Slashdot Mirror
iPhone 7 Plus Makes Hissing Sound Under Load, Some Users Complain (businessinsider.com)
Several commendable users are complaining that their iPhone 7 Plus handsets are making a "hissing" noise especially when they do some heavy weight work. Some users note that this issue extends to the iPhone 7 as well. BusinessInsider reports:Stephen Hackett, cofounder of podcast network Relay FM, tweeted that his iPhone 7 Plus "makes terrible noises when under load," and shared an audio clip of the noise. TechCrunch writer (and former Apple employee) Darrell Etherington responded that his "brand new, just-unboxed [device is] doing the same thing right now." It sounds like the problem isn't affecting all devices, and it's not immediately clear what's behind it. Hackett said on Twitter that Apple will be replacing his device with a new one, which suggests it's a defect rather than just an unexpected quirk of the new smartphone's design. There's some speculation out there as to what's causing it - but nothing concrete yet. Engadget's Jon Fingas suggests it could be "coil whine," a process where electronics make an unintended noise while working, for example.
I would guess they chose an inductor too small and it is vibrating.
It's the sound of your soul being sucked into the device. That's why it's noticeable on a "brand new, just-unboxed device". Should go away after a few days, once you are completely soul-less.
I eat only the real part of complex carbohydrates.
That's actually an explanation for their use of the word courage. It takes courage to release a product full of design flaws, masked by removing features you can't get working.
The CPU gets hot when in use. To overcome thermal noise, the voltage on the CPU goes up a little.
There is an expected tolerance band for operation, and the control of the voltage on the CPU and ram has upper bounds for very good reasons.
I am suggesting that the normal operation of this voltage regulation under computational load results in increased bus noise due to a saturated ground, and that apple considered this acceptable because most users will just be listening to mp3s, or playing casual crap on Facebook, and not taxing the system this way, making the issue statistically ignorable.
The way you deal with signal bleed on a device that cannot be earth grounded is to have a very large conductor inside that serves as a floating ground. Usually this is sandwiched inside the PCB as a good thick copper layer. Apple wants a device that is practically lighter than air, and thin as a straight razor. Copper is pretty heavy, and extra layers inside the PCB add thickness. Both are things the idiots, I mean, "geniuses" in Cupertino think are trendy to do away with. As a consequence, I expect the grounding layer to be thinner than what is actually needed for the proper operation of the device at heavy load, resulting in ground saturation. When the ground saturates, coupled with a hot CPU from heavy load, the regulator pumps up the voltage to try to assure reliable signals are being generated. This adds to the problem, because now more heat is being added and the ground is already saturated, so rf noise leaks everywhere. Throw in a densely packed PCB, where lots of devices will pick up the noise, and you have a recipie for early component failure.
All the devices are working within design, but the design is poorly considered.
Rather than admit that the design is poorly considered, due to the absurdity of trying to make a high performance device that thin, I expect apple to blame users for overloading the phone instead. The noise only happens when the system is taxed, because it was designed to play on facebook, not number crunch.
I expect the engineers decided that transient loads of 100% were acceptable because it takes time to saturate the ground, and most things a user will do won't saturate the CPU like that.
For reference, most CPUs run between 1.2 and 1.5 vdc, with subtle changes up and down based on activity and temperature. This is normal, expected operation. In this case, the ground saturates when the CPU sits at 100% for a long time, and the voltage sits at 1.5v, with transient voltages from rf buildup pushing components outside that from the saturated ground.
Hypothetically.