On Feb 5, 2017, 15:06, at 15:06, Julius Lehmann julius@devpi.de wrote:
Regarding USB C you are right that there are four differential pairs on the connector. However, for communication only two pairs are used because the USB 3.1 protocol only uses two. On the device that has a USB C connector you need a dedicated controller that manages the orientation of the plug. Then the two data lines are connected to the right pins on the connector through a multiplexer. The other two pairs on the connector are for no use in USB mode but if you want to use USB C in alternate mode (e.g. DisplayPort), signals can be rerouted.
You can find all the information on usb.org (zip archive with full specification).
This basically means that the EOMA68 standard has two differential pairs that are not used for USB but if your intention was to make it available for USB C, an internal multiplexer would be needed.
Kind regards Julius Lehmann
On Feb 5, 2017, 14:20, at 14:20, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware:
https://www.crowdsupply.com/eoma68
On Sun, Feb 5, 2017 at 11:44 AM, Internet internet@devpi.de wrote:
Hey there!
I was wondering about the USB3.1 pinout in the EOMA68 standard. In
my
research on the USB Type C connector I found out that USB3.1 (Superspeed+) has only two differential signal pairs (one for
receiving,
one for transmitting)
http://pinoutguide.com/Slots/usb-type-c_pinout.shtml
two for receiving, two for transmitting, each 5gbit/s for a total of 10gb/sec. the connector and thus the pinouts are rotatable by 180 degrees so the pinouts on the first 12 pins (top row) are *identical* layout to the second 12 pins (bottom row) except in the reverse order. D+ and D- (the USB2 pins) are *missing* on one row of the receptacle so that you don't connect them up twice (which would be bad).
you *may* be thinking of USB 3.0 which is connector-compatible with USB 2.0 and is what you find on most laptops and desktop computers these days:
http://pinoutguide.com/Slots/usb_3_0_connector_pinout.shtml
that is *only* one set of (5gbit/sec) tx-rx differential pairs.
scared me for a minute that i'd messed it up... :)
l.
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Feb 5, 2017 at 9:30 PM, Internet internet@devpi.de wrote:
Regarding USB C you are right that there are four differential pairs on
the connector. However, for communication only two pairs are used because the USB 3.1 protocol only uses two. On the device that has a USB C connector you need a dedicated controller that manages the orientation of the plug. Then the two data lines are connected to the right pins on the connector through a multiplexer. The other two pairs on the connector are for no use in USB mode but if you want to use USB C in alternate mode (e.g. DisplayPort), signals can be rerouted.
You can find all the information on usb.org http://t224.mjt.lu/lnk/AEUAHWykN94AAAAAAAAAAGbE2wkAASHlFNUAAAAAAAZHgABYl5lfeaEfW2hMTR63ZGLMHX5RrAAGAjk/1/ZviPu0ichkU-Gm4llF2oBw/aHR0cDovL3VzYi5vcmc (zip archive with full specification).
This basically means that the EOMA68 standard has two differential pairs that are not used for USB but if your intention was to make it available for USB C, an internal multiplexer would be needed.
argh! no, it was a misunderstanding. ha. very funny. it means i accidentally made room for 4 extra wires... which could be used to upgrade the *second* USB port to USB3.1 as well.
that's frickin hilarious.
l.
I'm glad to hear the standrd itself is correct. The only thing I would still like to know is the mapping of the two USB ports. So, which superspeed wires correspond to which usb 2.0 wires?
Julius Lehmann
On Feb 5, 2017, 22:52, at 22:52, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Feb 5, 2017 at 9:30 PM, Internet internet@devpi.de wrote:
Regarding USB C you are right that there are four differential pairs
on
the connector. However, for communication only two pairs are used
because
the USB 3.1 protocol only uses two. On the device that has a USB C connector you need a dedicated controller that manages the
orientation of
the plug. Then the two data lines are connected to the right pins on
the
connector through a multiplexer. The other two pairs on the
connector are
for no use in USB mode but if you want to use USB C in alternate
mode (e.g.
DisplayPort), signals can be rerouted.
You can find all the information on usb.org
(zip archive with full specification).
This basically means that the EOMA68 standard has two differential
pairs
that are not used for USB but if your intention was to make it
available
for USB C, an internal multiplexer would be needed.
argh! no, it was a misunderstanding. ha. very funny. it means i accidentally made room for 4 extra wires... which could be used to upgrade the *second* USB port to USB3.1 as well.
that's frickin hilarious.
l.
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Mon, Feb 6, 2017 at 5:57 AM, Internet internet@devpi.de wrote:
I'm glad to hear the standrd itself is correct.
it's not!!! it just doesn't affect the current cards.
The only thing I would still like to know is the mapping of the two USB ports.
me too
So, which superspeed wires correspond to which usb 2.0 wires?
i'll have to revise the standard.
l.
A standard that provides for two USB3.1 ports is a good thing. Isn’t it?
In fact, USB3.1 availability might be unique in this small board world. I watched a long presentation where Jon Hall said he couldn’t find one (OpenHours #32 - 96Boards discussion with Jon "maddog” Hall).
No idea what standard revision and chain of components is required to make this work, so I’ve got no idea how to help here. Sorry.
Thanks, Nick
On 6 Feb 2017, at 06:04, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
On Mon, Feb 6, 2017 at 5:57 AM, Internet internet@devpi.de wrote: I'm glad to hear the standrd itself is correct.
it's not!!! it just doesn't affect the current cards.
The only thing I would still like to know is the mapping of the two USB ports.
me too
So, which superspeed wires correspond to which usb 2.0 wires?
i'll have to revise the standard.
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Mon, Feb 6, 2017 at 10:21 AM, Nick Hardiman nick@internetmachines.co.uk wrote:
A standard that provides for two USB3.1 ports is a good thing. Isn’t it?
yeah it is :)
In fact, USB3.1 availability might be unique in this small board world.
well there first has to actually *be* a SoC that has USB 3.1!
I watched a long presentation where Jon Hall said he couldn’t find one (OpenHours #32 - 96Boards discussion with Jon "maddog” Hall).
mmm... not a fan of 96boards after their spectacular failure to respond publicly to my questions about their incompetence in designing the standard *without consulting anybody*
big fan of jon hall though.
No idea what standard revision and chain of components is required to make this work, so I’ve got no idea how to help here. Sorry.
it's ok i just need to make a decision and document it. it'll be along the lines of "what would a typical SoC have as pinouts and how can PCB track cross-overs be minimised under the circumstances".
i _should_ have placed the pins for the SStx/rx next door to each of their respective USB2 tracks... but that's not possible, now.
arm-netbook@lists.phcomp.co.uk
-
Internet -
Luke Kenneth Casson Leighton -
Nick Hardiman