Odd electrical problem

[RobertArthur]

@ Rabbit, a false positive is possible with these non-contact voltage testers. The type with a capacitive sensor (yours) is already better than the traditional little neon screwdrivers. To see where there may be a broken wire or loose connection, voltage not present at some point, they give you a quick and dirty indication. You are however not measuring with a real current, producing a "hard" voltage. One of the reasons why in industrial settings they don't like these testers and high impedance multimeters, Duspols and their lookalikes, a simple yes/no is better than reading a voltage in many digits, pretending that they are more precise. To measure electronic equipment, dc voltages, they are excellent, not loading the circuit under test. In the rather dangerous environment of 230 V and 400 V (3- phase) you want to be 100 % sure: hard voltage present?

Even my 40 years old Steinel voltage tester with two red leds - you need a real current (in this case about 1.5 mA AC) to turn them on - is still hanging around in my toolbox. Also handy for old school DC voltage testing in cars: with a 12 V input still a good visibility of only one of the red leds (AC: both on), below 5 V it starts dimming.

Reasons for getting a false positive.

Voltage testers compared.



Department low impedance and loading continued:

The effect of measuring with a high - or low impedance, let the current flow. This time a DC test, a CR2032 coin battery almost at the nominal 3.2 V, looks almost healthy, under a little load a different story. Should have been very close to the nominal voltage.

[RobertArthur]

And much more information with my vintage meter (left) made by the battery experts, Every battery type its own specific load.

[ajm]

Our odd experience was with the pump in the well. After 16+years it pcjed in and was replaced by a new identical model. This worked fine for about 18 months and then tripped out after a couple of seconds. Still in guarantee I took it back to LM who tested it and said there was nothing wrong with it. Back home plugged it into a different socket and it was OK. Checked the contacts in the original socket OK. Only thing I could think of was to change the polarity and,heyho, it now works fine.

[Biggles]

Thank you everyone for your contributions which have resulted in some very useful information, mainly contributed by Robert, for which I am very grateful. I used my cheapy DVM to test for power on this occasion, and either failed to connect to the domino properly (it was in a very difficult and restricted situation) or the -ve lead on the DVM had detached itself, as it is prone to do, without me noticing. No other evidence of a voltage surge has been found anywhere, so essentially it is still a mystery, as the lights were 100% when I used the shower previously.

It has just occurred to me that a surge would not affect lights when they are switched 'off' would it?

(I'm sure there used to be a head-scratching Smiley!)

[RobertArthur]

Biggles, it is something of a black art to understand how lightning-induced overvoltages choose their victims. Turn left, turn right in the wiring? Let's jump over these two open switch contacts, only several millimeters, we just travelled from far above....easy.

Led lights, electronics inside, only a very small overvoltage above a certain level is enough to destroy critical insulation layers inside semiconductors. This is one of the weaknesses of LED lights and also a design problem, for a few dollar cents more you can do already a lot. Stiff competition out there, marketing guys and quality control.....

[RobertArthur]

Out of interest, part II, an additional remark about this simple old school Steinel voltage tester.

1/ It has a voltage range from 4.5 V to 450 V AC/DC, testing in a three-phase house wiring also possible.




2/ The little PTC resistor inside, low value when cold, does indeed trip a 30 mA RCD when measuring between phase and earth, inrush current. One school of thought calls this a feature, the other doesn't like it at all. That's why several modern types include an extra current limiting resistor. A way around is preheating the tester, phase - neutral for about 15 seconds and then check for the presence of earth, phase - earth, no problem at all.

And if you really want to check a RCD for its correct functioning it is easier to push the test button. It doesn't tell you everything, it measures with quite a high test current of 2.5 x I nominal (30 mA), about 75 mA. With the help of a little resistor inside. These RCDs are considered to be defective if the response time is too slow or switch off above 30 mA , special test instrument needed.

To cut a long story short: real hard voltage testing possible for only a fistful of dollars (or any other currency).