Leakage Current to Earth Through Mains-Powered Multimeter?

[Galgso]

I was using my isolated high-voltage supply earlier and I had this strange issue where I would feel tingling upon touching some other part of the circuit and the earthed metal case of the supply simultaneously. When I connected the probes of my battery-powered handheld digital multimeter between the highest voltage output and the metal case with the supply turned on I measured just over 100 volts DC. I was able to finally narrow down the issue to my bench top multimeter (Fluke 8010A). Removing the bench top multimeter dropped the voltage down to under half a volt. I can measure around 100v DC between the metal enclosure and the isolated DC output of the supply whenever the bench top meter is connected to the circuit regardless of whether the meter is powered on or not. I can actually turn off the outlet and as long as the meter's ground pin on the plug is connected to earth I will get this voltage. I can even measure the same voltage when a single probe from the bench top meter is connected to the 0v output on my supply. It seems like there's leakage current between the meter's inputs and mains earth.

I measured the resistance between ground on the plug and the negative input of the multimeter and it was greater than 20M ohms (my handheld DMM cannot measure above 20M ohms). Any thoughts on this? I can probably get access to a PAT tester on Monday if that would help but I really did not expect the multimeter to fail this way.

 

[Nigel Goodwin]

The multimeter almost certainly hasn't 'failed', it sounds like it's working perfectly.

There are two types of mains appliances, Class I (earthed) and class II (not earthed), the meter will be class II, as an earthed multimeter would be a VERY serious safety hazard.

Because Class II devices aren't earthed they can gain a high static charge, to prevent this there are special safety components connecting from chassis to the incoming mains - these are usually a high value resistor and a capacitor in parallel, and are vitally important, and drain the static charge to ground (via the mains).

The result of this is that the chassis of the appliance floats at about half mains potential, as a VERY low leakage current, hence the small tingle you can feel.

Perfectly normal, nothing to worry about - and certainly something you shouldn't try and 'cure', as you would be making it dangerous.

 

[Galgso]

The multimeter almost certainly hasn't 'failed', it sounds like it's working perfectly.

There are two types of mains appliances, Class I (earthed) and class II (not earthed), the meter will be class II, as an earthed multimeter would be a VERY serious safety hazard.

Because Class II devices aren't earthed they can gain a high static charge, to prevent this there are special safety components connecting from chassis to the incoming mains - these are usually a high value resistor and a capacitor in parallel, and are vitally important, and drain the static charge to ground (via the mains).

The result of this is that the chassis of the appliance floats at about half mains potential, as a VERY low leakage current, hence the small tingle you can feel.

Perfectly normal, nothing to worry about - and certainly something you shouldn't try and 'cure', as you would be making it dangerous.

Thanks for the respone. The multimeter itself has a plastic case and I've never had any issues using with the meter itself. The issue is that when the multimeter is used to measure voltage coming from a different supply with a grounded metal case current seems to be finding a path to earth through the multimeter's inputs. Normally I wouldn't be able to measure any voltage greater than half a volt between earth and the supply's outputs but with the meter connected to the supply I can measure 100v DC between the supply's isolated outputs and the earthed case. This seems to indicate the meter is affecting the isolation between mains and the supply's outputs. Is this actually a problem or do you think it is probably something that can be ignored? I assumed that the inputs of the multimeter are completely isolated with no way for any measurable current to flow to earth through them.

 

[Nigel Goodwin]

Thanks for the respone. The multimeter itself has a plastic case and I've never had any issues using with the meter itself. The issue is that when the multimeter is used to measure voltage coming from a different supply with a grounded metal case current seems to be finding a path to earth through the multimeter's inputs. Normally I wouldn't be able to measure any voltage greater than half a volt between earth and the supply's outputs but with the meter connected to the supply I can measure 100v DC between the supply's isolated outputs and the earthed case. This seems to indicate the meter is affecting the isolation between mains and the supply's outputs. Is this actually a problem or do you think it is probably something that can be ignored? I assumed that the inputs of the multimeter are completely isolated with no way for any measurable current to flow to earth through them.

I've already fully explained the reason - it's perfectly fine, working safely and exactly as intended.

If you don't understand the basics, perhaps you shouldn't have a multimeter?.

 

[Tony Stewart]

I've already fully explained the reason - it's perfectly fine, working safely and exactly as intended.

If you don't understand the basics, perhaps you shouldn't have a multimeter?.

Without specifics on DC & AC current levels to divert impulse from the HV boost converter using a common mode noise filter to earth and the capacitor leakage resistance and the meter's leakage resistance , how can you be certain your explanation was complete. Perhaps someone should moderate the moderator's remarks..

BTW you can feel a similar tingling leakage current on any laptop floating output to earth ground or grass between the ~19V DC or 0V DC to your knee or wrist (if you had bare feet on the grass). This is due to the AC CM Pi filter diverting switcher impulse noise to PE on the AC input and limited to safe leakage current by the 1 to 4.7 nF Caps to PE and the coupling capacitance of the pulse isolation transformer CM coupling capacitance. So there is a regulated safe current to filter out noise to the grid while creating a loop leakage current between DC and you with coupling by a single wire capacitance to you and earth by the PI cap filter values caused by the input switcher noise. When connected to a laptop, it is safest to use this isolated SMPS charger , but since only the input is PE grounded, you still can feel a tingle from the laptop case with damp feet to earth or touching any earthed device and your wrist or knee touching the corner of the laptop case. The smaller the area, the higher the current density and tingle feeling. (Even this is a not a thorough answer, lacking a block diagram)

 

[Galgso]

Without specifics on DC & AC current levels to divert impulse from the HV boost converter using a common mode noise filter to earth and the capacitor leakage resistance and the meter's leakage resistance

My high voltage supply is homemade and is linear with no switching of any sort. It only uses voltage doublers and rectifiers supplied by 240V 50Hz AC from the output of two mains step-down transfomers connected back-to-back.

I've just tried connecting the same multimeter to a mains transformer connected to a half wave rectifier. In the photos, the black probe on the isolated side goes to the common input of the bench multimeter. No other inputs on the multimeter are connected to anything and the meter is not powered on. With the meter's plug connected to the socket, I get higher 6v with respect to mains earth on the output.

With the meter not plugged into mains, I see much less voltage which indicates less leakage. I am mostly accustomed to using handheld battery-powered multimeters and I expected my recently acquired bench multimeter to act like one. I suppose I was wrong. I was just concerned because I was able to read 100v between the output of my supply and earth with the meter connected where I would have expected the leakage between the meter inputs and earth to be small enough to drop less than a volt on a DMM's internal impedance. Thanks Nigel and Tony for your helpful replies.

 

[Nigel Goodwin]

Much of your home electronics is Class II, such as TV, VCR?, DVD, Surround Sound System etc. - and the fact that many of these are connected together via cables, means that the tiny leakage is multiplied, and thus can give you a slightly higher 'tickle' when you touch them.

One problem of this is the aerial system, aerial systems are commonly earthed (and it's an actual requirement in parts of the EU) - so you can unplug the aerial from your TV (holding the TV in one hand) while pulling the aerial plug out with your other hand - this gives a nice direct leakage connection from hand to hand, for an even better 'tickle' - you might even see a tiny spark when you plug the aerial in.

 

[ahsrabrifat]

I was using my isolated high-voltage supply earlier and I had this strange issue where I would feel tingling upon touching some other part of the circuit and the earthed metal case of the supply simultaneously. When I connected the probes of my battery-powered handheld digital multimeter between the highest voltage output and the metal case with the supply turned on I measured just over 100 volts DC. I was able to finally narrow down the issue to my bench top multimeter (Fluke 8010A). Removing the bench top multimeter dropped the voltage down to under half a volt. I can measure around 100v DC between the metal enclosure and the isolated DC output of the supply whenever the bench top meter is connected to the circuit regardless of whether the meter is powered on or not. I can actually turn off the outlet and as long as the meter's ground pin on the plug is connected to earth I will get this voltage. I can even measure the same voltage when a single probe from the bench top meter is connected to the 0v output on my supply. It seems like there's leakage current between the meter's inputs and mains earth.

I measured the resistance between ground on the plug and the negative input of the multimeter and it was greater than 20M ohms (my handheld DMM cannot measure above 20M ohms). Any thoughts on this? I can probably get access to a PAT tester on Monday if that would help but I really did not expect the multimeter to fail this way.

Check for RFI suppression caps (Y-class) inside the Fluke. These can cause small leakage even when the unit is off.