Floating earth

Me! One probe on mains earth and one to my index finger. I suppose to be accurate I was representing the "real" earth and would have got the same result had I leant out of the window and shoved the probe into a flower bed.

Unless I'm simply full of electricity ;-)

Was it a high impedance digital meter - if so the reading is probably false and caused by your body picking up mains hum. Do it with an 'old fashioned' analogue Avometer and get the same results and I'll be convinced.

Andrew

I can't be totally sure, but nothing you have posted makes me think there is anything dangerous or particularly abnormal about this installation. The amp is fed with by a double insulated PSU with no earth, right?

Good point, overlooked that one. But the shock certainly felt more than a bit of inductive pickup. Same meter at home gets me 5v in the same setup, rising to 15v if I grasp a mains cable in my hand, quite a bit lower than earlier.

Yep, that's the one.

Yeah, but he said he got a belt off it, which implies something other than just a duff reading of a phantom voltage. I personally can't feel anything less than 80v; for me, that's when a tingle first becomes perceptible.

Unless you're really thick-skinned, I'd be surprised if you couldn't feel 80V from a low-impedance source.

Oil-cooled welding transformer low impedance enough for you? The point I was trying to make was that I can't feel anything *at all* below 80v (provided my hands are dry) YMMV. I have a 3A mains variac as well. Maybe over xmas I'll try out some higher voltages and see where it becomes uncomfortable. We'll see... I don't recommend others experimenting, though!

Flowers to the crematorium please.

;-)

What sort of audio cable? Balanced (two conductors plus screen) or unbalanced (single plus screen)?

I'd want to know what the other end of the screen was connected to.

If unbalanced you will get hum if you connect the mains earth to the signal screen at more than one point.

Unbalanced, jack to phono, a short one. The amp's powered off a laptop PSU brick so unearthed and had the faintest signs of hum when on, but as soon as the lead was plugged in it got worse. Usually only suffer this when there's a 10/20m run involved. When it's terminated it'll shut up but, being for laptop presentations, that's not always.

It really wants better equipment but this is the bodget and scarper end of the market and the powers that be don't know balanced from unbalanced from an acre of danelion and burdock.

Is the source on a different mains phase? Put a DI box, with earth lift, in the way.

It's not uncommon to have high value resistors between line and chassis on some designs - which will show volts on the chassis if the true earth isn't good. Enough to feel or measure - but not enough to do you damage. Unless you drop the amp on your foot in surprise...

You should tell him about balanced. Would allow even cheaper cable for the interconnect. ;-)

what is even MORE common is an RF filter with caps between line and neutral to the chassis.

Puts the chassis and often the equipment ground at 110VAC with respect to a proper earth. Enough as we discovered to blow the serial port of a system connected to the offending laptop.

Or a combination of resistors and capacitors. ;-)

The volts don't matter - it's the current which does.

The serial cable should have connected both chassis together.

In message , "Dave Plowman (News)" writes

You can still get a lively tickle.

In a range of equipment I used to work with, the mains input had something like 0.01uf capacitors from the live and neutral to chassis. You certainly knew about it if the earth lead wasn't connected, and you 'offered up' the first interconnection with other earthed equipment. We found that when there were about 15 if these units all running together, the 30mA earth leakage trip used to pop.

See above. If you interconnect unearthed-chassis equipment, until the first connection is firmly made, you can certainly damage something fragile depending on which pin makes contact first.

It did eventually, *after* the shield had striped the signal pins and blown the input chip...

Nope, same twin socket!