Apple's Lightning connector detailed in extensive new patent filings

Posted:
in General Discussion edited January 2014
A trio of Apple patent applications filed with the U.S. Patent and Trademark Office on Thursday reveals the intricacies of the company's latest input/output protocol for iOS devices: the Lightning connector.

Lightning
Source: USPTO


When Apple introduced Lightning alongside the iPhone 5 in September 2012, not much was known of the protocol besides its small size and orientation agnostic design, two vast departures from the venerable 30-pin dock connector it replaced.

Further investigation, including a teardown that found a built-in authentication chip, revealed that the new connector dynamically assigned contact pins to allow for reversible use. Processing was thought to be executed by a control unit onboard the iPhone, though the true inner workings of the protocol remained largely unknown.

The three patent applications discovered today yield a detailed look into how Lightning works, both from the accessory side and the iPhone or host device side, as well as how the physical connector is designed to facilitate reliable operation.

Starting with construction, Apple's patent filing for a "Dual orientation connector with external contacts and a conductive frame" is an incredibly in depth overview of Lightning's hardware design, build and proposed functionality.

Lightning


With Lightning, Apple looked to overcome deficiencies with previous connectors, including industry stalwarts like the ubiquitous 3.5mm headphone jack and USB. In designing the new protocol, the company needed something short enough to allow for the inclusion of increasingly large mobile device screens, and thin enough to accommodate consumer demand of ultra-slim form factors.

Also considered was signal interference due to the accumulation of debris in connector cavities and ease of use associated with how a plug is oriented when inserted into a device. Finally, the "feel" of a connector was accounted for, with unwanted wobble and imprecise insertion and extraction being a sticking point.

In order to retain the tab, for example, detents are employed on the metal sheath, while arms on the female side hold the connector in place. A similar mechanism was used in Apple's 30-pin connector.

Lightning


One the most important considerations was creating a multiple orientation connector. In many of the invention's embodiments, the Lightning connector's pin arrangement and overall "tab" design are discussed. While each has its own pin layout, all point to a flattened part with two large surfaces flanked by two smaller surfaces, or sidewalls.

Contacts on the connector include a ground, multiple data signal pins and ID pins to denote orientation. The problem with creating an orientation agnostic design is how to facilitate the identification of each pin on the connector. To solve this dilemma, Apple disposed an ID module that is operatively coupled to the pin contact on the device side.

In one embodiment, the invention calls for a set of rules regarding the location of data signal pins and ground contacts, with the positioning either mirrored on both sides of the tab, or disposed opposite each other. The wiring within the tab, which is consequently associated with the correct pinout, connects contacts on one side of the tab with its partner contact on the opposite side. However, it should be pointed out that two contacts, accessory ID and accessory power, are located kitty-corner to each other on opposing sides of the tab. This configuration allows for the dual orientation compatibility seen in Lightning.

Lightning
Lightning pinout as per one embodiment of the patent.


The application goes on to explain the various electrical requirements needed to make the system work, including per-pin voltages and internal wiring. Also noted are manufacturing techniques to produce the new standard.

Lightning


Apple's Lightning hardware patent was first filed for in November 2012 and credits Albert J.Golko, Eric S. Jol, Mathias W. Schmidt and Jeffrey J.Terlizzi as its inventors.

The two remaining patent applications (1, 2), both titled "Techniques for configuring contacts of a connector," deal with the dynamic assignment of pins on the connector by a host side identification system.

Basically, the properties describe the host and accessory systems needed to dynamically assign pins a certain role, whether it be as a data signal contact or a power contact.

Lightning
Various pin designs for Lightning.


According to the patent language, in one embodiment the contacts are configured by first detecting the mating of two conncectors. In response, the host unit sends a command through two host-side contacts and, depending on which returns a valid response is received, determines the orientation of the connector.

In operation, a switch for a first orientation detection contact is opened and a command signal is sent from the host device. If a viable signal is returned, which can be translated by the ID module, the orientation process stops and orientation is set. If, however, no signal is received, as would be the case if the connection is with a power line, the OD1 switch is closed and OD2 repeats the process.

As noted above, the accessory power and accessory communication contacts are located kitty-corner to one another, thus if an OD contact finds the accessory comm pin, the correct orientation of the connector can be deduced.

From this information, the remaining connectors can be dynamically assigned with correct power and data signal pathing.

Lightning


After describing the hardware requirements, the filing goes into detail describing how the aforementioned ID module can be programmed with identification and configuration information to successfully communicate with the connector. An authentication module is used to serve up an encryption key routine with the host device, while a current regulator embedded in one of the ID chips can function as a voltage regulator.

The patent applications were filed for in November and December 2012, and credit Jeffrey Terlizzi, Scott Mullins, Alexei Kosut, and Jahan Minoo as their inventors.
«13

Comments

  • Reply 1 of 59
    wyredwyred Posts: 10member
    Next feat for Apple would be to come up with a new headphone jack standard and force it onto the market. I would support that completely...
  • Reply 2 of 59
    asciiascii Posts: 5,936member
    It's a real contrast. Lighting is almost perfect from a physical point of view: small, reversible and a firm fit, but has a relatively low data rate. Thunderbolt on the other hand is wobbly even when plugged in, and has big chunky connectors in order to move certain circuitry external, but it has a screaming data rate, the fastest of all ports. Take your pick...
  • Reply 3 of 59
    rcoleman1rcoleman1 Posts: 153member
    ascii wrote: »
    It's a real contrast. Lighting is almost perfect from a physical point of view: small, reversible and a firm fit, but has a relatively low data rate. Thunderbolt on the other hand is wobbly even when plugged in, and has big chunky connectors in order to move certain circuitry external, but it has a screaming data rate, the fastest of all ports. Take your pick...
  • Reply 5 of 59
    rcoleman1rcoleman1 Posts: 153member
    I totally agree. Put that kinda speed in the Lightning connector and you'd have a new industry standard for sure. Apple PLEASE READ THIS!!!!
  • Reply 6 of 59
    anantksundaramanantksundaram Posts: 20,411member
    Trust Apple to take something as simple and pedestrian as a 'connector' to the next level.

    Five years from now, most of the rest of the world will claim that it was all, of course, obvious.

    The copying machines are running overtime everywhere, I am guessing.
  • Reply 7 of 59
    solipsismxsolipsismx Posts: 19,566member
    It's been a long time since I've seen a patent that had excellent drawings.

    ascii wrote: »
    It's a real contrast. Lighting is almost perfect from a physical point of view: small, reversible and a firm fit, but has a relatively low data rate. Thunderbolt on the other hand is wobbly even when plugged in, and has big chunky connectors in order to move certain circuitry external, but it has a screaming data rate, the fastest of all ports. Take your pick...

    I don't understand why people assume Lightning is only capable of relatively low data rates? Are you judging the NAND bottleneck? Isn't that like claiming Usain Bolt is slow if he's queued up for a hot dog in Times Square?
  • Reply 8 of 59
    chandra69chandra69 Posts: 638member

    Quote:

    Originally Posted by SolipsismX View Post



    It's been a long time since I've seen a patent that had excellent drawings.

    I don't understand why people assume Lightning is only capable of relatively low data rates? Are you judging the NAND bottleneck? Isn't that like claiming Usain Bolt is slow if he's queued up for a hot dog in Times Square?


    I felt the same. May be because of too much minute details, they have to draw it neatly.

  • Reply 9 of 59
    asciiascii Posts: 5,936member

    Quote:

    Originally Posted by SolipsismX View Post



    I don't understand why people assume Lightning is only capable of relatively low data rates? Are you judging the NAND bottleneck? Isn't that like claiming Usain Bolt is slow if he's queued up for a hot dog in Times Square?


    I said it was *relatively* slow in a post comparing it with Thunderbolt, certainly didn't expect any disagreement on that one.

  • Reply 10 of 59
    solipsismxsolipsismx Posts: 19,566member
    ascii wrote: »
    I said it was *relatively* slow in a post comparing it with Thunderbolt, certainly didn't expect any disagreement on that one.

    I wasn't sure exactly what you were getting at hence my questions. Well, the last one does make a presumption but I had hoped you'd see that as an attempt at humour.
  • Reply 11 of 59
    gustavgustav Posts: 828member
    "Further investigation, including a teardown that found a built-in authentication chip, revealed that the new connector dynamically assigned contact pins to allow for reversible use. Processing was thought to be executed by a control unit onboard the iPhone, though the true inner workings of the protocol remained largely unknown. "

    From the start, the theory that is was solely an authentication chip was challenged. Most people knew it was to offload the signal processing to the cable, resulting in more capabilities of the connector itself. In theory, Apple could now issue a USB 3 cable, though I think the bottleneck is still the speed of flash memory in the devices.
  • Reply 12 of 59
    mrallistermrallister Posts: 27member
    wyred wrote: »
    Next feat for Apple would be to come up with a new headphone jack standard and force it onto the market. I would support that completely...
  • Reply 13 of 59
    iaeeniaeen Posts: 588member
    Does anyone know why they didn't just mirror the pins on both sides? Dynamically assigning pins is neat and all, but it seems to me to be an overly complicated solution to a simple problem.
  • Reply 14 of 59
    solipsismxsolipsismx Posts: 19,566member
    mrallister wrote: »
    Apple already did, when Apple released I forget what generation iPod, it introduced it with the 3.5 millimeter headphone jack, setting the standard still used today

    Was the 3.5mm not used before it was in an iPod?

    iaeen wrote: »
    Does anyone know why they didn't just mirror the pins on both sides? Dynamically assigning pins is neat and all, but it seems to me to be an overly complicated solution to a simple problem.

    Mirroring would appear to be more complicated to build since that would mean a lot more overlapping pins in such a small space. What I'm seeing in the design that each side has the exact same order of pins save for the ground pin which will dictate which side is pugged in. Since the ground accounts for one of the pins the other power pins are shifted over by one in the same power pin slots for one side. Doesn't this allow all the complexity to be in the chip itself?

    The only way I could see it be less complex is if they do a direct mirror on the male plug, as you state, but wouldn't that mean the device would need a 9h pin so that it could use that dedicated ground on each side to know which orientation it was plugged in?
  • Reply 15 of 59
    tallest skiltallest skil Posts: 43,388member


    Originally Posted by MrAllister View Post

    Apple already did, when Apple released I forget what generation iPod, it introduced it with the 3.5 millimeter headphone jack, setting the standard still used today


     


    How in the world are you confusing that with anything?


     


    I can't find anything on the release date of 3.5mm (which is ludicrous to me; that should be easy to have somewhere), but I have a walkman from the early '80s that has 3.5mm.

  • Reply 16 of 59
    normmnormm Posts: 653member
    iaeen wrote: »
    Does anyone know why they didn't just mirror the pins on both sides? Dynamically assigning pins is neat and all, but it seems to me to be an overly complicated solution to a simple problem.

    By dynamically assigning pins, they have eight distinct configurable "data" pins. If they required hardwired mirrored pins, there would only be four. Maybe they're contemplating using these eight pins, for example, as four differential pairs (i.e., four pairs of complementary signals) in order to implement the high speed Thunderbolt interface.
  • Reply 17 of 59
    solipsismxsolipsismx Posts: 19,566member
    normm wrote: »
    By dynamically assigning pins, they have eight distinct configurable "data" pins. If they required hardwired mirrored pins, there would only be four. Maybe they're contemplating using these eight pins, for example, as four differential pairs (i.e., four pairs of complementary signals) in order to implement the high speed Thunderbolt interface.

    I guess there is a potential for them to make the receptacle have connectors on both sides thus allowing for 8 data pins but is that really necessary for higher speeds once the NAND can handle it? You can get at least 1Gb on 4 pins with Ethernet.
  • Reply 18 of 59
    iaeeniaeen Posts: 588member
    solipsismx wrote: »
    Was the 3.5mm not used before it was in an iPod?
    Mirroring would appear to be more complicated to build since that would mean a lot more overlapping pins in such a small space. What I'm seeing in the design that each side has the exact same order of pins save for the ground pin which will dictate which side is pugged in. Since the ground accounts for one of the pins the other power pins are shifted over by one in the same power pin slots for one side. Doesn't this allow all the complexity to be in the chip itself?

    The only way I could see it be less complex is if they do a direct mirror on the male plug, as you state, but wouldn't that mean the device would need a 9h pin so that it could use that dedicated ground on each side to know which orientation it was plugged in?

    The way I am reading the diagram is that the ground is wired so that it is the same regardless of orientation (being on the far right regardless of which direction is considered up). My question is why not make all the pins like this. If they did, I don't see any reason why they would need to determine orientation because either direction is wired the same. This might be mechanically more difficult due to all the wire crossings, but at the same time you wouldn't need it to be a smart cable or any of the mechanisms to determine orientation.

    I'm not an electrical engineer though, so maybe it really is easier to do all this with chips than what I'm thinking it is.
    normm wrote: »
    By dynamically assigning pins, they have eight distinct configurable "data" pins. If they required hardwired mirrored pins, there would only be four. Maybe they're contemplating using these eight pins, for example, as four differential pairs (i.e., four pairs of complementary signals) in order to implement the high speed Thunderbolt interface.

    This is what I was thinking before we knew anything about the cable. I remember when everyone was saying that it was an 8 (or 9) pin connecter I was thinking "no, there are 16 (or 17) pins, and if they are going to go through all the trouble of making a smart cable that dynamically assigns pins surely they won't leave half of them unused." But they didn't do that. The current lightning ports on the iPhone/iPad have only 8 contacts, and the schematic on the patent clearly shows only 4 data pins. This is really why this design confuses me so much.
  • Reply 18 of 59
    drblankdrblank Posts: 3,385member

    Quote:

    Originally Posted by ascii View Post



    It's a real contrast. Lighting is almost perfect from a physical point of view: small, reversible and a firm fit, but has a relatively low data rate. Thunderbolt on the other hand is wobbly even when plugged in, and has big chunky connectors in order to move certain circuitry external, but it has a screaming data rate, the fastest of all ports. Take your pick...


    Both have different applications, one is for charging the battery and small data transfers, the other is a high end I/O for large data transfers in both directions at the same time.  Hence why they put Thunderbolt on desktops/laptops and Lightning on mobile devices.


     


    The two main drawbacks I've seen on the Lightning is that the connector, under the right (wrong, depending on how you look at it) can be broken. I've seen them broken on displays at the local Apple store because a kid grabs the phone and doesn't lift up and they end up putting too much pressure on the connector and it ends up snapping.  The other drawback is that it isn't a standard whereas Micro or Mini USB is.  That's a dilemma that Apple has had to deal with.  I don't know if I would have chosen for the Lightning route if it were up to me.  it's just a problem when trying to deal with these standards.  


     


    Thunderbolt, since that's an Intel I/O like USB was, it's a lot easier for them to get it to be a standard.  All it takes is Apple and other PC mfg to embrace it which they have and it's just one of many i/O ports that are being used.

  • Reply 20 of 59
    solipsismxsolipsismx Posts: 19,566member
    iaeen wrote: »
    The way I am reading the diagram is that the ground is wired so that it is the same regardless of orientation (being on the far right regardless of which direction is considered up). My question is why not make all the pins like this. If they did, I don't see any reason why they would need to determine orientation because either direction is wired the same. This might be mechanically more difficult due to all the wire crossings, but at the same time you wouldn't need it to be a smart cable or any of the mechanisms to determine orientation.

    Let's say you decide to make this cable have a more complex wiring scheme that isn't handled by an internet chip. How much does this add to the cost of the device? How much does this potentially affect the failure rate in manufacturing? How does this affect the size of the cable since you'll need to have more shielding for the extra wires that cross each other?

    Then you still need a chip in the cable so Apple can do what it wants to do with this future-froward design. Their solution seems like the only feasible solution and frankly it's long overdue.
Sign In or Register to comment.