Here's some data of interest for you Apple fans. Below is a capture of a trace from a (fully charged) Apple 2016 Macbook Pro 13 w/ Bar, its 2m "rectangular box" cord, and 61w charger. There's some interesting behaviors.
tl;dr: there's a lot more going on with an Apple charger than with a generic charger!

First, notice the Apple 61w charger lacks a 15v voltage level! This means it will not charge certain laptops optimally. This also means the Nintendo Switch dock will very likely not work properly. :(

(Corrected 5/14:) It also only has 5v/2.4a. This means some devices like Google Pixel C tablet owners are sightly impacted. It supports 9v/2.6a... but only with the screen off. With the screen on, it goes to 5v/2.4a, slower than stock, because of said quirk with the Apple charger. Same applies for Nexus phones which rely on 5v, but with non-PD current advertisements.

Note the PD1.0 spec required 12v, but was quickly split to 9v/15v in PD2.0. I think this is because the first products to hit the market encountered issues; from what I understand 12v may have been too high for certain applications, but too low for others. For phones, 12v is considered "extremely high voltage" and special power ICs are needed -- which take up more space, cause more heat, and are more expensive. All of those are bad given the constraints.

Second, note the extremely atypical traffic! The charger does not probe for the eMarker. (This isn't strictly needed as all USB-C cords are required to be good for at least 3a/60w.) Instead, it begins a very noisy back-and-forth communication with the Macbook lasting 1000+ messages over a period of 5 seconds. It uses multiple Vendor Defined Messages (VDM) in rapid sequence, utilizing an Apple-specific AltMode. This includes significant use of Unstructured VDMs. As these are vendor-specific and proprietary, (and Unstructured VDMs are by definition,) I have few clues what this traffic is for. I'm attempting to locate relevant documentation now.

Typically a charger and device merely negotiate a power level. If thorough, they might also do some probing back and forth for SNK and SRC capabilities. However the Apple charger and laptop do a much more complicated mating ritual. (Pun intentional. I make no apologies.)

Third, as a minor note, check out the voltage clamping when the cord is disconnected. Unlike most other chargers on the market, the Apple USB-C chargers use a constant-current source to advertise basic Type-C current level. This is just interesting to note from an electrical perspective, as it hints at the level of attention to engineering detail in the adapter.

Fun fact: per its DISC_ID reply, the Macbook Pro does not claim to have been submitted to the USB-IF Certification process. :) This XID is merely a placeholder test ID number, but is left blank here. I am certain it was rigorously engineered nonetheless.

Also note the Macbook claims to be both a USB host, and USB device! This is likely refers to the "Target Disk" mode. As I only have one Macbook I won't be able to play around with this, but it's interesting to see traces of it in the... well... traces. The presence of Intel Thunderbolt 3, DisplayPort, and some weird "Apple mode" are advertised as well.

(Post corrected to fix a bitmask error with DISC_ID command. This is the "Apple mode" of interest. I regret the error.)

#USB #TypeC #USBC
PhotoPhotoPhotoPhoto
5/14/17
4 Photos - View album
Shared publiclyView activity