to to make this work. I think that not having jaws is probably a big help, and being able to ensure good physical symetry must also be required.)