7 ms·
This is a feature of newer usb-c pd chipsets. See, e.g., https://www.ti.com/product/TPS25751 https://www.ti.com/product/TPS25751 in the features section. Appl
by DannyBee 3y ago
This is a feature of newer usb-c pd chipsets.
See, e.g., https://www.ti.com/product/TPS25751 https://www.ti.com/product/TPS25751 in the features section.
Apple has historically used TI chipsets for their USB power controllers. The feature likely comes from this.
- userbinator 3y agoHow it works (on page 52): "The TPS25751 supports liquid detection by measuring the voltage level across the SBU pins of the Type-C connector. When a short occurs across the SBU pins to another Type-C pin(s), the TPS25751 takes action to protect the the Type-C port by disabling the power paths and notifies the embedded controller when applicable. Once liquid has been detected, the TPS25751 continuously monitors the SBU voltage. During the SBU voltage monitoring, if the liquid/short is no longer present the TPS25751 takes action to return to normal operation." In other words, it's not actually detecting "liquids" but more of any conductive short. Non-conductive liquids, like oils and many other petroleum products, won't be detected.
- jeanlucas 3y agoDumb question: wouldn't non-conductive liquids be harmless? i.e., the USB port can in theory work even if it is immersed in some non-conductive liquid, no?
- ianferrel 3y agoNon-conductive liquids still might be corrosive, or might provide enough thermal insulation to cause heat build-up, or might leave some kind of residue that would be problematic, or...
- 13of40 3y agoOr dissolve plastics and rubber, maybe delaminate circuit boards, destroy any tape and glue...
- disillusioned 3y agoSure, this is the same principle of, eg, the mineral oil cooled PC: https://www.pugetsystems.com/mineral-oil-pc/ https://www.pugetsystems.com/mineral-oil-pc/
- c0n5pir4cy 3y agoAlso some companies like 3M actually make a collection of liquids specifically for immersion cooling in data centres, so it's mature enough that people are happy to use it even in a data centre setting.
- monocasa 3y agoVery mature, going back to at least 1985 in the Cray-2.
- CamperBob2 3y agoI wonder if any of those coolant varieties survived the recent jihad against fluorochemicals. AFAIK 3M is out of the business entirely at this point.
- redox99 3y ago> Dumb question: wouldn't non-conductive liquids be harmless? Mostly yes, except for things like concentrated sulfuric acid which is highly corrosive but a bad conductor.
- robocat 3y agoAgree: bad conductor - https://chemistry.stackexchange.com/questions/153107/why-does-the-graph-of-the-electrical-conductivity-of-sulfuric-acid-water-solutio https://chemistry.stackexchange.com/questions/153107/why-doe...
- geon 3y agoI'll keep that in mind.
- flir 3y agoSomewhere, a YouTube content creator scribbles Macbook Pro in concentrated sulphuric acid. (And I'd probably watch it).
- Crosseye_Jack 3y agoAs long as we get a Steve Mould cross-section made from 3D printed parts, I’m on board!
- adolph 3y agoNileRed for HN, paging NileRed for a video about non conductive liquids in USBC PD
- Crosseye_Jack 3y agoNileRed: I turned this MacBook Pro into Cotton Candy!
- erik_seaberg 3y agoAt least bigstackD can recover the aluminum afterward.
- PeterStuer 3y agoBut you would need to make sure that the pins still contact though.
- pvaldes 3y agoWould depend on the nature of the liquids. A harmless liquid can still damage the plastic or be toxic for the owner.
- gardaani 3y agoDoes the USB-C specification allow that? Can such devices get USB-C certification? OTOH, I haven't seen any USB certifications on Apple USB cables or devices, so most likely Apple doesn't care.
- tiagod 3y agoAlthough Apple is a board member of USB-IF, at least their cables are indeed not USB-IF certified.
- yjftsjthsd-h 3y agoWhy would that not be allowed? In isolation, not running power through a short seems like a great safety feature for the hardware.
- rickdeckard 3y agoThis is a standard feature of all Smartphones with IPxx certification. It prevents corrosion on the connector contacts.
- explodingwaffle 3y agoYes: because Apple actually got this added to the spec: https://www.usb.org/document-library/usb-type-cr-cable-and-connector-specification-release-23 https://www.usb.org/document-library/usb-type-cr-cable-and-c... which is the real story IMO (I am impressed no one bothered looking at the Type-C spec update after the Type-C iPhone finally came out) This actually replaced the analog audio mode (good riddance- whoever thought adding analog audio into a port that also does digital audio was a good idea?)
- tgv 3y agoMe with an analog headphone?
- crote 3y agoThe problem is that it's a rather confusing Alt Mode. Basically, it's only supposed to be used for USB-C-to-3.5mm adapters, and it was only implemented by some smartphones. It's not a reliable way to get audio out of a USB-C port, and because it's using the internal DAC the audio quality isn't going to be good anyways. The proper solution is to use a USB DAC. This provides the same user experience but it actually works with all USB-C devices. Considering the official Apple USB DAC is available for $9, the Audio Alt Mode really doesn't have a reason to exist.
- ajross 3y agoSo it's a software visible GFCI circuit. Not super interesting, though sure, if the chip does it the OS should probably listen. The spin in the article about how this is all because of warranty denial revenue is pretty weird though.
- nucleardog 3y agoReally nitpicky, but not really a GFCI. The USB connector is detecting a short circuit--a direct connection between the two pins--and cutting power while the condition is present. If you wanted to put this in terms of household electrical, this would be closest to a self-resetting breaker. (And actually, most devices you have already _have_ this. There's usually a self-resetting fuse/PTC thermistor protecting your USB ports from over-current conditions. The only novel thing here seems to be that it's done in software.) A GFCI, meanwhile, works by detecting in imbalance between the hot and neutral lines. The amount of current flowing through one should match the amount flowing through the other. If power is being drawn through the hot pin but not returning through the neutral, it has to be going _somewhere_ and it's not where we intended it to go. It cuts the power off rather than let it continue to flow out and find a path to ground through your kitchen sink/etc.
- ajross 3y agoEven more nitpicky: that's totally a GFCI. You "detect a short circuit", by definition, by detecting current flowing between pins that shouldn't carry current. GFCI is the same thing, just for the special case of three-wire power delivery where all it needs to do is test one ground line.
- burnerthrow008 3y agoGelman amnesia and all that… Even more nitpicky: it’s not a GFCI because GFCI does not detect current leakage between pins that shouldn’t carry current. It measures current imbalance between pins that should be carrying current. The distinction is that a real GFCI (which this USB port is not) does not care whether the imbalance is returning through a third conductor on the power cord or some other path. If this were a real GFCI, it would be able to detect when current flows out through the case, to your hand, though your body and then back though the puddle of watch you’re standing in. That’s not what this is doing. It is measuring unexpected current between the pins within the port for the purpose of avoiding galvanic corrosion (the hint is in the daemon name reported by macrumors).
- David_FF 3y agoThanks for posting this. I was curious how they were able to detect liquid in the USB port
- krackers 3y agoThe binary is at "/System/Library/CoreServices/liquiddetectiond". I took a quick look at disassembly, it has some functions for getting voltage, an amusingly named selector called "setIsWet". There's an "AnalyticsSendEvent" in there (and some corresponding obj-c methods). Maybe marcan will do a deeper disassembly one day. There's some reference to TI stuff in the selector names so you're probably right about that.
- londons_explore 3y agoKinda surprising they need an entire process for this... Smells of bad engineering - do they also have a batteryislowd? or a someonepressedtheMkeyonthekeyboardd?
- sneak 3y agoThey do indeed have a lot (dozens) of small, single purpose daemon processes in modern macOS. Many of them access the network, too.
- macNchz 3y agoOver the years those processes have also frequently–at least for me–been the answer to “where did all my battery life go and why is my laptop so hot?”, as they’d rarely but inexplicably go haywire and peg the cpu at 100% until being killed manually.
- spease 3y agoDepends on what you're optimizing for. Good for security and stability, bad for performance. > someonepressedtheMkeyonthekeyboardd "Pages would like to access the 'M' key" "Open System Preferences" "Deny"
- rickdeckard 3y agoFWIW, liquid (pin-shortage) detection on the USB-port is a standard feature of all Smartphones with IPxx water-ingress certification. It prevents the device from applying/draining power from any pin in such a state, mainly to reduce corrosion of the contacts and increase longevity. (Without such a mechanism, an IPxx device may be unable to charge within a few weeks as the contacts of the connector can deteriorate quite fast) As to why Apple ported the daemon from iOS to macOS....well, for sure a good indicator to reduce warranty repair-costs...
- DannyBee 3y agoYes, it is, and most of them get it from the same chipsets :) They are not making their own, it's a function of the controller of the USB-C PD port they use.
- dharma1 3y agoOn iOS you can thankfully override it (the alert dialog lets you do it) in case of an emergency - which is pretty handy if you’ve had the phone in the pool/shower and really need to charge it and don’t have a blow dryer to hand. Less frequent of an issue for laptops I would hope :D
- rickdeckard 3y agoNow that override is something I would log as a vendor, because THAT's when the corrosion happens on the connector...