I can’t imagine having 110v inside the device. What is a “figure eight plug”?
Here is the pinout for the connector I’m familiar with.
The device doesn’t have 120v in it, unless the PSU is damaged. The figure 8 plug:
That plugs into the power brick, and the power brick plugs into the device. The figure 8 plug isn’t polarized, so it’s possible to swap the hot/neutral pins. That doesn’t matter if the PSU is working correctly, but it might if there’s something wrong with it.
This is a situation where a multimeter is really handy. If you don’t have one, get one. (Simple volts/ohms/amps.) Put it on a high AC setting (at least 120). Test between the place you get a buzz and a separate ground. If nothing big try a modest DC setting and the ground. That should tell you if there’s a voltage leak. (Note that DC will show up on the AC setting with a different value but it shouldn’t be anything well above the charger’s output voltage.)
As noted, the problem is likely from the charger. You can test across the barrel plug on the adapter. And AC or DC voltage above the adapters listed output is bad.
A bad charger should have it’s cord cut immediately and then thrown out. It can damage you and your device.
That brings up the question, “What is ground” in a two wire house system. With a three wire system, that plug has no connection to ground.
Even the neutral wire can be dangerous. If when you touch it, you are providing a better path to ground than its termination, you can be shocked. In a properly installed system this should not happen. But never assume.
I don’t have the device with me right now, but in light of the very helpful explanations given above, I wouldn’t expect that. The connector between the charger and the tablet is symmetrical.
Sure, that’s the obvious and easy solution to the problem, but I was simply curious what was going on there. I have learned a lot in this thread already, but I want to go and try out ftg’s suggestion, too.
I’ve have 2 surface pro laptops, both do this.
See Microsoft explanation
Turning the lead from the mains to the brick (figure 8 connector) by 180 degrees does stop/reduce the effect.
Essentially it is a known potential (pun?) situation and not dangerous. Well I’m still alive… 
As noted, the ground socket of an outlet is a good ground. Another is a (dry) faucet.
Of course you can have a two wire system. And you can have plastic pipe plumbing instead of metal pipes. But the two tend not to occur in the same house.
Does the tingling only occur when the person is grounded?
Yeah, you absolutely can.
It also happens as a failure of design on some of the cheap USB chargers - here’s an example where the circuit is built into a rechargeable lantern that also functions as a powerbank and a USB supply (but the same circuit can be found in some really cheap Chinese-made USB adaptors - this one is capable of outputting live mains voltage on the grounded shield of the USB socket (so the external metal parts of a device plugged into this could become live at mains voltage).
I don’t imagine this is relevant to the OP’s case, but I feel it’s important to point out that getting a dangerous shock from a USB charger is not limited to the freak accident/damage scenarios.
I have watched so many Big Clive videos of various death chargers that I’ve possibly become oversensitized to posts like the OP’s.
I am a bit surprised at Steven_G’s post that this is a common/known problem with the Surface Pro charger. I’ve never encountered this on any charger of mine and I’d consider it a defect. But apparently Microsoft does not.
So, to the OP: carry on, I suppose. The buzzing is apparently considered normal by Microsoft, and a replacement will probably not fix it.
It’s caused by a “Y” capacitor, and is considered normal, if annoying. Y capacitors are supposed to be fail-safe.
Supposed to be.
Looks like Y capacitors are required for common-mode noise suppression (and that, as you say, can cause this problem).
I wonder if this problem is getting more prevalent as chargers increase in wattage (especially due to high-density components like GaN), and decrease in size–which pressures them into leaving out a ground pin (which would solve the problem here).
Yep.
Here’s a schematic of a typical SMPS. Note the isolation barrier. CY1 is the “Y capacitor.”
If a Y capacitor is not included, the small capacitance between the primary and secondary windings of the transformer will cause the supply to generate a significant amount of high frequency EMI, and will subsequently fail EMI/EMC testing. To make it pass the tests, engineers put a “capacitor” between the the primary and secondary windings of the transformer. The capacitance is chosen to be much larger than the transformer’s capacitance, hence “swamping it out” and lowering the frequency of the EMI. It is called a “Y capacitor” because it is safety-rated.
There’s a downside to including the capacitor, though.
If,
-
The output of the power supply is floating, i.e. none of the terminals on the output are connected to earth ground, and
-
one side of the capacitor is connected to the “hot” side of the mains, and
-
you are grounded (e.g. standing barefoot on a wood floor), and
-
you touch any terminal on the output of the power supply,
you may receive a shock.
And that’s the tradeoff: the larger the capacitance of the capacitor is, the easier it will be for the power supply to pass EMI/EMC testing, but the lower its reactance will be, and the greater the shock hazard.
As can be seen in the schematic, CY1 connects to COM. And the voltage of COM is always very close to one of the mains (AC2). If AC2 is neutral, you won’t receive a shock, because neutral is connected to earth ground at your home’s circuit breaker panel. However, if AC2 is “hot,” you might feel something.
If the plug on the power cord is comprised of just two, unpolarized prongs, then you have to try both orientations to figure out which one connects AC2 to neutral.
It would increase the complexity a bit, but it seems like a power supply could be designed to selectively connect the Y capacitor to the neutral pin. You would have to figure out which is the hot wire, which of course isn’t trivial, but is possible. Ironically, those hot wire detector sticks work by measuring the tiny leakage you get from capacitive coupling, which is essentially just the problem here. So maybe the answer is for the PSU to detect this leakage and try coupling to the other pin if the current exceeds some threshold (i.e., turn the plug around like the OP did)
Well, two-pong plugs are supposed to be polarized…
Once prong is wider than the other. But, maybe the manufacturers decided that the power supply must be safe regardless of the polarity of the prongs, so they made them with the same-sized prongs. I know Apple chargers with the two-prong adapter have identically-sized prongs.
Well, the other side matters also. No use polarizing the plug if the other end can be flipped. Polarized versions of the figure 8 plug (what are those things called, anyway?) do exist, with one end squared off. But they don’t seem very common.
Yea, trying to determine which is hot and which is neutral without a real ground connection is difficult. As you mentioned, you’d have to rely on a tiny bit of capacitive coupling to earth ground, which adds cost, size, and complexity to the circuit.
I don’t know why they don’t use polarized power plugs. Seems like it would be an easy way to ensure the Y capacitor is tied to neutral (assuming the outlet is wired correctly). I suppose you could also use a three prong plug, and simply not use the ground terminal for anything.
On Apple chargers, they have a removable adapter for either two-prong or 3.
The three-prong cord is grounded, while the two-prong is not (clearly) and also not polarized.
I almost always use the grounded extension cord, since it gives some extra room, but more importantly, gets the power supply away from the outlet, where it is likely to get bumped and damage the prongs, or fall out.
So the power cord will guarantee polarity, but the two-prong adapter will not. Interesting.