- Reaction score
- 13,475
Look at the response from others who replied with an explanation showing that there is an issue.
Some relevant comments there but also a lot of errors. I'm not picking this apart to be awkward, I was genuinely surprised by your original comment and it set me thinking about the possibilities. But I could not find any valid ones, and neither did these other threads. Let's look at the key points raised:
Capacitor charging current. Yes, true, but only for a few seconds unless your IR tester is faulty. Capacitors wired L+N->E are never very large in value, to limit AC leakage and because of the job they do (RF suppression for EMC). I often test equipment with quite high capacitance to earth and it only takes seconds for the reading to settle with a normal tester.
Capacitor leakage. I don't know how many capacitors these guys test but at a guess I've tested many tens of thousands in my life, of all kinds, sizes and ages. Any of the types of capacitor that are used L+N->E (normally class Y construction) that test under tens or hundreds of megohms, are faulty. But I've never found any that are not also physically destroyed, because they are designed specifically to have high integrity against developing leakage. Other kinds of capacitors do develop significant leakage, e.g. wax-impregnated papers in vintage radios over 60 years of age; if the radio works today most likely they have been changed already and they weren't usually used L+N->E anyway. 'Aluminium foil jobby' is not a kind of capacitor I recognise but if it means aluminium electrolytic, they do have high leakage but are never used L+N->E for a bunch of reasons.
Bleeder resistors. Exist, often across X-caps (L-N) which are typically much larger, but not commonly class Y capacitors. The numerical example given contains errors: The initial voltage of interest is Vpk not Vrms, 325V for 230V AC, 50/325=15% so 2RC<5s and R<=250MΩ. A resistor that high is expensive and potentially unstable, so practical designs use lower resistors on the rare occasion they are fitted at all. I mentioned 7MΩ - that's a favourite of Sony that one still sees occasionally, more often its >99MΩ. If we take your '0.01' reading as 20kΩ, as it could well be, it would still need 7000/20 = 350 pieces of Sony equipment of these specific models on the circuit to cause it.
Inductors and zener diodes. Since when were inductors ever connected L+N->E in appliances? Nonsense. Zener diodes / VDRs can be present as surge protection and I did mention these as a typical special case where rated for 120V mains only. But testing at 250V should not cause 230V-rated surge arrestors to conduct significantly, or else they would try to clamp the 325Vpk of normal mains. Therefore they will not badly affect the reading at 250V (They can ruin it at 500V though).
Filters. Well these are where the said class Y caps are usually located, usually with some inductors, often built into the IEC inlet etc as a power entry module. They are no more and no less than the components inside them. All reputable filter makers will permit insulation tests, many filters are tested to >2kV in the factory (although repeated, prolonged tests at this voltage will cause damage). Even the cheap nasty unknown brand ones survive 250V just fine, or they wouldn't last on the mains at all.
There are other errors in posts in those threads (e.g. confusing different types of capacitors and their functions) but they are not so relevant to my point. Which is, that ordinary non-faulty equipment that might be connected to ordinary circuits (I am not talking about multi-megawatt induction furnaces or electrode boilers!) will not under normal conditions present a DC resistance of tens of kilohms to a 250V IR test L+N->E. Unless you can find real-world examples.
E2A OK, OK, I've thought of one - old sheathed heating elements that have not been used recently. Leave a Baby Belling from the 80's plugged in but switched off only in the line, and you'll get a low reading due to moisture absorption in the mineral insulant. They used to fail PAT IR tests regularly, although not often at 10k . Run them for a few hours, at first the value falls as the moisture condenses at the cold tails, and then they will usually come up fine once it's evaporated again. But they are pretty easy to spot and unplug.
Last edited: