USB type c , USB3.0/3.1 and USB Power delivery.

May 11, 2008
22,531
1,468
126
I have to read it all to help co-develop a circuit for work.

The differences of how USB Power Delivery is communicated makes it not that easy. My head hurts from reading all day. And trying to find all the chips for the desired proto design.

A small summary :
The newest usb standard for power delivery is 2.0.
But it is interesting.
The old USB power requirements were 5V @ 500mA.
Now we can choose by use of negotiation :
5V @ 900mA.
5V @ 1.5A.
5V @ 3A.
12V @ 3A
12V @ 5A.
20V @ 3A
20V @ 5A.

And even have 12V at the usb power pins of the new type c connector with a maximum current of 5A equally divided over 4 pins.
With a maximum voltage of 20V @ 5A for 100W of power that can be provided when supported. 5V @ 500mA (or 900mA, not sure about that)is the default voltage. But the USB PD protocol allows to communicate for higher voltages and currents when more power needed.





Through the USB PD there is even more possible, the superspeed differential data line pairs (4 of them for a total of 10Gb/second) can also be used for other signals than USB , for example display port signals or even pci-e signals. This will be negotiated over the USB PD communication channel.

The new type c connector is a great design, just like the micro usb connector. The plug on the cable holds the spring contacts, the receptable on the device the sturdy contacts. Since the springs will wear out quicker, this means that the devices do not have to be replaced just because of a worn out connector. Just a new cable will do.

I am really happy to be honest. Another step up for me. The last designs i helped with turned out ok. I am getting a littler more confidence (That i do not have that nasty feeling that i have forgotten something )each iteration.
With software development, when you make an error, you reprogram the MCU, test the new code and it works. Software development allows for trial and error methods when needed. But an expensive hardware design with schematics and with printed circuit boards that have been drawn and made, every HW engineers wants to get it right the first time. Even in the prototype stage.
That means going through datasheets with a fine comb.
 
Last edited:
May 11, 2008
22,531
1,468
126
This application note sums it up nicely :
I started with wiki then this Application Note AN1953 and now i am at the USB org documents and have been doing some selection of a few chips .

Introduction to USB Type-C - Microchip

http://www.microchip.com//wwwAppNotes/AppNotes.aspx?appnote=en574276


edit :
note that the drawings Figure4 and Figure 5 of the type c receptacle contains an error with respect to the naming convention of the superspeed data pairs
 
Last edited:
May 11, 2008
22,531
1,468
126
Forgot to mention, that the USB 3.0 or 3.1 version data only use 2 pairs (4 data wires). One RX pair and one TX pair.
The type c connector has 4 pairs (8 wires) because the connector can be flipped. But when the alternate mode is used, these 8 pins and the sideband use pins can be configured for other uses.
 
May 11, 2008
22,531
1,468
126
Those documents of USB.org really help out. It is all a lot more clear now.
The protocol, the hardware, the mechanics.
Before yesterday i had heard of USB 3.0 and some way of providing power but i had no idea about how it works or why. It is fun to see that from the beginning of USB to now, there is a clear evolution going on. Especially because backwards compatibility must be maintained.

Now i know enough to select the appropriate electronic parts and make a first concept design as per customer wishes to discuss further at work with my colleagues. But that is for Monday.