dervu wrote: ↑04 Aug 2022, 07:02
Chief Blur Buster wrote: ↑04 Aug 2022, 05:30
Some electronics have become so sensitive to interference, that they just blankout (screen goes temporarily blank, etc) or crash (bluescreen, etc) -- I know a computer that reliably bluescreens when a nearby 1980s vaccuum is turned on within 12 inches of the computer tower, because it's a like mini-EMP bomb. No permanent damage, just computer crash. Very common if you pulse something very close to an unshielded computer (i.e. computers with a glass side are essentially unshielded computers that are not EMP-proofed against simple domestic-appliance interference).
If you know what device makes it crash, maybe it would be fun experiment to find out if moving it further from PC would cause lag?
Polite note about quality of your question:
- To a newbie, this is a legit question of curiously
- To a Ph.D university graduate, this can feel like a stupid question to them
Because it's a game of S/N margins and every people have different S/N margins, what happens with an identical rig for Person 1 will not happen for Person 2.
It can be as little as a little more photons or fields / a little fewer photons or fields. Every location have very different ambient electromagnetics (e.g. entire spectrum beyond visible light), with infinite numbers of frequencies of very different strengths in every different location. It's not as simple as a single wavelength like WiFi or reddish-orange light or 88.7MHz radio wave or whatnot.
So trying to come up with reliable EMI reproduction that works in all locations and cases is hard, unless there's extremely defectively-designed equipment involved (like the LG 5K recall).
Since there is an infinite amount of ambient electromagnectic-radiation differences in every cubic nanometer of the entire universe (more dramatically different than two random cloud-precipitation ice crystals aka snowflakes, the old adage "no two snowflakes are alike") -- consequently, you can't reproduce reliably, unless you use overkill (e.g. nuclear EMP) or you intentionally cherrypick defectively-designed equipment (e.g. early LG 5K display) in order to get reproducibility between two different users in two different locations, despite identical rigs.
Simplified example: A lightbulb is emitting light, which is a bunch of electromagnetic radiation (photons). But you know the whole electromagnetic spectrum (radio waves and microwaves are longer wavelength than visible light, while x-rays and gamma-rays are shorter wavelength than visible light). In this analog continuum of electromagnetic spectrum, there's an infinite number of frequencies. You call WiFi 2.4 GHz, but it's actually an infinite number of frequencies between exactly 2400 MHz thru 2483.5 MHz, and a wireless modem inside a router creates a WiFi signal.
That's the microwave frequency range (that's why turning on a microwave oven often weakens your houses' 2.4 GHz WiFi signal). But 0 Hz thru infinite Hz is an infinitely long band of infinite number of frequencies -- and any random frequency has a stronger signal. A specific microwave signal (e.g. 2453.3 Mhz might be strong) while concurrently a specific frequency xray signal is medium-strong (1.943252352358239582395823 x 10^17 Hz), while a visible light signal is strong (whatever frequencies from whatever type your lightbulb is emitting, and how far you're from it, etc), etc.
If you've studied high school sufficiently, you begin to realize there's an infinite amount of possible electromagnetic-spectrum signals. An infinite number of different signal strengths along an infinite number of frequencies along the entire electromagnetic spectrum, and manufacturing variances may pick up / reject frequencies better (e.g. a different motherboard may be able to accept a 10% stronger electromagnetic-spectrum without failing, because its transistors/VRMs are better-binned specimens, or whatever).
If you graduated science/physics, you're familiar with the electromagnetic spectrum and how infinitely wide of an analog continuum it is -- and electromagnetic-spectrum can be zillons of simultaneous frequencies of randomly different strengths in every single cubic nanometer of the entire volume of the entire universe. Move a bit further to the side, you're farther from the lightbulb but closer to the radio tower next door and the radium emissions from the earth below, and slightly closer to the WiFi router, and the emissions emitted by your fridge's motor. This is hugely simplified -- but there are an infinite number of electromagnetic sources -- even being a millimeter closer to the star in the sky, could tip the domino on an ultra-threshold situation.
Infinite number of simultaneous electromagnetic wavelengths, infinite number of signal strengths for each, infinite number of concurrent combinations -- and due to inverse square law (away from sources = weaker = fewer photons per second received from that source for that specific wavelength of the infinite number of electromagnetic wavelengths) all of the strengths for all of these infinite number of wavelengths change (up/down) everytime you even move a single atom to the left or down or thataway. Every single location of the universe is completely different.
So even shifting a computer to the side -- you might not be able to reproduce a crash, or whatever, unless one specific electromagnetic-spectrum source is overstrengthed (e.g. like a nuclear EMP) or if the hardware is already so defectively weakly-rejecting EMI, that it's easy to sue smalltime pulses (like from a motor or an antenna) to reliably reproduce the same thing in all locations for all users. Overkill is your friend.
Some YouTuber could perhaps use some giant appliance motor (e.g. unshielded clothes dryer motor while plugging/unplugging starting/stopping/etc) only 1-2 inches away from a chip surface, and maybe semi-reliably create crashes. Who knows? Maybe a YouTuber will come up with reliable instructions, but it definitely won't be easy to reproduce -- since every single location point in the entire volume of universe is all completely different ambient electromagnetic regimes.
But there are a lot of semi-defective (weak EMI rejection) product on the market lurking, not quite as sensitive as the LG 5K recall, but sufficient to produce reliable-reproduction instruction. Unfortuantely, since there are thousands of computer products on the market, you are going to have purchase all of them (2 or 3 of each) and test all of them in many different locations, with the same "force an EMI crash" technique.
It's a common problem but it's not an easily reproducible problem because of binning differences (e.g. different chips will be a few % more/less tolerant) and because every single different point of universe is completely different ambient electromagnetic regimes.
And electronics may be more sensitive to certain frequencies than others, and it's just not possible to test EMI-resistance for every single frequency. Engineers can only do their best, to a baseline home. But you're possibly SOL if you are living outlier (e.g. computer is in outerspace, or next to a nuclear bomb testing site, or computer is only 10 feet away from a 500 kilovolt power line, or whatever). I'm being dramatic -- of extremes -- but in Earth of nearly 8 billion population, more than 1 million people are living underneath ultra-high-intensity EMI like a hospital scanner next door or house under half-a-million-volt power transmission tower, etc. So interference often creates electronics glitches of all kinds -- sometimes as mudane as a temporary screen blankout (one of the more common symptoms)
The number of EMI combinations is much bigger than the number of subatomic particles in the known universe -- so it just is impossible to perfectly reproduce EMI regimes. So we fallback to cherrypicking interference-sensitive equipment (defective manufactured), or cherrypicking extreme-interference-strengths, in order to write reliable reproduction instructions. Despite one-off-interference situations being extremely common.
If you flunked university, you won't understand this post.
If you passed university/college and understand the electromagnetic spectrum well (and how electricity, you'll get the idea.
Those in between might be interested in reading the basics --
https://en.wikipedia.org/wiki/Electromagnetic_spectrum
https://en.wikipedia.org/wiki/Electromagnetic_radiation
https://en.wikipedia.org/wiki/Electromagnetic_field
It will help people understand why it's far more difficult to create two identical EMI regimes (infinite number of frequencies being made perfectly identical) and making sure that both identical rigs have exactly the same tolerances (identical resistance/overlcocking behaviors), necessary to create identical reproducible EMI behaviors. It's infinitely harder than trying to create two identical nowflakes. You need use the weak-link effect to create reproducibility (e.g. ultrastrong EMI far beyond norms, or defective equipment like LG 5K, or both in combination), despite one-off common-ness.
