Column heater leaking voltage, possibly from switch?

[User276354]

I have a heater like this,

The timer dial was broken, so I opened up the heater, eliminated the dial from the wiring, and tested it. Unfortunately there was some voltage registering at some places I didn't expect. I labelled the following diagram:

F is the component no longer included. C is a tipping-switch, working fine. The dial-thermostat at E is working fine, the safety-thermostat at G is working fine. But at D, there is a voltage at the side of the brown wire. It should be 0, but it isn't. D is a switch, giving 0, 1,2,3 as settings (see first piccy). When this switch is set to 0, voltage at the brown wire connection to D is about 50V, but i thought it should be zero. When the switch is set to 1, 2, or 3, the voltage at the brown wire reduces to about 23 volts. BTW, I live in a 240 V country and the input voltage seems to measure fine. Whatever voltage I read at the junction between D and the brown wire shows up also at A (the earth wire).

Here is a close-up of the switch.

I thought that any voltage appearing at earth was a problem.

There is one possible complication: many of the appliances in my house have 'live' casings, because I live in a dodgy country with very lows standards in electrics (Egypt). For example, the macbook pro I am using right now , when I rest my hands on the keyboard, is generally around 20-50v (so I use a wireless keyboard!). But it's not usually a problem unless I take my shoes off.

But I didn't think this could be the full explanation, since the voltage to the heater measures zero all the way from the entry (B), through the tip-switch and thermostat; it is only at the switch (D) that voltage seems to be 'lost'.

Any ideas?

tim

 

[alec_t]

I thought that any voltage appearing at earth was a problem.

Normally you treat earth as 0V and measure with respect to that reference. So what are you using as a reference?

 

[User276354]

I am using the metal water pipe which runs down the apartment block to the ground. I attach one lead from the multimeter to this. It indicates the 240/0 V from the mains correctly, so I think everything is fine

 

[alec_t]

That pipe should be bonded to the earth connection at the consumer unit.

I thought that any voltage appearing at earth was a problem.

It is. It seems either your 'earth' connection or the pipe bonding is faulty or even missing. Apart from safety considerations, without a good earth your measurements are likely to be affected by your body acting as either an antenna or as a capacitive coupler to mains voltage.

 

[User276354]

Given the problems with the 'general earth', there still remains the problem of why there is a drop of voltage over the switch in the off position. Any ideas on that aspect? There's not much I can do about the bigger problem of the earth of the building.

 

[JimB]

I am using the metal water pipe which runs down the apartment block to the ground. I attach one lead from the multimeter to this.

That sounds like a very poor technique.

It indicates the 240/0 V from the mains correctly, so I think everything is fine

Maybe it does, but that is a bold assumption.

You have two supply wires from the mains, the live wire (usually Brown colour) and the neutral wire (usually Blue colour).
When faultfinding around the mains power supply wiring inside a piece of equipment, it is usual to refer all measurements to the neutral.
So in this case, for correct fault diagnosis, one lead from the meter should be connected to the neutral wire where it enters the equipment under test.

When this switch is set to 0, voltage at the brown wire connection to D is about 50V, but i thought it should be zero.

Indeed it should, but...

What kind of meter are you using?
Your picture shows a digital multimeter.
One of the big problems with this type of meter is that they have a very high input impedance.
Normally this is a good thing, because the meter will not load the circuit under test and give lower readings than what is actually there.
But in mains powered equipment, when the supply switch is open (off), very small (harmless) currents flow through the stray capacitance between wiring and switch contacts.
The result is that a DMM like yours will show stray voltage readings.

If you had an old analogue type meter, which has a much lower input impedance, you would not see these stray voltages.

JimB

 

[Nigel Goodwin]

If you had an old analogue type meter, which has a much lower input impedance, you would not see these stray voltages.

Not strictly true, particularly for the subject under discussion here.

A digital meter is 10,000,000 ohms input on all ranges, an analogue meter (Avo 8 for example) is 20,000 ohms per volt, so on the 1000V range (such as you might be using for mains testing) it's actually 20,000,000 ohms, twice that of a digital meter.

However, I fully agree that an analogue meter is much better for purposes such as this.

 

[User276354]

I never knew the difference before between the usefulness of analog MM. Thx

 

[JimB]

Not strictly true, particularly for the subject under discussion here.

A digital meter is 10,000,000 ohms input on all ranges, an analogue meter (Avo 8 for example) is 20,000 ohms per volt, so on the 1000V range (such as you might be using for mains testing) it's actually 20,000,000 ohms, twice that of a digital meter.

Hmmm...
Nigel is correct.
However, having said all that, I cannot remember experiencing this problem when using an analogue meter such as an AVO 8.
How many years is it since I used an AVO8? 25 or so!

I do however stand by my statements that the stray voltage which you are seeing are due to the high impedance of the DMM.

JimB