Weird result determining order of sockets on ring circuit

[jackhaughton]

Hello,

I am trying to work out what order the sockets are connected in on the ring circuit in my house. I have been trying to do this by connecting the line and earth pins on a 13A plug together, plugging it into a socket (with the circuit breaker open naturally!) and then measuring the resistance between earth and line at each of the other sockets. The higher the resistance, the further away round the ring, right?

But I've run into a puzzle. There are two adjacent sockets in my kitchen. If I short one of them, and measure the resistance at the other, I get 0 milliohms. However, if I short a socket on the other side of the ring, and measure the resistance at one of the adjacent sockets, and then at the other, I get 470 milliohms at one and 380 at the other. Doesn't seem to make sense!

I have disconnected nearly all equipment from the circuit (there are a couple of sockets with built in USB chargers which I haven't disconnected, but the resistance between line and earth with no shorts is >200 ohms, so I don't think they're significant). I would suspect that there was something wrong with my tester, but the readings are reliably reproducible - if you blindfolded me and told me the reading on the tester, I could tell you which socket it was plugged into. I don't see how this is possible when there's no resistance between the two sockets. Any ideas?

Thanks
Jack

 

[bernardgreen]

The resistance readings are being affected by stray voltages induced in the cable from other cables. I assume you are using a multimeter with a 9 volt or lower battery. A multi-meter sends a test current through the circuit being measured. With the majority of multi-meters this current is a few milli-amps or less. Currents induced by stray voltages can be higher than this current and thus the meter gets incorrect readings.

To trace the route of a ring final requires a test current of at least one Amp to prevent stray induced current affecting the readings enough to make them un-usable.

 

[securespark]

If you isolate your supply, there would be no stray voltages.

 

[jackhaughton]

Thanks for the replies. I'm using a Di-Log 9083P, manual says it pushes 200mA into 2 ohms on the continuity test, but maybe it's not enough to accurately trace the circuit in this way. Other possibility is that I've missed a piece of equipment on the circuit somehow. I tried opening all the CU switches and the isolator on the DNO's board, just for belt and braces, but still got ~ 100 milliohms difference between the two sockets. Guess I'll have to pull up some floorboards to trace the section I'm interested in, doesn't seem like these results are to be relied on.

Thanks
Jack

 

[John D v2.0]

Have you zeroed the leads correctly?

Also not affecting your situation but are you certain it's actually a ring and not a circuit spurred off a ring?

As others have said completely disconnect the circuit from the neutral and if possible just remove all the sockets to make the process simpler. Ours was all over the place and half the sockets were spurred under the floor so I just disconnected them all because I didn't trust the readings, but should be OK if your kit is good.

 

[JohnW2]

Thanks for the replies. I'm using a Di-Log 9083P, manual says it pushes 200mA into 2 ohms on the continuity test, but maybe it's not enough to accurately trace the circuit in this way.

Whilst everything bernard wrote was theoretically correct, I have often been amazed/impressed to the extent that even cheap multimeters are relatively unaffected by incidental induced AC in the conductors concerned when measuring resistance. However, having said that, I can't think of any better explanation for what you have observed!

Other possibility is that I've missed a piece of equipment on the circuit somehow.

Even that could not, in itself, explain your measuring different resistances from a fixed point to two other points that had 'zero ohms' resistance between them.

Kind Regards, John

 

[JohnW2]

Have you zeroed the leads correctly?

Even if he hadn't, provided he was always using the same leads, I don't see how that could bring about the results he has observed. Both of the resistances he measured would be 'wrong' if zeroing had not been done correctly, but the two figures should nevertheless have been the same.

Kind Regards, John

 

[bernardgreen]

With all the power off and a 200mA test current stray voltages from other cables are going to be insignificant.

Other possible causes are radio frequency energy picked up by the cables and rectified into a DC voltage by poor joints ( copper oxide diode ) or filter components in the USB circuits. Are you close to a transmitter site ?

But as has been said disconnecting sockets and testing end to end of the cables is the better option. It has the advantage that any suspect joints , loose terminals can be found and repaired.

 

[jackhaughton]

Hi,

Yes, I zeroed the leads. These two sockets are spurred off the ring, both on the same spur. The circuit's neutral goes to an RCBO. The neutral was definitely isolated from the supply neutral when I did the test (by both the CU main switch and the DNO's isolator), but according to the diagram on the front of the RCBO it doesn't isolate the neutral, just the line. So the neutral would still have been connected to the neutral of other circuits, just not to the supply.

I just tried disconnecting the RCBO's neutral and rerunning the test, no effect. Disconnecting the sockets sounds like a good plan. Thanks all for your help.

Jack

 

[JohnW2]

These two sockets are spurred off the ring, both on the same spur.

That probably makes your observations even more surprising. However, in passing, is there a fuse (i.e. an FCU) at the origin of that spur? If not, you should not have two sockets on the same ('unfused') spur.

Kind Regards, John

 

[jackhaughton]

Hi John,

Yes, I wondered about that. I don't know where the spur originates from yet (probably under the floor upstairs), but there aren't any unexplained FCUs in the house, so I imagine it's unfused. Not sure when the last rewire of the house was, but it's all red and black cabling, so I guess it's at least 10 years. Not sure when the regulation about multiple sockets on a spur came in. The reason I'm trying to trace the circuit is that this spur needs removing before the extension that it's in gets knocked down, so it will cease to be a problem shortly!

Thanks
Jack

 

[securespark]

14th Ed., 1966:

A.40 .....Not more than two socket-outlets, or one twin socket-outlet, or one stationary appliance shall be fed from each non-fused spur.


This allowance for two sockets on a non-fused spur had disappeared from the regs by the 15th Ed., 1981

 

[EFLImpudence]

How are you measuring?
With the probes on the conductors or through a plug-in socket?


Is no one suspicious of the 0mΩ reading which will actually be 0.00Ω?

According to the spec sheet, the meter only measures to two decimal places.
Depending on how it works 0.09Ω (0.47 - 0.38) could be displayed as 0.00.

 

[jackhaughton]

Using a plug-in test lead. I zeroed the leads by plugging my shorted plug into one socket of a double outlet, and the test lead into the other. So I guess there was no measurable resistance between the two double sockets. They are less than a meter apart.

You're correct about the decimal places of the meter. The reading was sometimes 0.00, sometimes 0.01, so I guess we can say it's <15 milliohms.

Thanks
Jack

 

[EFLImpudence]

You're correct about the decimal places of the meter. The reading was sometimes 0.00, sometimes 0.01, so I guess we can say it's <15 milliohms.

Your meter may not round-off but simply chop-off (I don't know) so 0.00 could be anything from 0.0000 to 0.0099 and 0.01 anything from 0.0100 to 0.0199.

So a resistance of 0.0090 could easily jump between 0.00 and 0.01.

The resistance of two metres of cable, two socket switches and four plug contacts may easily be 90mΩ.
It definitely can't be zero.