This vacuum had the most five-star reviews on Amazon, but then a minority of people started complaining and leaving one-star reviews, saying they were getting shocked.
The reactions from the fanboy crowd were non-sympathetic ("you're imagining it," "it's just static," etc.), and they mass downvoted every one-star review. Some of the people claimed they were getting shocked PRETTY HARD, too, as in it knocked them over. IIRC, even Shark responded saying that it was impossible for the vacuum to shock you.
I remember seeing this while looking at vacuums for my elderly mother who has a heart condition. To the electrical engineer in me, everything made sense!
I started defending every single one-star reviewer person, and I wrote my own very detailed review with photos of the handle. Soon, the tide turned, and more people started upvoting my review than downvoting it; and I even received personal "thank you" messages from the people who got shocked ("finally someone believes me").
It eventually got to the point where a year later in 2014, Shark released an updated model with no more metal strip. Seeing this, I bought the new model for my mother. I'd like to think I can give myself SOME credit for causing them to do that.
Not only that, the upvotes from that alone were a good part of the reason I became an Amazon Top Reviewer.
And later released a workaround to make the hardware work in a slower, more power consuming backwards compatibility mode so Intel can continue not properly supporting the hardware in Linux.
But sure, it was totally "Lenovo blocking Linux".
Lenovo initially released a BIOS with a goto statement added to jmp out of the disk controller mode setting (which is normally present and allows the user to change back to AHCI mode). Reverse engineers on the Lenovo forums discovered this modification while studying the disassembly.
Then, one user patched the BIOS and manually reflashed it using an SPI flasher and some soldering, and Linux worked just fine by detecting the drive.
Yes, the ideal solution is that Intel gives specs to OSS devs to build a driver (or even builds on themselves), but the path of least resistance is to have Lenovo unblock that BIOS setting, which amounts to changing one line of code, or 15 minutes of one engineer's time.
Because switching to AHCI mode with Windows installed would brick the machine (Yay, Microsoft), and Lenovo foolishly assumed that people who wanted to use Linux would buy one of the Linux certified machines instead (X1 Yoga and 60 series Yoga devices are all certified: https://support.lenovo.com/us/en/documents/pd031426 – but I'm sure that's just more proof that Lenovo hates Linux for some reason).
> the path of least resistance is to have Lenovo unblock that BIOS setting, which amounts to changing one line of code, or 15 minutes of one engineer's time.
Which is exactly what Lenovo did: https://forums.lenovo.com/t5/Lenovo-Yoga-Series-Notebooks/Yo...
There are tons of BIOS settings that could "brick" my machine though, it's a weak argument, really. My car didn't come with a lock on the hood.
"Which is exactly what Lenovo did"
Okay, you're acting like it didn't require any teeth pulling on our end to get them to release it (check out the 30+ page forum topics).
The hot water pipes were being used as the earth, and the electricity could leap across the pipe in bad weather.
I had a grounding issue and I used to get shocked by my tower all the time, this is pretty common.
It's the devices responsibility to design for that.
You can fix it by bonding the neutral to ground at the breaker box.
You are wrong. In the US the NEC requires this. If you bother to look inside your breaker panel you will find the neutral and ground wires are literally screwed down to the same bus bar.
Any path your body (or the soil outside) can provide back to the generator pales in comparison to the neutral tap provided by the transformer. Grounding to dirt is about static electrical discharge, not the voltage provided by the generating station.
The safety of providing a grounding wire is that metal parts of appliances must be connected to that ground. The ground wire itself is both bonded to neutral and connected to a metal grounding rod driven into the dirt. If any part of the appliance's internal works becomes energized and touches the metal casing the direct path to ground is designed to create a short circuit and trip the breaker immediately, rather than sitting there waiting for a human to come along and complete the circuit.
If ground weren't bonded and the ground dried out or the grounding rod weren't properly installed then the scenario you describe would occur. A single faulty appliance would back-feed voltage onto every grounded appliance in the building. So in fact bonding is what creates safety, not the other way around.
It could be different in US, but in modern European installations the PE and N bars are separate; older installations use common PEN; and even older ones don't even have PE. Each of the wiring schemes requires different approach and poses different hazards. There's no single recipe that would work on all of them.
My bottom line is: don't make arbitrary connections in the breaker box without proper knowledge and training; and don't give random electrical advice on the internet without knowledge of local codes.
In the US a single PE+N wire feeds into the building. All building wiring is separated as in Europe. They are tied together in only one location, usually wherever the main switch is.
Just like Europe, really old installs in the US don't have ground (PE). Some are retrofitted with ground like my house in SF. One thing people should never do is connect neutral to ground at a secondary point. It can be tempting when you don't have access to run a ground wire but creates a serious potential for safety problems. The way to handle that is to install a GFCI (RCD) plug and put a sticker on it stating "No equipment ground" - the outlets actually come with those stickers for that purpose :)
As far as I know in European systems where the building has separate PE+N feeds they are still bonded together, it just happens at the transformer. I could be wrong about that though!
For example in the UK neutral is traditionally grounded at the electrical substation, which serves several hundred homes, and all lines are separate / unbonded within the house.
In the most recent PME standard in the UK the earth and neutral are bonded back to the substation as the Protected Earthed Neutral but that is achieved on the supply-side, not the in consumer equipment.
OP's very first sentence clearly stated they are in India, where this is obviously dangerous advice because most devices (and piping) connect their surfaces to Earth, which is new Neutral.
Earth here was a metal rod into the ground, and I only found out much later they fixed the RCD issue by disconnecting the rod and I started getting electric shocks from touching my fridge while my crappy toaster was connected.
That's when I decided to not mess with electrics in India. It's dangerous enough as it is.
As for the 70V, I eventually managed to reach someone in the electrics company that understood my issue, and they told me a local transformer for the whole street was broken. And they didn't have any plan to fix it.
I got a multimeter and measured. The neutral was 70V and the phase 300ish. So the difference was roughly 230V but a switched power adapter would make this hover around 70 too (I believe an old coil based would probably been fine).
The Indian electrician didn't understand my problem because his only tool was a regular bulb with two exposed wires that he jammed into the socket.
"Look sir, light, power working"
sigh.