AC electrical leakage from Surface Pro 3!

[Alphas]

When the power cable is connected, AC current flowing from the power brick on surface pro 3 will cause electrical leakage.



The voltage here is 240V, enough to kill you.   

The orange light from the test pen shows AC current  flowing through the whole case!

The power brick has 3 pins plug, but the cord is only 2 pins, that means there is no earth wire into the surface pro 3 power cable. M$ should have made a earth wire into the power brick and grounding the metal chassis.

RMA is useless since a replacement adapter will still have same leakage problem.   

[fubar.gr]

The power brick has 3 pins plug, but the cord is only 2 pins, that means there is no earth wire into the surface pro 3 power cable. M$ should have made a earth wire into the power brick and grounding the metal chassis.

The mains earth prong is usually tied to the output's negative. You can test that with a multimeter.

Use the test light at the earth prong of your wall socket. Does it light up there too?

[andtfoot]

As far as I know, it would probably be leakage in the capacitor between the primary and secondary of the power supply.
It would be low current, so unlikely to cause any problems, but still a bit concerning for a new power supply.

[wraper]

As far as I know, it would probably be leakage in the capacitor between the primary and secondary of the power supply.
It would be low current, so unlikely to cause any problems, but still a bit concerning for a new power supply.

99% it is a cause, you likely have a half mains voltage there, not full. That's OK BTW, most such PSUs will do that.
Didn't found any photos of surface power brick having 3 pin plug, only 2 pin. However if it have the third pin and it is connected to the output GND, then just buy a cable which have a ground for a few bucks.

[Richard Head]

If the earth pin isn't connected then as mentioned it's probably at half the supply voltage and is due to the two mains Y-class capacitors that are from L-E and N-E. When the earth isn't connected they are in series and act as a capacitive voltage divider thereby splitting the mains AC in half. The Y-caps are normally 4.7nF or less and result in a 0.8mA RMS leakage current. It won't kill you but may make you spill your coffee!
I think that this situation sometimes occurs due to incompatible two pin plugs and two pin sockets.

Dick

[deephaven]

https://www.eevblog.com/forum/chat/earthing-rant/

[AndyC_772]

So what if you can measure a voltage? Without also measuring the source impedance, you have no grounds whatsoever to claim that the product is actually or potentially harmful.

It's current through the human body that's harmful. Put your multimeter in AC current mode, and measure between the case of the computer and earth. Let us know, quantitatively, what you measure.

It's very common indeed to have a high - but not infinitely high - impedance path between the primary and secondary sides of a PSU. Provided, that the leakage current is lower than that prescribed by the applicable safety standard, there is no issue.

[amyk]

It's current through the human body that's harmful. Put your multimeter in AC current mode, and measure between the case of the computer and earth. Let us know, quantitatively, what you measure.

Or more precisely, energy (which is a combination of current, voltage, and time). A static discharge can be several kV, with peak currents several tens of A, but it's not lethal because of the extremely small timescale involved.

[rsjsouza]

One other detail: usually Neon lamps can be discharged by a nearby electric field. Does the lamp lights up if you simply put it close to the fluorescent backlight - i.e., not touching anything but your finger on one side?

Mine does light up when I do the same with several devices that have backlight - the worst is my old Palm Lifedrive.

[Simon]

So what if you can measure a voltage? Without also measuring the source impedance, you have no grounds whatsoever to claim that the product is actually or potentially harmful.

It's current through the human body that's harmful. Put your multimeter in AC current mode, and measure between the case of the computer and earth. Let us know, quantitatively, what you measure.

It's very common indeed to have a high - but not infinitely high - impedance path between the primary and secondary sides of a PSU. Provided, that the leakage current is lower than that prescribed by the applicable safety standard, there is no issue.

Yep, what will happen when you pick up a scope probe and touch the end and suddenly find a perfect sine on the screen, are you going to shit yourself ?

[deephaven]

That impedance is a lot higher than that provided by a 2 wire input mains power supply. I don't like 'feeling' the voltage.

[Simon]

I don't think the Op said he could feel it. I am surprised you can get that screw driver to light but then i think I had fun with one when I had one and was surprised how easily they are lit up. surely it would never have passed CE.

[deephaven]

I don't think the Op said he could feel it. I am surprised you can get that screw driver to light but then i think I had fun with one when I had one and was surprised how easily they are lit up. surely it would never have passed CE.

It's been my experience with any such power supply that you can feel it. Shops with product plugged into chargers always give 'tingles'.

[Simon]

Well they should not, my laptop charger has never done any such thing.

[IanB]

It's normal for power supplies to let mains voltage through like this. I've mentioned before that my iPhone makes voltage detectors go crazy when it is on charge. I touch the detector to the phone and it flashes and beeps just like it does on mains wiring.

[Simon]

eh, I still remember the early shoddy SMPS for mobile phones that totally screwed AM radio reception for 10's of meters around, I don't know if they are any better now.

[Rufus]

So what if you can measure a voltage? Without also measuring the source impedance, you have no grounds whatsoever to claim that the product is actually or potentially harmful.

Shitty earthless leaky PSUs are a pain. Yes you can feel the leakage of bad ones, you can measure 10's of volts with respect to earth using a 10M multimeter.

A laptop I have measures around 80vac with respect to earth. In a darkened room I can see small sparks if I earth the shell of a USB cable. Using it with a USB ICE those sparks could be applied directly to the pins of the target processor depending on the order of pin connection. I always had to run a separate earth from the laptop to the target to avoid the risk of killing it. The laptop I mostly use now has a 3 pin power brick and measures about 1k between laptop ground and mains earth - much less of a pain.

So yes the leakage won't kill you (as long as you trust the capacitor responsible for it) but I would rather it wasn't there.

[Precipice]

Well they should not, my laptop charger has never done any such thing.

I'm guessing you tend to wear shoes, or at least socks, most of the time?
Most things tingle when I'm barefoot, and don't when I'm not.

[Simon]

Well they should not, my laptop charger has never done any such thing.

I'm guessing you tend to wear shoes, or at least socks, most of the time?
Most things tingle when I'm barefoot, and don't when I'm not.

good point, I remember kneeling on a stone floor and hooking up the house earth wire to the mains water pipe. I ended up having to kneel on a peice of cardboard because of the leakage from the wiring of the whole house.

[denelec]

But sometimes, the leakage can be serious...
http://www.theage.com.au/victoria/australian-woman-electrocuted-by-usb-charger-20140627-zsnm7.html

I once measured 60V between earth and the output of a cheap ($3.99) USB charger.  The current wasn't enough to electrocute me but I could clearly feel it. 

[DanielS]

A static discharge can be several kV, with peak currents several tens of A, but it's not lethal because of the extremely small timescale involved.

With the tiny capacitance usually involved in a static discharge, most of the energy is lost simply striking the arc and then you have a capacitive voltage divider between the ESD source and your body further reducing the net effect. Also, with such fast rise/fall times, most of the energy is subject to skin effect and this helps with making the mother of all ESDs, lightning strikes, survivable with some luck.

[Shock]

I'm suspicious of this thread and if this would have been tested a long time ago.
Would the magnesium back on the Surface Pro 3 even be conductive?

This thread was also posted 2m29s after Alphas registered on the forums.

[Simon]

Yes it's interesting that he has not come back and only made the one post since registering. Mind you I think whatever the intent "it" failed.

[Rasz]

But sometimes, the leakage can be serious...
http://www.theage.com.au/victoria/australian-woman-electrocuted-by-usb-charger-20140627-zsnm7.html

I once measured 60V between earth and the output of a cheap ($3.99) USB charger.  The current wasn't enough to electrocute me but I could clearly feel it. 

http://www.fakt.pl/Pania-Grazyne-poparzyla-ladowarka-od-komorki-,artykuly,99478,1.html
http://www.se.pl/multimedia/galeria/57078/108893/adowarka-spalia-mi-twarz/

dont click on photos if you are sensitive, raw meat

[amyk]

About feeling it or not: the threshold of perception varies for different people, and for double-insulated devices is limited to 250uA (although in practice good PSUs will have far less). Some people can feel this if they're well-grounded.

[dentaku]

Yes it's interesting that he has not come back and only made the one post since registering. Mind you I think whatever the intent "it" failed.

Maybe the original poster just wanted to see who would be the first stuck in the 90's person to type M$ instead of MS.

[Simon]

Oh sorry was this topic about a microsoft product ?

Now under the threat of open source software i guess prices have dropped a bit and at last they "had" their act together with windows 7 but they seem to have lost it in traditional M$ style with windows 8 err 8.1 as 8 was uncontrollable.......

[Shock]

MS alternates releases with what the public want and what they want, so in this case a wait for Windows 9.

[dentaku]

Oh sorry was this topic about a microsoft product ?

Now under the threat of open source software i guess prices have dropped a bit and at last they "had" their act together with windows 7 but they seem to have lost it in traditional M$ style with windows 8 err 8.1 as 8 was uncontrollable.......

Looking back it looks the original poster is the first person to write M$ so I guess Alphas is the one suck in the 90's
MS has changed allot in the last few years and the Surface Pro 3 is the first tablet I can say I would actually like to buy.

I wonder just how conductive the magnesium shell is?

[Alphas]

Yes the power brick on Surface pro 3 is a 2 pins plug, even my old fujitsu laptop adapter has it earthed. I can get a shock just by touching the mag alloy case on the Surface pro 3.




The voltage reads over 110V, I certainly dont feel good having such high voltage in my body.

[Rufus]

The voltage reads over 110V, I certainly dont feel good having such high voltage in my body.

How bad do you feel wearing nylon socks which can generate thousands of volts in your body?

[denelec]

110V can be lethal or harmless.  So does 1000V.  But you'd have to be very unlucky to hurt yourself with 10V.
It depends on the intensity of the current that goes trough your body. And also the duration.
If the impedance of the source limits the current below 1mA, you won't feel much.
If it limits the current below 10mA, you won't be in danger.
http://www.hubbellpowersystems.com/literature/encyclopedia-grounding/pdfs/07-0801-02.pdf

With only a voltmeter you can't determine if the 110V you measured is safe.  You must measure the maximum current this 110V can generate.