All About USB-C: Power Delivery

USB-C eliminates proprietary barrel plug chargers that we’ve been using for laptops and myriads of other devices. It fights proprietary phone charger standards by explicitly making them non-compliant, bullying companies into making their devices work with widely available chargers. As a hobbyist, you no longer need to push 3 A through tiny MicroUSB connectors and underspecced cables to power a current-hungry Pi 4. Today, all you need is a USB-C socket with two resistors – or a somewhat special chip in case the resistors don’t quite get you where you want to be.

You get way more bang for your buck with USB-C. This applies to power too; after all, not all devices will subsist on 15 W – some will want more. If 15 W isn’t enough for your device, let’s see how we can get you beyond.

Reaching Higher

USB-C power supplies always support 5 V and some are limited to that, but support for higher voltages is where it’s at. The usual voltage steps of USB-C are 5 V, 9 V, 15 V and 20 V ; 12V support is optional and is more of a convention. These steps are referred to as SPR, and EPR adds 28 V, 36 V and 48 V steps into the mix – for up to 240 W; necessitating new cables, but being fully backwards and forwards compatible, and fully safe to use due to cable and device checks that USB-C lets you perform.

A charger has to support all steps below its highest step, which means that 20 V-capable chargers also have to support 5 V, 9 V, and 15 V as well – in practice, most of them indeed do, and only some might skip a step or two. You can also get voltages in-between, down to 3.3 V, even, using a PD standard called PPS (or the AVS standard for EPR-range chargers) – it’s not a requirement, but you’ll find that quite a few USB-C PSUs will oblige, and PPS support is usually written on the label.

You can’t get beyond 5 V with just resistors though – you’ll need digital communications over the CC line, using a protocol called USB PD (Power Delivery) – which lets a device and a PSU negotiate power requirements in a featureful way. It’s a constant-baudrate bidirectional protocol, there’s CRC checks, response timing requirements, and it’s used for basically everything USB-C – even high-speed protocol negotiations. Most importantly, USB-C PD is immensely powerful.

There’s a mindboggling amount of communication possible through USB-C, letting us build devices smarter than ever before. It’s thoughtfully designed – the communications are forwards- and backwards-compatible, with new 140 W EPR chargers happily charging old 60 W devices at 60 W rates, and 60 W chargers still usable for slower charging of 140 W-craving devices. USB PD capabilities are way beyond “give me this voltage limited by this current”, too – devices can query each other’s power role preferences, charging states, swap power roles on the fly, determine what a cable is capable of, and do all of that with safety in mind.

Docks For Everyone, Let Nobody Leave Dissatisfied

If you don’t yet appreciate the complexity, let me walk you through a complex scenario that is made easy by USB-C. Imagine a USB-C dock with a USB3 port, a HDMI port, and a USB-C charger input port. If you connect such a dock to your laptop without plugging a charger into it, the laptop will feed this dock with 5 V, capped at either 1.5 A or 3 A, depending on the laptop. That alone is quite useful when you need HDMI or an extra USB3 port on the go, and some such docks have Ethernet too – hard to pass up on that. All in all, the dock acts as a power sink, and the laptop acts as a power source. The arrangement is turned onto its head, however, when you plug a high-voltage-capable charger into this dock’s charger input port.

The IC inside the dock will detect the charger and act as an intermediary, talking PD with the laptop and the charger, determining each other’s power requirements and capabilities. Say, the laptop’s interested in the 20 V, 3 A that the charger can provide. The intermediary IC will tell the laptop to stop supplying 5 V and prepare to charge from a higher voltage instead, and once the laptop agrees, will tell the charger to ramp up the voltage. Within a second, the system’s changed completely – instead of running on 5 V from the laptop, the dock now passes 20 V from the charger to the laptop through itself, and gets power for its own needs from the same 20 V.

This power role swap is designed to be completed within a small timeframe – your USB3 and HDMI devices won’t experience brownouts, and high-speed communications aren’t interrupted as power role swap happens. 20 V doesn’t get onto the 5 V-only pins, either – the dock has power regulation and gating inside to prevent that. The dock will also work if your laptop or phone doesn’t support video output over USB-C, in which case, everything but the HDMI port will function as intended, and if the laptop doesn’t support charging over USB-C, then charging negotiations will fail in a safe way.

The dock IC knows the power budget for its own USB3 and HDMI ports, and subtracts it from the charger’s power budget before presenting it to the laptop, with the laptop knowing exactly how much current it’s able to consume, and being able to stick to that for powering everything, keeping the charger happy and not overloading it. If the gear used is up to USB-C standards, the user doesn’t have to take any precautions – everything is safe.

Best of all, you can buy a dock like this for just $10 online. All of these features aren’t a fluke or a hack – USB-C is designed to do all of this and much more. As a user, you no longer need to buy a business class laptop to get a fully featured single-cable dock – many consumer-grade laptops and phones with USB-C data and charging chops can handle all things listed above without breaking a sweat.

How Do They Talk?

If you like geeking out about consumer and hacker tech advances, you will find USB-C firmly within the futuristic-technology territory as you get to know it. You deserve to know how all of this works, too – let’s go through the basics of power communications with PD.

Say, you plug a charger into a laptop. The charger provides a pull up on the CC line, the laptop will detect that, and if your laptop is USB-C charging-capable, it will put a pull down onto the CC line. The charger will then provide 5 V, and will get ready to talk PD for anything higher than that. As you can see, before higher voltages are possible, you first have to go the 5 V and analog communications way. Put your 5.1 kΩ resistors onto CC pins, get 5 V, and talk the power supply into giving you a higher voltage from there.

The resistors remain connected during communications – in fact, the combination of the pullup and pulldown resistor has to be present constantly for the charger to keep supplying current, even at voltages way beyond 5 V. The CC line voltage is used so that the PSU can quickly determine when a device has been unplugged, and vice-versa – this helps with safety features like reducing contact arcing at higher voltages when unplugging the cable, and making sure the charger goes out of its higher-voltage modes so that it doesn’t destroy the next device connected – which used to be a failure mode for some very early USB-C chargers.

The idle voltage on the bus is the resistor divider-created voltage

As a result, the digital signalling is overlaid onto the CC line voltage. You could say that PD a half-digital protocol. This requirement of PD, plus demands like applying VCONN to currently unused pin for checking high-speed or > 3 A cables, make it hard to just wire the CC pins up to a UART peripheral or the like. However, there are microcontrollers with PD peripherals, both of Western and Eastern design, and there’s PD-talking chips you can hook up over I2C if your microcontroller lacks one.

The ability to talk PD is instrumental for making USB-C hacker-accessible. You don’t have to know the PD lingo if high voltage out of a USB-C port is all you need, however.

Just Give Me Twenty Volts

You can buy a friendly IC to speak PD for you. You might’ve heard of “PD trigger boards” – a small board with a USB-C socket, two pins for voltage output, some solder jumpers if you’re lucky, and an IC with just enough PD vocabulary to convince a power supply it should give us 20 V or 9 V. We’ve seen hackers use PD trigger boards and chips for converting old laptops and higher-powered devices to get power from USB-C power supplies, in situations where the manufacturer most absolutely didn’t intend for that to happen.

A custom breakout for CH224K, assembled

There’s a number of trigger ICs around. We’ve seen hackers successfully use IP2721, and recently, exploring ICs like the CH224K. Most of us just buy a small PCB with everything soldered on, however, and don’t fret over the specific ICs; that said, you’ll want to have a few suggestions whenever you set out to design a device with a USB-C port high voltage demands, what’s with surprise out-of-stock events and all.

PD trigger boards are not perfect for everything. Say, you want to power a resistive load of 8 Ω through USB-C – let’s use a PD-capable soldering iron as an example. The tip needs 2.5 A at 20 V, and 1.9 A at 15 V. A 60 W charger can provide 3 A at 20 V and will happily power your load, so you can safely set your trigger board to 20 V with a solder jumper. But then, a 45 W charger can only offer 2.25 A on 20 V, however, and a 8 Ohm load will put it into overcurrent protection mode – you need to use 15 V even though the 20 V mode is available. Trigger board ICs aren’t designed with logic like this in mind – thus, projects like the PD Buddy Sink exist, pairing a PD-talking chip with a MCU that can handle more complex logic.

In fact, trigger boards don’t tend to give you any current limit information – if there’s risk of overcurrent because the charger cannot provide 3 A at the voltage the trigger board request, you’ll just have to make sure the charger you’re using implements overcurrent protections; those are mandated by USB-C and done by most, that said, you don’t want your device powercycling every few seconds, either. Also, if you want a dual-role port that works with OTG adapters or perhaps can do high-speed lanes, a trigger IC won’t be able to help you, and you can’t just connect multiple PD ICs requesting different functions in parallel to your CC pins. We can try and surpass these limitations in later articles – for now, know that trigger boards have their well-defined place in your USB-C arsenal, and they will help you with your USB-C projects as far as higher voltages are concerned.

I Want To Give Something Twenty Volts

What if you have a 20 V barrel jack PSU that you’d like to turn into a USB-C charger? After all, the voltages match. Things aren’t as simple – you can’t just shove 20 V into a USB-C port and expect your laptop to charge. In fact, your laptop might die if you do that – it’s not guaranteed that magic smoke won’t be released; some laptops have protections against that, but I’ve witnessed this arrangement cause magic smoke emissions, and I’d rather not have that happen to you, too. You want to do it properly, and for that, you need to go through the PD negotiation song and dance first, limiting the supplied voltage to 5 V while at it.

There are adapters online that seem to do at least the 5 V and PD parts for you – without the intermediate steps, but hey, you can omit them safely if you know that 20 V is what your device needs. You do also need to make sure that your 20 V is within a reasonable range; sometimes the adapter will check that for you, and sometimes the device won’t mind. There’s also quite a few USB-C PSU designs with DC input, that will have active conversion onboard – so your 12 V, 19 V and 24 V PSUs can be put to good use with USB-C stuff.

Power, But Not Always For Good

There are caveats to USB-C power. One of them is connecting two dual-role ports together – say, a laptop and a powerbank. Both of them are able to either provide power or use power for charging, and with USB-C, they use the same port for both of these functions. As a result, if you want to charge your fancy laptop from your fancy powerbank, you might be surprised to find your laptop charging your powerbank instead – and same might happen if you connect your phone to your laptop. It’s not often that it happens and it seems that some powerbank manufacturers manage to avoid such an arrangement; however, other products might not fare as well.

This and more fun tales of customers reverse-engineering their own laptops’ third-party lockout behaviour can easily be found

There are provisions for battery level information exchange in USB-C, but I’d guess they are not utilized or implemented by all. Some devices that expose the USB-C layer controls to users, will show you a menu where you can check – in part, modern phones and Chromebooks do that. Powerbanks with their single-button controls, if they even have a button, might not do that, and many laptops aren’t able to ask anything either, so we’ll just have to wait and see towards an eventual solution for manufacturers to standardize upon.

USB-C also lets you implement digital signing for device validity verification. If you can read between the lines, it smells of DRM, and that’s what it is. Some device manufacturers, especially from the HP/Dell/Lenovo dark triad, will implement DRM that makes their laptop throttle its CPU if the charger or the cable is third-party – even if it’s all the same 100 W. It sort of makes sense when Dell does that in cases where they push  6A through a verified-to-work combination of charger, cable, and laptop. But at this point, let’s be fair, the conscientious choice would’ve been to go for EPR and 140 W instead, and throttling is inexcusable either way.

On the topic of laptops – if you ever try to find a laptop that gives more than 5 V on its USB-C port, I wish you good luck. As far as I’m aware, such laptops do not exist, and you’ll want a charger and a powerbank.

Still, We Win

There’s little technical reasons that a switch-mode power supply has to be hard-wired to a certain voltage, and USB-C breaks these chains. Now, you can get 9 V or 20 V at 3 A from a gas station charger and a small cheap board, and you can even charge Li-Ion batteries from a PPS charger. There’s a potential in USB-C capabilities, that we’re only now starting to tap into. As times goes on, proprietary plug and wacky voltage chargers around us will completely die out, and we’ll forget about them, just like we’ve forgotten about all the cellphone data transfer cable standards after MicroUSB took over. How many proprietary data cable ports did Samsung alone create, again?

60 thoughts on “All About USB-C: Power Delivery

  1. I’m a big fan of USB-C for all the reasons you mention above. I do worry about its mechanical robustness though, as it seeing a lot more ‘action’ now that is is the only charging port on many laptops. I’ve noticed in laptops that the barrel jack is typically an easily replicable part/assembly (usually without soldering), whereas the USB-C connector is mostly soldered to the motherboard. That’s not going to be pleasant to replace! Hopefully manufacturers will see the light, and fit them to replicable flexis or carriers very soon.

    1. I am not a fan of USB-C, as I find its robustness a big problem: my phone will not hold a USB-C cable anymore (Two or three text messages is enough to shake the cable loose, and I have a near-empty phone in the morning). It’s hard to get lint out of a USB-C socket (a necessity for pocketable devices like phones), IMHO a lightning socket is mechanically way sturdier and simpler.

      I have not seen the need for more than 5V2A. But I don’t own a laptop, I can see it’s use there.
      Having to get all new cables and chargers is annoying too, as I currently have USB B, Mini-B, Micro-B devices on my desk. USB Micro was promised to be THE standard. When will USB D take over?
      Frankly, other than my 2 Rpi4’s, my phone is the only USBC-device I have.

      Yes, I’m a registered luddite, why?

      1. Yeah, sadly, USB-C on phones can be a pain, at least for now. It seems like a lot of phone manufacturers go for ports of mechanical form-factor that isn’t encountered anywhere else, “these ports don’t exist anywhere outside of this phone model” kind of deal, and of course, they go for the lowest bidder. I’m no mechanical engineer, but, it looks like many of those ports are not battle-tested mechanically, or well-integrated into the chassis to keep the port from flexing and breaking internally. I do hope that phone makers can get their shit together soon!

        1. Just for interest’s sake, I have an XP Pen Star 06 graphics tablet, and the charging cable for it uses a clever recessed micro-USB connector. The cable itself has to go in about 1cm before it reaches the actual socket, and this helps both guide the connector and provide protection from the cable getting yanked out. Quite a clever design. You can see it briefly in the video here at about 0:35:

          1. oh I see! this is cool! I’ve had similar devices.. built a few DIY devices like this too 😅 It’s a good decision for MicroUSB, the extra mechanical support is seriously helpful for keeping your port alive! There’s one tradeoff, you do need a specific cable – I still have an assortment of long solderable MicroUSB plugs from when I couldn’t find suitable cables for such a device and had to build my own.

          2. Haha, so likely only their own cables will fit in the port. Does the standard cover that, or can we expect this to be new a way for companies to ‘differentiate and innovate’ their products, aka try to create a monopoly.

            They do indeed need to be bullied to produce goods that society wants, and not just whatever will make the most profit for them.

          3. I love this! And a well-designed standard could allow for this, specifying not only the form factor of the last 8mm (the metal shell) but also of the next centimetre of plastic, so it could form a better strain relief.
            Otoh, a more fixed cable becomes a liability when tripping incidents happen

          4. This *is* covered by the USB standard (both type C and previous) – the overmold dimensions of the connector are specified.

            They’re maximums, of course, so many cables are smaller, and in that kind of a situation they’ll still work, but be a little more awkward.

            And, of course, there are plenty of cables that violate the overmold specs, but that’s why you buy reputable cables.

      2. Get yourself some of those plastic floss pick things, Use the pick end to dig out the lint from the port. I have had USB-C phones for a while now and whenever the connector starts having the issue of not staying in very well, I clean it out using one of those. Also, get yourself a wireless charger stand for charging at night (assuming you phone supports it).

    2. I have had countless laptops with barrel style and mag-safe charge connectors, and zero failures on any of those despite many tripping over the charge cable incidents that sent the laptop flying. We have one 2.5 year old iPad with USB-C charging, and that port slowly self destructed. Took 2+ hours to replace due to the glued on screen. At least the port was a replaceable part rather than being soldered down. I’m just waiting to start losing USB-C ports on my laptops, as I’m not convinced they will make it past a few years. I don’t expect laptop ports are very replaceable…. We’ll see in the next few years how well it holds up as it becomes more widespread.

      1. My experience with barrel jacks hasn’t been nowhere near as wonderful – I’ve replaced many sockets and cable plugs with my laptops, spent quite a bit of time and money scouring local markets when something broke, and developed an entire ecosystem of barrel jacks adapters and chargers at home at some point just to avoid lugging proprietary chargers from room to room wherever I want to sit down and work. Nowadays, I just get USB-C chargers and they work with everything, and I no longer have to go online looking for replacement cables and sockets for a device of mine in advance. In my laptop repair experience, barrel jacks too have been a failure point and I’m quite glad to see go them go away, together with stuff like microHDMI and proprietary dock ports.

        1. A busted Barrel jack, which I’ve as far as i can recall seen only once is trivial to replace though (if you discount Dell’s pure BS odd version that was super delicate by design and only works properly when it can talk to its own specific ‘smart’ charger – that solution is so shit I’ve seen it fail on a machine that has never moved more than a few inches or been unplugged in its entire life…). It is a much much more common standard so compatible sockets to your PCB footprint are widely available, and a grand total or two conductors to worry about so trivial to fit an incorrect socket if you need/want to – the USB-C connector is full of so much crud from the point of view of being the power input that if it breaks its not very repairable…

          Also has to be said I’ve never seen Barrel jacks as a problem, the sane more standard designed ones are durable as can be, and for the most part devices are so common in barrel shape, voltage and power requirement its trivial and cheap to get extra chargers that work perfectly for any extra locations you want to wire up. USB-C chargers are a comparative minefield of incompatibility and never cheap…

          Proprietary docking ports I can almost wholehearted agree with, small caveat in that the quick latching and self locating nature of some proprietary docking systems are vastly superior to a more universal manual system for some users – and doing that without it being at least a little proprietary to suit the shape of the specific device would be nearly impossible. Though such a fragile solution as the USB-C connector and cable in comparison to most of the docking solutions I’ve seen I’m less sold on…

          1. The “Dell’s pure BS odd version” is just a three-connector barrel jack (pin, inner ring, outer ring). Connectors like that are readily available and have lifecycle ratings as high or higher than other barrel jacks.

            The biggest problem with Dell’s power supply issue was actually at the power supply side (which was made by Delta), which didn’t strain relief the cable properly (… or at all). Which is why you would see failures even if you ripped off the plug end and entirely replaced it. Or ripped the connector off the PCB and replaced it with any other 3-contact connector.

            Barrel jacks actually have lower insert/remove cycle ratings than USB because of the overall play in the connector – you essentially flex the center pin slightly each time.

            Really though the best solution for a device that needs to be charged constantly (rather than a permanently powered device) is a sacrificial connection inside at the connector (via like a flex cable or direct wires). Definitely an advantage for barrel jacks that you could do that easily.

    3. Yeah, it’s a problem – encountered it myself, with a laptop’s USB-C port that was impossible to find a replacement for; it’s a pain in the ass. Macbooks do “carriers”, Framework does too with their expansion cards, but given the high-speed signals, I don’t know if other mfgs will go the same route. Fingers crossed they see the light!

      Also, I used to do laptop repair for a good while, and my observation is that, outside of Thinkpads and a few other kinds of laptops, barrel jacks are typically soldered to the mainboard (and can be a bother to replace too). My guess would be that you’re a Thinkpad user ;-P

  2. I love, love, love USB-C PD. I have it on my phone, my earbuds (well, USB-C but not PD), my (newest) laptop and my soldering iron (it works great!). It’s only failed me with a non-compliant specialty appliance. I’m never buying another consumer device laptop or smaller, that is powered by anything other than compliant USB-C, PD if necessary (devboards and similar, get a pass, though I’m already prioritizing buying ones with a USB-C plug pretty heavily).

    One thing that really aggravates me is non-PD devices/chargers that use USB-C, such as my work laptop, a dell with the double-USB-C docking ports. The split cable, and the smaller single-charger, are non-interoperable with USB PD devices (and it won’t use USB PD itself), meaning I have to lug around two *separate* USB-C charging setups if I’m going anywhere for work more than a day, one for my phone (and maybe personal laptop), and the proprietary one for my work laptop.

    I’m not sure where this fits into the upcoming European regulations. I really hope they regulate out non-standard charging setups that use USB-C

      1. Their top reply in the chain is the exact connector, yes, it’s the one that physically splits in half with a cable stemming from each (probably more manufacturable, as the two halves/cables have rotational symmetry, but the split is useless in function). My previous work laptop used the exact connector in the tweet you posted with the single cable and no split.

      2. Todays Dell’s double USB-C are splitable. I can connect my work laptop with just one USB-C cable (I’m using it regularly with Thinkpad dock) and I can connect one-USB-C devices to the original Dell dock and it works including 2x DP video and Ethernet even with my Samsung S10.

      3. My work bought thousands of this exact charger/dock. Worst experience of my life. For months I could not charge my laptop, or if I could I couldn’t get external displays working. Randomly my ethernet would drop out. Every time the system booted it would throw a warning saying that the power supply wasn’t good enough for the laptop even though it was the power supply sold with the laptop.

        6 Months later. Update firmware for the chip on the motherboard, update the driver on the PC, and update the firmware inside the docking station. Finally the dock sold with the laptop worked. Your articles make me not hate USB-C anymore but I think my frustration was understandable with how absurd that situation is.

  3. I really, really want to be able to use this to make accessories for my phone that don’t just drain its batteries. Is there a USB-C microcontroller that can do this out of the box? Or is that too much to ask?

      1. I watch a lot of videos on my Android tablet, so having a device with physical buttons for the primary media controls (play/pause, volume up/down, etc) would be very useful. Plus, this model of tablet has a bug with its BlueTooth system somehow, so wireless is out for me.

        Not having finished making the first experiment, I have no idea how long the tablet’s battery will last with this plugged in.

        I already have a bunch of TTP223 touch modules for the inputs, but choosing an effective microcontroller is the problem. Being broke and not knowing I was eventually going to want to do this means all the microcontrollers I do have aren’t cooperating…
        >.<

        I've tried the original-version Adafruit Pro Trinket modules, but found out that the keyboard module no longer compiles, and I can't figure out why!

    1. I own one of these “USB3, HDMI and charger input” docks, bought it in a local shop for 9EUR, disassembled it a few times – the build quality is seriously good for what it is and how much it cost! It also helps a lot that USB-C has the entire ‘smart’ part to it, because that alone lets you bypass a lot of safeguards you’d need to spend money implementing in hardware. You can ask questions about what the port is ready to accept, see what the charger is ready to supply, and implement safeguards easily. Basically, majority of what makes this dock have all these features, is the firmware – most of it factory-flashed into the ICs that do USB-C stuff, and that’s quite inexpensive.

  4. “48 V steps into the mix – for up to 240 W”

    That means you can supply nearly a horsepower with three USB-C cables now? Bonkers.

    750 Watt Ebike could use 3 USB-C cables to deliver power to the motor. I know it’s now realistic, but think about that.

  5. Pff. They moved the multiple connector problem to the electronic and software domain, not solving the problem, but hiding it, so earlier you could at least see if you got the wrong power supply, now you don’t. Its a lot neater indeed.
    The EU should have started with a standarised set of power requirements for dc devices and after that mandated a connector for the different requirements. Like for example the eec connector variants.
    Oh well. All will be water under the bridge in 10 years, when something way more easier ( ie more complex) is the new hot thing in power delivery.

    1. It’s definitely has been solved in a significant way, purely because we can now get rid of many barrel jacks and video connectors, as well as a wide variety of different chargers and adapters. Overall, hardware becomes more way more compatible with USB-C. You can still see when the PSU is wrong – the requirements tend to be written on both of the devices, and phones will generally straight up tell you. Plus, a pleasant thing – devices like laptops can nowadays at least be trickle charged from lower-voltage USB-C PSUs, unlike before, when it was “20V or bust”! Making problems be a matter of software/firmware in particular, with all the upgradability available nowadays, does solve quite a bit – if not today, then with tomorrow’s firmware update.

    2. +1 to this almost sufficiently cynical take.
      Though I think really there were already enough stardards for DC power the simple one is just picking say the ATX voltages – been in existence forever, already really really commonly correct for every device so all you need is a connector in a form factor more suitable for external (and probably much much lower power) use.

    3. You absolutely can see if you got the wrong one, it doesn’t work or tells you it’s slow charging. No damage can be done by getting it wrong (Unless you buy noncompliant stuff, which appears to be uncommon).

      Most people just use it to charge things, and transfer data at 2.0 speeds. Anything beyond that requires keeping track of “The good cable” that can do it, but that’s only minorly more work than keeping track of where you put your RCA to 3.5mm cable. It does the common use case amazingly well, and the advanced use case acceptably.

      Having different connectors for different requirements increases the number of items you have to own. Why would I want a 5v and a 12v adapter when I can have 2 adapters that can do either, serving as backup to each other in a pinch?

      What if I want to take my laptop somewhere? Now I have to bring a separate phone and laptop charger, or a 2 in 1 specific to that combination of phone and laptop, that my friends can’t use if they have a tablet that needs a different one of the options?

      Yeah, it’s not like the world would end if I had to bring 2 chargers, but tech isn’t about “good enough”, it’s about making the best product you can without costing too much, and the PD negotiation adds a lot of convenience.

      1. You can have that universal charger without all the failure points and added complexity of negotiation for the voltages and really really easily when you are apparently happy to have so many conductors and more expensive sockets and cables anyway! Just spec one or more of those many many pins as x,y,z voltages! Plus 12V isn’t actually part of the official USB spec anyway, so actually getting adaptors that will do that isn’t certain. Even more easy if you are happy with USB 2 data speeds, so most of those extra conductors are pointless extra expense anyway…

        And you really can’t easily see if you have the wrong one – the quagmire that USB has become means the cable may well be working perfectly for data but won’t do the voltage negotiation BS properly so doesn’t actually provide the right power at all – but the host device won’t know or care, to it its just an older spec USB cable doing its thing. And even if it does pop up that freindly message of ‘this cable is shit or the power supply too weedy’ you then have to spend ages cycling through the heap of these damn cables looking for the one that isn’t broken yet and actually was built to spec in the first place…

        And damage absolutely can be done by getting the ‘wrong’ one – that power supply is now actually capable of dumping way way way more voltage than your device is designed to take, and all that is required for it to do so any one of a myriad of little failures in its initial design or software. Where if the power supply is designed to only produce x volts with dumb parts it actually has to be a real hardware failure to spike excess into the device, and because the expected voltage and power limits are consistent it is simple to provide overload protection to the device from some a failure of power supply…

        1. It’s quite easy to provide overvoltage protections inside a USB-C device, I’ve done it! Add a crowbar with a zener, or one of those “protect from overvoltage” chips, problem solved.

          > the cable may well be working perfectly for data but won’t do the voltage negotiation BS properly

          that’d require a missing/broken CC pin, either a misdesign or a failure.

          > actually getting adaptors that will do that isn’t certain

          you can simply take look at the store listing description, or on the adapter’s box/label, when you buy it! Then, you’ll see if 12V is available – it often is.

          If you have a heap of USB-C cables and none of them are built to spec, you ought to reconsider your purchasing or cable marking decisions! I have a heap of USB-C cables, all of them built to spec, and it’s not hard to remember or see which ones do what. Most of them are USB2, all of those can do up to 60W charging. Then, there’s a few USB3 ones, and one EPR+Thunderbolt one I can use for fancy stuff. I also haven’t had failures with USB-C supplies, despite the complexity – there are quite a few safeguards in USB-C designs, if you look into it!

          The comment reads like quite an exaggeration – I use USB-C day to day, and I just, don’t experience problems like that. I have fried some devices due to barrel jack overvoltage though, spent a lot of time&money fixing broke/nincompatible laptop power supplies, and getting adapters to convert between high-speed data standards of different kinds. Quite glad I no longer have to do any of these things!

          1. All the cables I’ve ever bought are supposed to be full fat do everything in the USB-C spec cables and not from cheap shitty sources either, and they almost without fail have been shit, either not full spec in the first place or broken in short order – probably because the full fat cables require so damn many conductors but even the more premium cables want to stay thin and flexible… (I have exactly 1 cable that actually still bloody works, its doing well having lasted at least a month now and I’ve only been using USB-C at all for about a year, and I don’t even use that device or cables heavily!)

            And as 12v isn’t officially part of the spec anything can say USB-C PD and not do it, while still being 100% compliant. You can’t just read the box.

            Providing protection inside the device is quite possible – even possible to provide protection that should survive contact with the 48v or perhaps it will be 100v in future and reset safely. But that adds costs to a device and as devices SHOULD outlast their spec at the time they were built they really need to be tolerant of the future spec’s top end to survive reliably – which means guessing wildly or having the device rendered useless as that protection system hopefully turns into a single use fuse rather than fails to protect the device at all. Its so much cheaper and easier if the protection ONLY HAS to exist in the power supply, and the devices can then choose to add extra if they feel the need, while knowing the protection they put in will remain relevent or actually require extra smooth power, perhaps for audio reasons.

        2. Real world testing seems to show that damage due to the negotiation failure is very rare.

          And the extra conductors needed for multiple dedicated voltages in one connector would be thicker and heavier than the ones needed for a negotiation line, plus you would need multiple converters in your adapter, which would probably be completely impossible without making them bigger, since chargers are all about pushing the limit of size already. Chargers would cost more, and devices couldn’t take advantage of PPS, they’d have to always regulate current on-device, and not have access to nonstandard voltages.

          The extra connectors aren’t pointless extra expense even though 3.0 is still rare, they are essential for video output on phones and such without having an extra connector, which would be terrible as micro-hdmi doesn’t appear to be mechanically reliable, at least not the common cheap ones.

          The extra complexity doesn’t seem to increase the cost much. Complexity is cheap when you make a billion, physical substance tends to stay expensive. And it doesn’t seem to affect reliability much either.

  6. Respectfully, does anyone else find this writing style jarring? I know nothing about USB-C and I’m sure not going to learn about it from this article- I had to re-read every sentence a few times and paragraphs start with half sentences and thoughts. I didn’t make it far and had to give up. Sorry and thanks though

    1. Sorry to hear that! Not everything is for everyone, and if you have any specific criticisms, please do tell! I do recommend you start with the intro article – my impression has been that it does present a more clear picture of what USB-C does.

      1. It is definitely helpful to type out acronyms the 1st time they show up.
        SPR – Standard Power Range
        EPR – Extended Power Range
        PPS – Programmable Power Supply
        AVS – Adjustable Voltage Supply

        And there is a big difference between TLA (Tree Letter Acronym) and TLA (Two Letter Acronym ) ;)

      2. Thanks for being awesome and open to suggestions. I’d say it is more (hopefully) constructive feedback to make it more accessible to new learners than criticism!
        I’ll definitely go back to the start of the article series. I tried to dip my toe in and instead tried to drown myself.

        I think in general it is a topic with a lot of common sense knowledge that isn’t so common to me. The acronym soup is murder: “Imagine a USB-C dock with a USB3 port, a HDMI port, and a USB-C charger input port” but I’m not sure how to get around that other than periodically re-defining the acronyms (PD, CC, etc).

        I’m not a professional writer or anything, but a recurring theme is using the same word twice in a sentence which is hard to follow, as well as including more than one thought per sentence. One random example:

        “USB-C power supplies always support 5 V and some are limited to that, but support for higher voltages is where it’s at” is challenging to parse. It uses “support” twice, but in two different ways: power supplies support 5 V but also, ?the USB-C standard? supports higher voltages?… I can’t figure that one out and again, it is probably obvious to the non-ignorant. The first “limited to that” isn’t really clear if it referring to UCB-C itself, the relationship of the power supply to USB-C, or what. I’m sure it is not technically correct but more clear to me would be something like “The USB-C standard requires the support of 5V. Some power supplies, however, can provide higher voltages in discrete steps, specifically, 5, 9, 15 and 20V.”

        Anyway, it is a cool topic and thanks again.

  7. I wonder if I will ever accept any C device. Those “contacts” half as big as they would have been are smaller than fleas. The only problem with barrel jacks was the bastardization that happened with them. And even those negative pin ones, guitar FX pedals I’m looking at you.

    Hams have this problem so does the rest of the electric world. We got 120 and 230 AC but outside of some big things everything on wheels anywhere Earth gives 12Volts. Yet there is no worthwhile connector to the humble car battery. Cigar sockets suck and there is some obscure Molex or something connector that some hams use to be able to connect up various gear and power up on field day. 12Volts is left out of USB PD and power in general.

    1. That obscure molex is the ARES standard for Anderson PowerPoles(You can click them together in different ways, ARES has a standard hams like). It’s a very good connector, but it’s not 12v, it’s 13.8ish, and it’s definitely not 12v when it comes from a car(Car power is noisy as all heck).

      It’s a wonderful standard but it’s in its own category, since it’s unregulated and could be like 14.4v, and it does nothing to keep you from misconnecting stuff.

      USB-Cs flea sized connectors have proven themselves to be really reliable, even in the real world on cheap equipment using maybe counterfeit parts, which is something many connectors don’t have, things like 1/4″ are only reliable with high quality cables. Most any braided C with a strain relief will probably last years.

      I was worried about the digitally negotiated power messing up and frying stuff, but that’s just not happening in the real world at any noticable rate.

      I do think leaving real regulated 12v out of PD was a mistake due to all the legacy stuff like routers.

      I would have also liked to see a bit more explicit support for solar panels in the protocol, and for daisy chaining power banks(For things like strings of light fixtures that can pass power to the next in a sane efficient way so they split the available power).

    2. The ability to avoid it is going to start looking like black magic I expect, at least for remotely portable and/or low enough power devices. As it does seem everything is going to go to this oh so standard standard where the only way to know it actually does something is try it. Which seems to go for the cables and devices both.

      I want to like USB-C as it does have a few nice points, but it is just so daft overall.

  8. I, for one, am damned uncomfortable with having a single connector type through which very different voltages can be sent. Seems very risky to have nothing more than some negotiations in software preventing 20V being unleashed on to a device which can’t tolerate more than 5V. They should not have designed a single cable for all these different voltages, there should have been a 5V connector, a 9V connector, a 20V connector… all mechanically incompatible with a requirement that any converter adapter would contain step-down or step-up circuits as appropriate. But if they insisted on it then the spec should have atleast required all lower voltage devices with a USB-C port on them to have the correct protection circuitry to be unharmed even if the device only wants 5V and gets exposed to the maximum voltage of the extended range, this could perhaps be done with certain zener diode arrangements.

  9. Very nice, my favourite part of this series so far. I too am amazed by the futuristic capabilities of USB-PD docks, and how cheap they are!

    I am wondering, though – is there such a thing as a USB-C charger that could output 5v and 20v (or 15v, or 12v) *at the same time*? Currently I’m powering a thin client (acting as a file server) via a single Xbox PSU, which outputs both 12v – great for the PC itself as well as the HDD’s – and also 5v which helps for those same HDD’s, because the thin client itself can only support powering a single SATA drive – the other drives need external power.

    It works fine so far, though I my with 3 HDD’s max because the aforementioned Xbox PSU only has 1A on the 5v output. And the typical HDD will eat 0.3A from what I recall measuring.

    So, is this such a thing? Could it be possible to get a USB-C power supply that outputs both 5v and 12v (or higher, I do have DC-DC voltage converters)? Or will I have to bite the bullet and just get a traditional ATX power supply?

    1. With a dc-dc converter you should be able to just feed the needed voltage from the existing voltage … oh wait, with USB-C you have to negotiate that first. I’ll skip, my old-school wiring just works and is repairable without the need of unverifiable and in times of need probably unobtainable silicon/software blobs.

  10. Calling big dongles for docks is one of those things that somehow invokes the emotional response that can best be described as “I’m not mad, I’m dissapointed, VERY dissapointed”

  11. This is amazing and clever until there’s a firmware vulnerability in a USB-PD charger and some rogue android app starts reflagging them to ignore PD communications and just supply the max voltage. But don’t worry, because everything is secure, particularly “embedded” stuff like chargers.

    Or a Chinese supplier cheap out and has their cable report it handles a higher wattage than it does. But don’t worry, they’d never do that.

    1. People also worried about laptop batteries being reprogrammed to explode. Just because you can think up a scenario doesn’t mean it’s likely. No competent hardware engineer leaves power supply protection measures purely up to software, even if it’s just adding a fuse or a zener.

  12. If I have a AC to DC converter (rectifier) in my basement and power 20 LED recessed lights in my home with DC lines daisy chained from one to another then, aside from the basement, would my recessed lights need to be inspected (North America)? Would connectors have to exist within a junction box? (that would be weird). If an electrical inspector started to give you a hard time, could you just tell him/her, “bug off, this ain’t your domain”? What if I ran greater than 48v 5amps DC up from my basement and then split ’em off to other rooms? At what point would this be breaking “code” (in most cases N. America)??

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.