https://arstechnica.com/gadgets/2018/10/helium-implicated-in-weird-iphone-malfunctions/ Skip to main content * Biz & IT * Tech * Science * Policy * Cars * Gaming & Culture * Store * Forums Subscribe [ ] Close Navigate * Store * Subscribe * Videos * Features * Reviews * RSS Feeds * Mobile Site * About Ars * Staff Directory * Contact Us * Advertise with Ars * Reprints Filter by topic * Biz & IT * Tech * Science * Policy * Cars * Gaming & Culture * Store * Forums Settings Front page layout Grid List Site theme light dark Sign in Don't get high on your own supply -- Helium implicated in weird iPhone malfunctions This is one of the stranger stories we've heard of smartphones malfunctioning. Ars Staff - Oct 31, 2018 6:30 pm UTC Helium-filled balloons. Enlarge / Helium-filled balloons. Warren Rohner reader comments 126 Reddit user harritaco discovered something rather unusual about the iOS devices used at their place of work. iPhones and Apple Watches stopped working unexpectedly, completely locking up and recovering only days later, sometimes suffering long-term harm. The failures appeared to coincide with the installation of a new MRI (magnetic resonance imaging) machine. MRIs use powerful magnetic fields and helium-cooled superconductors, and something about the presence of this new machine was upsetting the Apple hardware. That magnets can be a problem for electronic devices is no big surprise--they can damage magnetic media, confuse compasses, and induce electric currents in harmful ways--but surprisingly, it's not the magnets that seem to be the problem this time--it's the helium. The iPhone user guide warns that proximity to helium can impair functionality and that to recover, devices should be left to air out for a week or so in an environment far away from the rogue helium. Harritaco discovered that, during installation of the MRI machine, some 120 litres of liquid helium leaked and vented into the environment. This created a relatively high helium concentration, and any Apple hardware exposed to that helium stopped working. To test this hypothesis, harritaco conducted some experiments in which an iPhone was put in a sealed bag of helium; after a few minutes, it stopped working. This leaves one big question: why? Redditor captaincool offered a plausible theory. Smartphones contain microelectromechanical systems (MEMS): tiny mechanical systems that are integrated into chips. These mechanical systems are used for things like the gyroscope and accelerometers. These devices operate in tiny cavities within the chip (they need space to flex and vibrate), with those cavities either evacuated to create a vacuum or filled with a gas with known properties. This means that the MEMS devices have a seal to plug the gap where the air was pumped out (or another gas pumped in). While those seals are able to keep most air constituents out, they're permeable to the very tiniest molecules. In particular, both diatomic hydrogen and monatomic helium can permeate the seals. At normal atmospheric concentrations of hydrogen and helium, this is no big deal, but, when exposed to high helium concentrations--such as the concentrations experienced by harritaco--enough helium could diffuse across the seal to render the MEMS components non-functional (at least temporarily, until the helium escapes). Our guess is that a MEMS oscillator used to generate a clock signal for the phones is misbehaving after exposure to helium, hence the system-wide failure. That's probably the weirdest way to make a smartphone stop working that we've heard. As for why non-Apple devices appear to escape without harm? They might use different seals or perhaps aren't using MEMS devices in such critical roles. Promoted Comments * norton_I Ars Praefectus jump to post redleader wrote: I don't know about this type of mems oscillator, but in telcom a common failure mode during testing is "snapdown" where a unit has too much electrostatic actuation and comes into contact with the opposing electrode and housing. At that point the attraction becomes vastly stronger due to the contact and the device is usually permanently fused. The hypothesis here (and which explains the reversibility) is that the MEMS oscillators are actually vacuum sealed for low loss, rather than filled with nitrogen or air like most accelerators (which want the damping to avoid oscillation). The helium is diffusing in and increasing drag (relative to vacuum) enough to kill the oscillator. Once the helium diffuses out they should work again. Helium isn't likely to be more than a fraction of a percent even in the situation referenced here, so I doubt it would have any effect on devices that are air filled. This makes the most sense. My speculation is this is because MEMS oscillators in phones are likely optimized for lowest possible power. Therefore, the loop gain would be as low as practical, and small increases in dissipation could cause it to drop below the oscillation threshold. Oscillators for less power constrained applications might have stronger drive capability and be less sensitive to this. That could be one reason some devices are more susceptible than others. 4086 posts | registered 4/30/2001 * tbenz9 Wise, Aged Ars Veteran et Subscriptor jump to post I saw this when it was originally asked on Reddit, I'll admit I didn't believe it then. I believe it now, Ifixit has written a very in-depth analysis that is definitely worth a read. https:// ifixit.org/blog/11986/iphones-a ... to-helium/ 102 posts | registered 10/14/2012 Promoted Comments * norton_I Ars Praefectus jump to post redleader wrote: I don't know about this type of mems oscillator, but in telcom a common failure mode during testing is "snapdown" where a unit has too much electrostatic actuation and comes into contact with the opposing electrode and housing. At that point the attraction becomes vastly stronger due to the contact and the device is usually permanently fused. The hypothesis here (and which explains the reversibility) is that the MEMS oscillators are actually vacuum sealed for low loss, rather than filled with nitrogen or air like most accelerators (which want the damping to avoid oscillation). The helium is diffusing in and increasing drag (relative to vacuum) enough to kill the oscillator. Once the helium diffuses out they should work again. Helium isn't likely to be more than a fraction of a percent even in the situation referenced here, so I doubt it would have any effect on devices that are air filled. This makes the most sense. My speculation is this is because MEMS oscillators in phones are likely optimized for lowest possible power. Therefore, the loop gain would be as low as practical, and small increases in dissipation could cause it to drop below the oscillation threshold. Oscillators for less power constrained applications might have stronger drive capability and be less sensitive to this. That could be one reason some devices are more susceptible than others. 4086 posts | registered 4/30/2001 * tbenz9 Wise, Aged Ars Veteran et Subscriptor jump to post I saw this when it was originally asked on Reddit, I'll admit I didn't believe it then. I believe it now, Ifixit has written a very in-depth analysis that is definitely worth a read. https:// ifixit.org/blog/11986/iphones-a ... to-helium/ 102 posts | registered 10/14/2012 reader comments 126 Advertisement Channel Ars Technica - Previous story Next story - Related Stories Today on Ars * Store * Subscribe * About Us * RSS Feeds * View Mobile Site * Contact Us * Staff * Advertise with us * Reprints Newsletter Signup Join the Ars Orbital Transmission mailing list to get weekly updates delivered to your inbox. Sign me up - CNMN Collection WIRED Media Group (c) 2024 Conde Nast. All rights reserved. Use of and/or registration on any portion of this site constitutes acceptance of our User Agreement (updated 1/1/20) and Privacy Policy and Ars Technica Addendum. Ars may earn compensation on sales from links on this site. Read our affiliate link policy. Your California Privacy Rights | [privacyopt] Do Not Sell My Personal Information The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of Conde Nast. Ad Choices