Helium Can Stop Your IPhone — Maybe Other MEMS, Too

Sometimes hacking isn’t as much about building something, it’s about getting to the root of a particularly difficult problem. [Erik Wooldrige] was facing a problem like that. He’s a system specialist at a hospital near Chicago. Suddenly a bunch of iPhones and Apple watches were failing or glitching. The only thing anyone could think of was the recent install of an MRI machine.

Sure, an MRI machine can put out some serious electromagnetic pulses, but why would that only affect Apple products? Everything else in the hospital, including Android phones, seemed to be OK. But about 40 Apple devices were either dead or misbehaving.

It took some detective work, but they think they know what was the cause. The MRI machine uses liquid helium to cool its powerful magnets. Turns out the helium had leaked and over 5 hours about 120 liters of liquid helium vented into the air. Helium is notoriously hard to contain because, like hydrogen, it is a tiny little atom even by atomic standards. It also expands about 750 times when it turns into a gas, according to the post’s analysis.

Gathering more data, they found that many of the phones would eventually recover and that all the devices were at least an iPhone 6 or an Apple Watch. So even older iPhones seemed to be immune. Some speculated that the helium is small enough to get into the MEMS devices like the accelerometer or gyroscope that is in most modern phones and affect its operation. But why would that effectively brick phones? And why wouldn’t that affect most phones Android or otherwise?

The best theory — and it seems plausible to us — is that Apple stopped using quartz crystals for the phone’s internal clocks. Instead, they are using MEMS oscillators from a company called SiTime. Supposedly the MEMS oscillators are smaller and work better at temperature extremes. If the mechanical clock element got gummed up with helium, that would explain all the observed evidence.

[Erik Wooldrige] reading about the issue on Reddit, did an experiment where he subjected an iPhone to helium in a plastic bag. Granted, this is a lot more concentration of helium than the hospital probably got. but they also had five hours of exposure. In the video, below, you can see Erik’s phone stopped keeping time just after the three-minute mark on the video, eight and a half minutes of exposure.

It turns out if you read the iPhone user’s guide it reportedly says:

“Exposing iPhone to environments having high concentrations of industrial chemicals, including near evaporating liquified gasses such as helium, may damage or impair iPhone functionality. … If your device has been affected and shows signs of not powering on, the device can typically be recovered. Leave the unit unconnected from a charging cable and let it air out for approximately one week. The helium must fully dissipate from the device, and the device battery should fully discharge in the process. After a week, plug your device directly into a power adapter and let it charge for up to one hour. Then the device can be turned on again.”

Apparently, SiTime also is aware of this problem and says its newer devices are “impervious to all small-molecule gasses.” But they admit older parts were not immune.

Unless you spend a lot of time blowing up balloon animals, this probably won’t affect you. Still, we thought it was an interesting piece of detective work and one of those things that you might remember in a few years when you have some wacky failure in your blimp fleet. Of course, we were supposed to be running out of helium, so if that were true, this problem would eventually take care of itself.




115 thoughts on “Helium Can Stop Your IPhone — Maybe Other MEMS, Too

        1. … and Apple is any better? Let’s all just agree that by this point just about every system is compromised by a myraid of bugs, faults, and backdoors. It’s just the world we live in, a sacrifice on the altar of technology.

          1. Desktop computers you assemble yourself from parts are still sort of okay (as long as you get an intel CPU you can use me_cleaner with to remove the IntelME backdoor). It’s mostly just “mobile” computers that are crap.

      1. Hmm fair question, can it ever happen can we get that satisfaction. Technology rolls on or rather superficial permutations adsorbing to the ideals of technology – as the contemporary philosophers the Rolling Stones appropriately offered a titbit oddly reminding me of the Killer’s lyrics too ‘.. are we human or are we dancer..’ ?
        The way some anonymous nickname cowards bully yet stupidly easily betraying their facile cognition there clearly is an even lower organic archetype indistinguishable from mere non-organic mechanisms and so easily stepped on, good thing for time being we are nice guys but, as they are superficial jealous incompetents just like shooting those fish in a barrel – if one had that personality trait that is :D

    1. Vacuum chambers are also used to get the little bubbles out when gluing a new touchscreen digitiser to an LCD display. Maybe a lot of phone repair places already have them. JUST COINCIDENCE???

    2. Hmm, that’s an interesting point and if it doesn’t work or takes too long for overall utility/stability then suggest cycle the chamber its in with dry N2 and up to approx one atmosphere above STP and also cycle the temp too from zero C to warming it up as well maybe 50C or whatever the chip and system specs are maxed to then return to vacuum with step and repeat few times finishing with dry N2, though make sure changes quite slow such as 10-20 minutes between cycles to avoid pressure buildup – all without batteries of course…

      1. Heating the device will help a little. Not necessarily a ton: the diffusion rate will be somewhere between linear (diffusion) and square root (conduction) with absolute temperature, so heating the device up to like, 50 C is only going to be around a 5-10% effect. However the seal might have more of a temperature coefficient so it’s hard to say.

        A nitrogen purge won’t do anything. You’re not talking about liquids here: it’s not like the nitrogen molecules shove the helium atoms around or anything. The nitrogen probably can’t even get *through* the seal. Vacuum won’t help much either – the partial pressure differential is already very low, so if you’ve got 10x normal helium concentration inside to begin with, lowering the surrounding pressure from ~4 mTorr (pressure differential = 40 mTorr internal – 4 mTorr external = 36 mTorr) to ~1 mTorr (pressure differential = 39 mTorr) isn’t going to do much.

        1. Fascinating Pat :-)
          Hope you can appreciate simple partial pressures don’t obviate the deal gas law in particular over the suggested temp cycle in this case whilst under vacuum influence (obviously due to equilibria issues progressing) in that I didn’t state a time factor for the setup cycles as I suggested and over the temperature range which is safe practical constraint also as starting point overall
          ie. Any temp differential helpful as gasses do respond to that with an He moderately porous seal which is far less permeable to N2 and H2O. Statistical mechanics offers that in relation to gas laws N2 will actually does move He around (equilibria) ie. mix so effective in purge aspects but, in any case not at all relevant here inside the MEMS seal so not sure what you imagine such collisions could achieve inside the offending device – I feel none of seal ‘good’, maybe that’s what you meant sure then I agree but, that’s not the point as I focused on outside.
          Also please note I did state Dry N2 for particular reason and I add for you in relation to seal permeability effects since H2O vapour not to be dismissed as part of designing an Experimental Methodology for gas equilibria investigations is borderline nonideal.
          Suffice to say I stand by my model for fixing the device as it stands as a dead simple starting point, up to others to augment as they see fit and cheapest possible other than waiting and hoping. It’s only a matter of setting up the experimental purge assembly properly and that means the requisite feedbacks from the device Eg. by powering the phone up between cycles (non battery) whilst at above STP for classic feedbacks to conclude the cycling as it affects clock operation if at all at earliest – which I will expect mostly similar to classic decay paradigms but, in this case clock errors/drift once it is out of the inoperable state until suitable tangent reached.
          Then it’s just a matter of applying statistical math in relation to diminishing returns such that some confidence level can be reached – my guess is 5-10 cycles on average maybe 30mins ea. Such an experimental setup as I touched upon is a ‘best minimal fit’ as starting point and I see nothing better as that starting point and for a few reasons outside this specific issue/example too.

          If you can suggest a better benign less invasive methodology Pat then please post it as alternative starting point great but, with basic procedural implementation constrained by the operators budget or patience as an additive for most part then I will be most interested to analyse, thanks :D

          1. “If you can suggest a better benign less invasive methodology Pat then please post it as alternative starting point ”

            Yeah, I can. Take it out of the high helium environment. Wait a week and hope. Thermal cycling isn’t going to accelerate the outgassing by anywhere near as much as it will decrease the lifespan of everything in the phone. There’s a reason why Apple suggested that.

            It’s also worth noting that because it’s a complete system, rather than a single part, the device might not even work at that point – if the oscillator was out of spec for a while, it’s always possible that the system did something Not Good to effectively brick the phone. With an RTC I’d say that’s unlikely, but who knows what it does in the background.

        2. Can you explain your vacuum figures in a bit more detail? I’m struggling to make sense of your figures.

          For example:

          I can’t see why you are stating that the surrounding external pressure would start at ~5mbar (4 mTorr) would it not be atmosphere which is 1000mbar? What am I missing here?

          Also, why stop at 1 mTorr when it’s possible to get chambers well below that? The chambers we build can get down to 1e-5 mbar or even 1e-6 mbar when clean & dry. Was that an arbitrary value or is there a common bit of vacuum equipment I’m not aware of that can’t get below 1 mTorr which is why you picked it? Or is it because you could start damaging other components below that pressure?

          Putting all that to the side for the moment, I think you would still see some good results from trying to purge the He in a vacuum chamber, I say that because we check our systems for leaks using He and He leak detectors and it doesn’t seem to take long for He to defuse between a vacuum differential of a few decades.

          1. “I can’t see why you are stating that the surrounding external pressure would start at ~5mbar (4 mTorr) would it not be atmosphere which is 1000mbar? What am I missing here?”

            It’s the partial pressure of helium that matters. This is Dalton’s law: the components of the gas don’t interact with each other. And while of course it’s true that this isn’t perfect, especially at high pressures, in this case it’ll be pretty darn close.

            I love explaining this because it’s fairly counterintuitive at first, until you realize what matters are the individual components crossing the barrier, not the total pressure.

            Imagine there’s a seal that only helium can get through, and inside is a perfect vacuum, and outside is normal atmosphere. Yes, air is 1000 mbar, but helium only makes up about 5 ppm. So if you think about it in terms of “particles hitting the seal”, only 5 out of every 1 million gas particles that hit the seal can make it through. And of course, the pressure is just directly related to how often the particles hit the wall – and the fraction of time the helium molecules hit the wall (and thus contribute to the pressure) is their partial pressure: (5 ppm*1000 mbar). Now, those particles will make it through occasionally, and the pressure of helium inside the seal rises.

            As that pressure rises, the helium atoms inside will also hit the seal and come back out. And you’ll hit equilibrium when the rate of particles coming in equals the rate of particles coming out. Which means the inside pressure will equilibrate at 5 millionths of the outside pressure… or the partial pressure of helium.

            “Also, why stop at 1 mTorr when it’s possible to get chambers well below that?”

            Gas diffusion rates are (basically) proportional to the difference in partial pressures (concentrations) – this is Fick’s first law of diffusion. So the problem here is that if you’ve got 40 mTorr of interior pressure, lowering the outside pressure to a billionth of a Torr isn’t going to improve the initial diffusion rate much compared to if the exterior pressure is 1 mTorr or 5 mTorr. It’s still just (40-(something small)).

            So while more vacuum of course will help, it just won’t help a lot for the diffusion rate unless you actually need to get down close to the outside partial pressure (or below it).

          2. “I say that because we check our systems for leaks using He and He leak detectors and it doesn’t seem to take long for He to defuse between a vacuum differential of a few decades.”

            Oops, forgot to respond to this – you’re absolutely right that of course a really good vacuum chamber can pull out helium quickly if there’s a leak, but the point here is that there wasn’t really a leak. It’s a good seal, it was just exposed to a high partial pressure differential for a while. Pulling a vacuum on it won’t really speed things up that much.

            I will say after reading the actual Reddit thread that the chips might be very helium-sensitive, so it’s possible that only a slight overconcentration caused the problem, in which case a vacuum definitely would help. But if it’s a larger exposure then the vacuum wouldn’t help speed things up all that much.

          3. Thanks for explaining that for me Pat, very interesting, and useful knowledge for my line of work.

            In hindsight I don’t think it would be a very good idea to expose a smartphone to a high vacuum anyway, there is some question over whether an LCD can survive in high vacuum and with most phones these days its not possible to

            In conclusion, do you think we could say that a vacuum purge will help a bit but it has limitations and could cause further damage to your phone?

    3. Would the battery in the phone be safe in a vacuum? I’d be worried about it expanding/exploding when put in a situation like that. I suppose you could remove the battery, but that would be a hassle.

    4. Diffusion should be a question of partial pressure. If this is true, than the vacuum would not change anything, as the partial pressure of helium in normal atmosphere is extremely low.

      1. Ah but, for what gasses – most appropriate even mildly explosive solvents and anaesthetics eg N2O etc far more likely most often, very unlikely Helium due to very rare release and if so rises quickly so not a breathing or explosives risk unlike H2 which can form explosive boundary layers which persist on triboelectric like materials, in any case helium rather difficult to detect economically in a personally attached sensor with its appropriate power supply, calibration and servicing overheads not trivial…

        1. Ah, the good old N2O. That was the funniest painful experience I ever had, when they pulled back in place my broken thumb. I never laughed so hard, it made breaking a thumb worth it.

          1. Helium is the gas with the highest heat conductivity. In school we had a gas chromatograph, that used helium to operate, because it used a hot-wire detector (for heat conductivity)

      2. fun fact, only apple devices starting with the iPhone 6 were affected by this.
        effectively becoming a sensor to detect a possible risk to the owners health.
        also with the iphone 6, apple also introduced “apple health” back in september 2014.
        could it all be a feature and not a bug ?

    1. Well maybe they fell over too easily as chipmunks some of the time since frequent use of Helium can displace some easily occulded air so your O2 sats would be down progressively. Reminds me of the oldie University open days where some techs would inhale Sulphur Hexafluoride to impress school kids with low voices vs high from helium and then get blue veins in their necks whilst appearing lightheaded quietly watching them hilarious wondering whether to intervene (their demo hubris failing risk assessment) that gas SF6 rather heavy and seeps around avioli and other porous tissues (though mostly benign short term) in lower parts of lungs displacing fair amount of O2 access. Lungs not laminar flow devices by any means when it comes to branches as they constrict chaotically thus one cannot deterministically expect that all helium or SF6 leaves fast enough to allow air back to some structures – ie Do not exert oneself such as lifting heavy items or going for a run when demonstrating yer chipmunk status to the equally clueless public :P

        1. Well ppffft dynamodan would that be satire cast with a long bow perchance from wayback ? If so, I’m still working on consuming more red whine, chilli then chocolate this evening (GMT+8,) so it’s anyone’s guess what might spout forth subsequently as Sol orbits to the dark north hemisphere – though I hasten to add my earlier paras across other threads that weren’t easily understood at the time most likely reflects on the inflexibility and feeble intent of various audience members consumed more by unhelpful personality traits safly in conjunction with misplaced ego/hubris not at all useful in a collaborative technical framework – as a start though that’s it for now – promise ;-)

          1. > though I hasten to add my earlier paras across other threads that weren’t easily understood at the time most likely reflects on the inflexibility and feeble intent of various audience members consumed more by unhelpful personality traits

            Nah, typically there is just too much fluff in your posts, ratio of information to data is too low. Sometimes there are even errors which look like whole sentences are missing. Requiring your readers to be in top mental condition just to properly parse some comment under article is too much. Engineers’ communication should be concise.

          2. -1
            Hey dude, stop making excuses for yourself and get a grip. God you’re like some old drunk dialling his ex. after a skinful. If you really need to show us how clever you are how about you start by getting some of the technical stuff right like “alveoli”

    1. If you want a government contract. Perhaps they wanted to sell to NASA, a deep sea diving company, nuclear research, analytical chemistry, or other similar field where helium atmospheres are more common.

      1. No, it’s more than that. You need the MEMS structure to be free from contaminants in the first place, and the best way to verify that is with a helium leak test – but in addition, many MEMS devices are assembled with vacuum in the first place, and the super-high performance ones (the ones with very low long-term drift) obviously would get degraded with any penetration, so you test against the worst-case.

      1. It’s probably not a cost issue. It’s a size and power issue. I’ve come pretty close to designing in a SiTime oscillator before. and they can perform as well if not better than a quartz oscillator at lower power and with less board area. At the power scale of an iPhone or Apple Watch, though, I have some trouble believing that the board space or power savings is really all that substantial for the additional price.

        At lower volumes, the SiTime MEMS oscillators are more expensive than crystals, but perhaps on Apple’s scale it’s reversed.

  1. So probably susceptible to hydrogen as well. I used to make a lot of that when I was younger, putting zinc from old batteries into an acid, then filling balloons or displacing water in test tubes submersed in water and then putting a match to them. POP!

    1. Very much doubt it as hydrogen diatomic whilst helium monatomic hence the MEMS seal less likely to allow hydrogen through as molecule larger and hydrogen bonding potentials too though if it did get through then it might introduce hydride effects so it might even speed up the clock or offer something else in terms of increasing sensitivity such as to beta particles, eg if it were D2O instead of H2O ;-)

    2. Hydrogen can be held in a latex balloon. Helium needs aluminized Mylar and still leaks out. Hydrogen comes in twin packs and is huge compared to Helium atoms, which is an alpha particle with occasional electrons somewhere from nearby to infinity. Helium is half the size of a hydrogen atom, and way smaller than H2 molecule.

      1. Hydrogen can be “held” in a latex balloon. It too leaks, and has the annoying property of diffusing through the aluminium mylar as well. Hydrogen in metals is an issue because it soaks into many alloys and causes them to turn brittle.

        1. very interesting, I think you may have mentioned here one of the reasons why Helium is so darn rare here on Earth. Not only is it lost into space once released into the air, but its hard to contain, even in the rocks, because it always wriggles itself out. It nearly always gets out. It always gets lost into space once released. It never (almost) combines chemically with other atoms to make heavier particles. Sounds like a formula for extreme scarcity to me, even for the second most common element in the universe (for now anyway)

  2. “Apparently, SiTime also is aware of this problem and says its newer devices are “impervious to all small-molecule gasses.” But they admit older parts were not immune.”

    SiTime upper management meeting:

    Sales: “We just scored the Apple contract!!!”
    *round of applause, pats on back*

    Manufacturing: “How many, and what is the delivery date?”

    Sales: “1 billion, and next month.”

    Manufacturing: “We don’t have the capacity!!!”

    Sales: “We HAVE to make the ship date!!! Its APPLE!!!”

    *Calls Engineering on the phone*
    “Hey what can we do to speed production on these clocks?”

    “The most time consuming part is laser welding the package…
    We could just spot weld the corners, but this would make it susceptible to…”

    “How many more could we make?”

    “Ten times more, but doing this would compromise the hermetic…”

    “Engineering says they can meet the deadline!”

    *champagne bottle cork pop*

  3. apple conspiracy theory: originally this mems was designed to make your phone get slower as it aged. in normal operation larger molecules would become trapped in it and the clock would run slower.

  4. Hmmm… interesting… but I’m more shocked about the story then by the fact that the device stopped working.

    “…40 Apple devices were either dead or misbehaving.”
    “over 5 hours about 120 liters of liquid helium vented into the air”

    This means that over a period of 5 hours more then 40 people were in the close surroundings of this machine. I’m no expert, but I imagine that the helium doesn’t stay around very long. The machine leaking gas is a “problem” to be expected, so ventilation should be present, simply to prevent people from being suffocated in case of a major leak. Though that is only for the close surroundings of the possible leaky area. So I expect the ventilation not to be very sophisticated if it blows the gas around a large area (otherwise how could it affect so many people). Then again, it wouldn’t be the first time people got sick by ventilation spreading germs, but that’s a different story.

    Now let’s say that Apple has a phone market share of 25%, this means that 1 in 4 people there are expected to carry an iphone. Then this phone has to be of a specific type to encounter this problem. So these assumption are all pretty shady, but it’s just for a simple calculation. So for 40 phones to fail… there should have been at least 4*40 = 160 people in the problem area at that moment. 160 people over a period of 5 hours, comes down to 1 person every 2 minutes in that specific area. Now that some heavy traffic for a location that houses this expensive and delicate equipment to be expected to be handled only by specialists. And these people need to stay there for a very long time for the helium to have effect!!! You must realize that the number of phones being reported probably weren’t all, unless they started a survey among all people that were possibly present at that specific day and location.

    So despite the very technically interesting effect of helium on the oscillator circuit, I have my doubts about the hospital/story. I’m not saying it isn’t true… but some how you would expect the helium to “go away” faster. Because in order for helium to penetrate into a the parts of a phone, you would expect relatively high levels of helium in the close surroundings of the phone for a substantial period of time. If it needs to be in a bag of helium for 3 minutes before it stops, then how long should it be exposed in the real situation and to what percentage of helium in the air. I simply don’t expect 160 people to be lingering around the gas leak all the time. All this fascinates me more then the mems oscillator being affected.
    So I guess I’m missing something, does anyone have a suggestion about what it could be?

    1. Your assumption of 1 in 4 iPhones in the medical field fails to consider the factor of pride. Having the priciest and biggest name counts. If my co-worker can afford a $1K phone and an $500 watch, I can too!

    2. So much text just because you didn’t read article. They had special vent to outside, but it looks like SMALL amounts leaked also into hvac system and were distributed in building. It was small enough quantities that no one saw anything suspicious, but it was enough to disable iphones.

        1. the leaking of helium is one thing I do not doubt
          stopping a resonator by filling it’s chamber with a gas that shouldn’t be there is one thing i do not doubt

          What I do doubt is the fact that for some reason a huge number of phones stopped working and that it seems to be the talk of the town… investigations start… 40 items had problems. Now seriously, this is the point the article doesn’t concentrate about, where there really 40 items that didn’t function? Or where there 40 claims of devices not doing what they should do. And if so were they all related to the helium incident?
          Lot’s of people aren’t technical and lot’s of people claim their device from not working properly all year round. It is impossible to proves afterwards if all devices that didn’t work properly were experiencing the same problem. But if you ask a person if they saw something strange with their phone then many will say yes, simply by confusing one problem with another.
          I don’t doubt the problem of helium in mems like circuitry, I doubt the scale it really happened on. That’s all.

    3. Yeah, I had the same thoughts after reading this: while I’m sure helium CAN stop an iPhone (video of iPhone in Ziploc baggie as proof), the idea of an entire hospital maintaining such levels that it would poison over a hundred phones throughout the building seems *really* far-fetched. Hospitals are BIG places.

      It seems like someone would have noticed this: either monitoring systems, or just the sheer amount of liquid helium being dumped into the system (and boiling off) over a 5 hour period, should have set off some bells.

      I DO doubt the story, and am leaning more towards “someone found an interesting note in the iPhone manual, and then concocted a story around it involving an MRI”.

      1. I totally agree with you, the ziploc bag video fails to demonstrate that helium affects iphones in real life at atmospheric pressure , the helium concentration is obviously way higher in the bag than could be achieved in a hospital and hospitals tend not to be at a higher pressure than the atmosphere, he did after all press down on the bag at 32 seconds and then handle the bag.
        My guess is that during a break a mortuary tech mentioned that helium kill iphones and 40 iphone users volunteered their phones, they do tend to make bad decisions new and then, then, in an attempt to get new iphones broke the mri and made up a story worthy of the fast and the furious.

  5. And now I must don my black hat and go buy a helium balloon. Oh the carnage that will ensue. Listen attentivly for the screams of fear and frustration from all Millenials I encounter…let the iPhone massacre begin!!!! All will remember the fearful name of Taylorian!!!!

  6. It’s interesting to note that near absolute vaccuum, atoms do not flow to a lower vacuum. Thwy stick to a wall for a while and thwn.escape, but in no oarticukar (reliable) direction… whereas initially, when oulling a vacuum, air flow can be said to occur.

  7. I’m skeptical on this. It seems to me that the amount of helium it would take to leak into devices and change the properties of the oscillator to the point of failure would have an obvious effect on people’s voices.

    “Hey Joe, why are you talking funny? Or for that matter, why am _I_ talking funny?”
    “Let’s ask Siri ……. hm.”

    1. Maybe. But you have to consider that the MEMS structure is very small, to the point that you have to use some sort of ballistic simulation instead of statistical mechanics (and consider actions at very low reynolds numbers?).

      How easily the helium can be pushed out of the way of the tuning fork as it moves will I think effect the frequency? Does the clock get faster as nitrogen is displaced? Or slower as helium leaks in but nitrogen can not leak out?

      1. Yes, the mems structure is small, but a) helium is way, way smaller than that, and b) N2, H20, and O2 are so much bigger, so much heavier, and so much more prevalent, I don’t see how you make the jump from statistical to ballistic modeling. Are the He somehow that way more energetic?

        And finally, don’t try to BS me with your low Reynolds numbers!

          1. Um, is your point ‘He vs N2’ in ref to my initial comment to deal with it on the dry N2 purge ?
            If so then that wasn’t my intent to imply its one gas simply displacing another as the mechanism for dealing with helium ingress. Fwiw: Although dry N2 in that environment is pretty inert I understand it’s a good way to remove H20 when following the dry N2 flood with vacuum ie My understanding is H20 vapour can hitch a ride (as it were) on N2 as in they bind temporarily (asymmetrical dipole interactions) being somewhat sticky. This is another reason (for Pat) why vacuum most helpful as it can release any adsorbed H20/N2/O2 on the seal of the MEMS device with perturbation offered by temperature ramps. By releasing any adsorbed gases incl O2 makes it easier for helium to diffuse out. Worth considering as we don’t know the precise physical chemical permeability properties of the MEMS seal normally exposed to outside air when it diffuses into the phone’s case.

            Also (for Pat) temperature cycles under vacuum and mild pressure as cycled also ameliorate adsorbance material aspects too – all these things considered offer a definitive controlled mechanism to Quantitatively deal with the issue other than any facile qualitative guess of mere hope in an uncontrolled environment – I still haven’t seen anyone offer anything better than my base mechanism for dealing with it for any sort of professional path to achieve a tangible outcome – hope ain’t it !

          2. No, I wasn’t referring to your mention of N2 at all; just to the fact that MY assumption was that most MEMS devices operate in air, which is mostly N2 and other gases of similar molecular weight. Which was a bad assumption – I should have realized that for reliable operation they would need to run in vacuum. I had made that assumption based on my own experience in (non-MEMS) wafers, which were handled and packaged in plain old air. These devices were all glass-passivated, so reactions with atmospheric gases weren’t an issue. Of course, NOW I realize you can’t passivate MEMS, any more than you can plastic-encapsulate watch movements.

  8. This is one more reason smart phones should be constructed to be hermetically sealed, a much more useful function than making them wibbly-wobbly, stupid thin, or with screens all the way to the edge so you can’t see all of it while protecting it in an otterbox (or equivalent) to protect its way-too-fragile frame. The only other feature that might be more important is long battery life.

    1. Yes, maybe, but the seals in question aren’t the seals of the phone – they’re the seals of the ICs themselves, which ARE supposedly hermetic. If a solid block of epoxy isn’t good enough, what kind of sealing would you suggest be applied to the phones?

      1. I think I thoroughly get your point. So if the phone is a hen house and helium is a fox, you may as well just give him all your chickens because he’s gonna get in all he wants anyway no matter how well you construct the fence. I never realized until this hackaday article that helium was such a juggernaut. :-) fascinating. Lots of it on Saturn btw.

  9. Best example, and missed use, of “RTFM” in years.

    From the reference made to the iPhone’s user guide, I am shocked that Hackaday didn’t use RTFM in the title as clickbait. Also super surprised that no one has called this out already. The testing to confirm was a wee bit unnecessary when the manual already specifies the known impact of He on the device. Still cool, just funny.

    1. The liquid helium is used to cool the superconducting magnet. Liquid nitrogen boils at 77.4 Kelvin, which is well above the critical temperature for most superconductors. Most superconductors need to be well below 40 K to operate. “High temperature” superconductors have limitations that are not suitable for MRI.

      1. Beg pardon, I think you really meant to refer to the Freezing point of nitrogen…

        The existing tech for high current superconducting magnets require a temperature much lower than the freezing point of nitrogen which is about 63K. ie. Had they used nitrogren they would have to find a suitable superconducting matetial which operates above 63K and possibly a bit higher due to pressurizing the gas. Its more expensive to find material formulations for higher temperature necessary to carry such high currents reliably especially if any hot spots if bordering on nitrogen’s boil point then a catastrophic thermal cascade.
        The only material that is a liquid below nitrogen’s freezing temperature of 63K is helium. In fact you have to go to 20 times atmospheric pressures to make helium freeze at the closest you can practically get to absolute zero.

        1. I actually meant what I wrote, but in any case, 63 K is still way too warm for the superconductors used in MRI machines.
          By the way, there are MRI machines made for locations where you can’t guarantee a continuous power supply, that use conventional electromagnets, but these consume considerably more power. (If you lose power to the cryogenic system in a helium-cooled superconducting magnet for a significant amount of time, you risk having it go resistive, which increases power dissipation in a positive feedback loop that results in catastrophic failure of the magnet, similar to what you say about higher-temperature superconductors.)

          1. Beg pardon, you meant to focus boiling point really (??) that doesn’t address primary issue as in being off target in terms of key design criteria. You wrote “..nitrogen boils at 77.4 Kelvin, which is well above the critical temperature for most superconductors..”. Boiling points well above any other gases transition from solid to liquid for coil operation therefore not design criteria connected.

            The boiling point of the fluid being above its freezing point (which is also above the superconducting coils operating point) is totally Irrelevant in respect of design and regular operation when staying as liquid and not freezing imperative. For the system to function you need a material that is a liquid and stays as such under easily maintained economical pressure at the temperature the (cost effective) superconducting coils operate. IOW. As case in point, if the nitrogen’s boiling point we’re lower or higher it still makes no difference to the essential fact you need a fluid that won’t freeze. ie The temperature at which the fluid freezes is obviously the primary design and operational issue and in that respect too all gases in that range have wide enough margins over fluidity range as a liquid and not solid before they boil.

            Sorry but, your remaining sentences appear to shift ground touching on magnets doesn’t help in any way to clarify your idea of boiling points. Its about the coils resistivity not the cores when it comes to helium vs nitrogen. Appreciate control system interlocks in place so the device either stays in its Safe Operating Area (SOA) or shuts down minus power with a UPS with minimal venting of gases under thermal inertia time factors.
            FWIW. Straightforward one could develop a means to shut down fully even via UPS for short period and not lose any helium but, is it cost effective for the pumping at higher pressures into larger volume above helium’s boiling point (?) presently an open question and the fact is hasn’t yet happened AFAIK is replacing helium is still comparative cheaper overall.

          2. Mike Massen: The temperatures of most cryogenically cooled systems are regulated by the boiling point of the fluid. This is because as long as the fluid is liquid, you know it’s below the boiling point. My POINT was that even if I am wrong about this, nitrogen is STILL completely inadequate for cooling magnets for MRI. So even if it were safe to operate a nitrogen cooling system near N2’s freezing point, it STILL WOULDN’T WORK.
            I feel that I had already clarified this well enough, and was moving on to other points.
            Your point about using UPS to guarantee safety is certainly valid; I was just pointing out that cryogenic magnets aren’t even necessary for MRI, as, and I repeat, there are MRI machines designed and sold for applications where continuous power cannot be guaranteed. This was in response to AVR’s post – it’s not all about you.

    2. LN2 isn’t cool enough for HIGH CURRENT superconductors and it takes high current to get the strong magnetic fields used in MRIs and supercolliders for that matter. Yea, there are superconductors that work at LN2 temps but they don’t “cut the mustard” in this context. :-)

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.