Hi,

This photo,
https://commons.wikimedia.org/wiki/File:USB-OTG_Setup_IMG_2342.JPG
suggests, to me at least, that the phone can be charged while the
keyboard, mouse and flash store are used. True?

If this earbud set is connected via a USB A to C adapter, can audio
output also be produced while power is connected?
https://www.ebay.ca/itm/285288167463

Thx, ... P.

On 2024-03-09, peter@easthope.ca <peter@easthope.ca> wrote:
>
> This photo,
> https://commons.wikimedia.org/wiki/File:USB-OTG_Setup_IMG_2342.JPG
> suggests, to me at least, that the phone can be charged while the
> keyboard, mouse and flash store are used. True?

There are altogether too many unknowns here. Looking at the date
of the photo - 2015 - some phones of that era could support USB
OTG but were unable to supply power to connected peripherals,
meaning you would need a Y cable to supply that separately. Others
could do the job inclusive of power delivery, others couldn't work
in OTG mode at all. Generally you couldn't use OTG and charge
simultaneously, it's a role reversal that early USB wasn't designed
for.

Things are different with USB-C but of course that is much later.

--
Andrew Smallshaw
andrews@sdf.org

On 9 Mar 2024 at 21:54:51 GMT, "Andrew Smallshaw" <andrews@sdf.org> wrote:

> On 2024-03-09, peter@easthope.ca <peter@easthope.ca> wrote:
>>
>> This photo,
>> https://commons.wikimedia.org/wiki/File:USB-OTG_Setup_IMG_2342.JPG
>> suggests, to me at least, that the phone can be charged while the
>> keyboard, mouse and flash store are used. True?
>
> There are altogether too many unknowns here. Looking at the date
> of the photo - 2015 - some phones of that era could support USB
> OTG but were unable to supply power to connected peripherals,
> meaning you would need a Y cable to supply that separately. Others
> could do the job inclusive of power delivery, others couldn't work
> in OTG mode at all. Generally you couldn't use OTG and charge
> simultaneously, it's a role reversal that early USB wasn't designed
> for.
>
> Things are different with USB-C but of course that is much later.

But can even USB-C accept a signal input (keyboard and mouse for instance) and
unrelated signal output (audio for instance) on the same physical socket? I
ask only out of curiosity.

--
Roger Hayter

In article <2219572717.2fe80b82@uninhabited.net>, Roger Hayter <roger@hayter.org> wrote:
> But can even USB-C accept a signal input (keyboard and mouse for
> instance) and unrelated signal output (audio for instance) on the
> same physical socket? I ask only out of curiosity.

The setup is rarely tried. Appears that nobody can reply.

A keyboard or mouse sends tiny amounts of data; a few bytes per
second. In principle the link should be able to fit that in without
significant drop-outs in audio output. WIth so many details involved,
a general answer is difficult. When all else fails, we can test and
see what happens. =8~)

Regards, ... P.

On 12 Mar 2024 at 16:29:31 GMT, "peter@easthope.ca" <peter@easthope.ca> wrote:

> In article <2219572717.2fe80b82@uninhabited.net>, Roger Hayter
> <roger@hayter.org> wrote:
>> But can even USB-C accept a signal input (keyboard and mouse for
>> instance) and unrelated signal output (audio for instance) on the
>> same physical socket? I ask only out of curiosity.
>
> The setup is rarely tried. Appears that nobody can reply.
>
> A keyboard or mouse sends tiny amounts of data; a few bytes per
> second. In principle the link should be able to fit that in without
> significant drop-outs in audio output. WIth so many details involved,
> a general answer is difficult. When all else fails, we can test and
> see what happens. =8~)
>
> Regards, ... P.

So you could use an overarching protocol which was bilateral (ethernet
anyone?) and send and separate messages both ways. What I was aksing was
whether within the USB protocol there was provision for using some pins in for
one electrical signal (bi or uni-directional) and other pins for another
electrical signal. Because that would be the only way to do it without special
software on at least one of the devices.

--
Roger Hayter

On 2024-03-12, Roger Hayter wrote:
> On 12 Mar 2024 at 16:29:31 GMT, "peter@easthope.ca" <peter@easthope.ca> wrote:
>
>> In article <2219572717.2fe80b82@uninhabited.net>, Roger Hayter
>> <roger@hayter.org> wrote:
>>> But can even USB-C accept a signal input (keyboard and mouse for
>>> instance) and unrelated signal output (audio for instance) on the
>>> same physical socket? I ask only out of curiosity.
>>
>> The setup is rarely tried. Appears that nobody can reply.
>>
>> A keyboard or mouse sends tiny amounts of data; a few bytes per
>> second. In principle the link should be able to fit that in without
>> significant drop-outs in audio output. WIth so many details involved,
>> a general answer is difficult. When all else fails, we can test and
>> see what happens. =8~)
>>
>> Regards, ... P.
>
> So you could use an overarching protocol which was bilateral (ethernet
> anyone?) and send and separate messages both ways. What I was aksing was
> whether within the USB protocol there was provision for using some pins in for
> one electrical signal (bi or uni-directional) and other pins for another
> electrical signal. Because that would be the only way to do it without special
> software on at least one of the devices.
>

There's only one set of pins (well, two sets if you count USB2 / USB3 --
but you can only use one set at a time).

USB2 -> D+/D- (Bi-directional / Half Duplex)
USB3 -> TX1+/TX1- and RX2+/RX2- (optionally Full Duplex, IIRC)

As I recall the "Tx" pair is "Host Transmit to Peripheral", and "Rx" is
"Host Receive from Peripheral", but it's been a while since I read up on
the USB3 / USB-C implementations.

--
|_|O|_|
|_|_|O| Github: https://github.com/dpurgert
|O|O|O| PGP: DDAB 23FB 19FA 7D85 1CC1 E067 6D65 70E5 4CE7 2860

On 12 Mar 2024 at 19:27:04 GMT, "Dan Purgert" <dan@djph.net> wrote:

> On 2024-03-12, Roger Hayter wrote:
>> On 12 Mar 2024 at 16:29:31 GMT, "peter@easthope.ca" <peter@easthope.ca> wrote:
>>
>>> In article <2219572717.2fe80b82@uninhabited.net>, Roger Hayter
>>> <roger@hayter.org> wrote:
>>>> But can even USB-C accept a signal input (keyboard and mouse for
>>>> instance) and unrelated signal output (audio for instance) on the
>>>> same physical socket? I ask only out of curiosity.
>>>
>>> The setup is rarely tried. Appears that nobody can reply.
>>>
>>> A keyboard or mouse sends tiny amounts of data; a few bytes per
>>> second. In principle the link should be able to fit that in without
>>> significant drop-outs in audio output. WIth so many details involved,
>>> a general answer is difficult. When all else fails, we can test and
>>> see what happens. =8~)
>>>
>>> Regards, ... P.
>>
>> So you could use an overarching protocol which was bilateral (ethernet
>> anyone?) and send and separate messages both ways. What I was aksing was
>> whether within the USB protocol there was provision for using some pins in for
>> one electrical signal (bi or uni-directional) and other pins for another
>> electrical signal. Because that would be the only way to do it without special
>> software on at least one of the devices.
>>
>
> There's only one set of pins (well, two sets if you count USB2 / USB3 --
> but you can only use one set at a time).
>
> USB2 -> D+/D- (Bi-directional / Half Duplex)
> USB3 -> TX1+/TX1- and RX2+/RX2- (optionally Full Duplex, IIRC)
>
> As I recall the "Tx" pair is "Host Transmit to Peripheral", and "Rx" is
> "Host Receive from Peripheral", but it's been a while since I read up on
> the USB3 / USB-C implementations.

Usbc seems to have about 20 pins, that's why I asked.

--
Roger Hayter

On 2024-03-12, Roger Hayter wrote:
> On 12 Mar 2024 at 19:27:04 GMT, "Dan Purgert" <dan@djph.net> wrote:
>> On 2024-03-12, Roger Hayter wrote:
>>> [...]
>>> So you could use an overarching protocol which was bilateral (ethernet
>>> anyone?) and send and separate messages both ways. What I was
>>> aksing was whether within the USB protocol there was provision for
>>> using some pins in for one electrical signal (bi or uni-directional)
>>> and other pins for another electrical signal. Because that would be
>>> the only way to do it without special software on at least one of
>>> the devices.
>>>
>>
>> There's only one set of pins (well, two sets if you count USB2 / USB3 --
>> but you can only use one set at a time).
>>
>> USB2 -> D+/D- (Bi-directional / Half Duplex)
>> USB3 -> TX1+/TX1- and RX2+/RX2- (optionally Full Duplex, IIRC)
>>
>> As I recall the "Tx" pair is "Host Transmit to Peripheral", and "Rx" is
>> "Host Receive from Peripheral", but it's been a while since I read up on
>> the USB3 / USB-C implementations.
>
> Usbc seems to have about 20 pins, that's why I asked.

Yeah, that's just to allow the connector to be flipped over, and still
connect to the host / peripheral. The host/peripheral ports themselves
only have one orientation.

--
|_|O|_|
|_|_|O| Github: https://github.com/dpurgert
|O|O|O| PGP: DDAB 23FB 19FA 7D85 1CC1 E067 6D65 70E5 4CE7 2860

On 2024-03-09 23:09, Roger Hayter wrote:
> On 9 Mar 2024 at 21:54:51 GMT, "Andrew Smallshaw" <andrews@sdf.org> wrote:
>
>> On 2024-03-09, peter@easthope.ca <peter@easthope.ca> wrote:
>>>
>>> This photo,
>>> https://commons.wikimedia.org/wiki/File:USB-OTG_Setup_IMG_2342.JPG
>>> suggests, to me at least, that the phone can be charged while the
>>> keyboard, mouse and flash store are used. True?
>>
>> There are altogether too many unknowns here. Looking at the date
>> of the photo - 2015 - some phones of that era could support USB
>> OTG but were unable to supply power to connected peripherals,
>> meaning you would need a Y cable to supply that separately. Others
>> could do the job inclusive of power delivery, others couldn't work
>> in OTG mode at all. Generally you couldn't use OTG and charge
>> simultaneously, it's a role reversal that early USB wasn't designed
>> for.
>>
>> Things are different with USB-C but of course that is much later.
>
> But can even USB-C accept a signal input (keyboard and mouse for instance) and
> unrelated signal output (audio for instance) on the same physical socket? I
> ask only out of curiosity.

Certainly.

That's normal USB functionality. It is a bus. You need a hub, though, in
order to split the cables, one for each device.

The only problem is which device supplies the power and in what direction.

--
Cheers, Carlos.

On 2024-03-12, peter@easthope.ca <peter@easthope.ca> wrote:
> In article <2219572717.2fe80b82@uninhabited.net>, Roger Hayter <roger@hayter.org> wrote:
>> But can even USB-C accept a signal input (keyboard and mouse for
>> instance) and unrelated signal output (audio for instance) on the
>> same physical socket? I ask only out of curiosity.
>
> The setup is rarely tried. Appears that nobody can reply.
>
> A keyboard or mouse sends tiny amounts of data; a few bytes per
> second. In principle the link should be able to fit that in without
> significant drop-outs in audio output. WIth so many details involved,
> a general answer is difficult. When all else fails, we can test and
> see what happens. =8~)

It'll be fine, I use my laptop all day everyday, USB headset for
phone, USB network adapter on the same host port for connectivity
(including the phone). Works fine. USB was designed for that from
the outset. The retro-fit is splitting power delivery from the
host vs device relationship. The early versions of USB always
contrived that the host provided the power, USB-C allows you to
reverse that relationship so you can have e.g. a phone or laptop
on charge via USB while still controlling connected peripherals.

--
Andrew Smallshaw
andrews@sdf.org

On 2024-03-13, Dan Purgert <dan@djph.net> wrote:
> On 2024-03-12, Roger Hayter wrote:
>> On 12 Mar 2024 at 19:27:04 GMT, "Dan Purgert" <dan@djph.net> wrote:
>>>
>>> There's only one set of pins (well, two sets if you count USB2 / USB3 --
>>> but you can only use one set at a time).
>>>
>>> USB2 -> D+/D- (Bi-directional / Half Duplex)
>>> USB3 -> TX1+/TX1- and RX2+/RX2- (optionally Full Duplex, IIRC)
>>>
>>> As I recall the "Tx" pair is "Host Transmit to Peripheral", and "Rx" is
>>> "Host Receive from Peripheral", but it's been a while since I read up on
>>> the USB3 / USB-C implementations.
>>
>> Usbc seems to have about 20 pins, that's why I asked.
>
> Yeah, that's just to allow the connector to be flipped over, and still
> connect to the host / peripheral. The host/peripheral ports themselves
> only have one orientation.

USB-C supports SuperSpeed which uses two additional pairs on top
of USB2 and also allows for full-duplex. That is what the additional
contacts on a USB3 A plug are for.

--
Andrew Smallshaw
andrews@sdf.org

On 2024-03-19, Andrew Smallshaw wrote:
> On 2024-03-13, Dan Purgert <dan@djph.net> wrote:
>> On 2024-03-12, Roger Hayter wrote:
>>> On 12 Mar 2024 at 19:27:04 GMT, "Dan Purgert" <dan@djph.net> wrote:
>>>>
>>>> There's only one set of pins (well, two sets if you count USB2 / USB3 --
>>>> but you can only use one set at a time).
>>>>
>>>> USB2 -> D+/D- (Bi-directional / Half Duplex)
>>>> USB3 -> TX1+/TX1- and RX2+/RX2- (optionally Full Duplex, IIRC)
>>>>
>>>> As I recall the "Tx" pair is "Host Transmit to Peripheral", and "Rx" is
>>>> "Host Receive from Peripheral", but it's been a while since I read up on
>>>> the USB3 / USB-C implementations.
>>>
>>> Usbc seems to have about 20 pins, that's why I asked.
>>
>> Yeah, that's just to allow the connector to be flipped over, and still
>> connect to the host / peripheral. The host/peripheral ports themselves
>> only have one orientation.
>
> USB-C supports SuperSpeed which uses two additional pairs on top
> of USB2 and also allows for full-duplex. That is what the additional
> contacts on a USB3 A plug are for.
>

Except what Roger is talking about is the "spare(tm)" contacts in a
USB-C cable :

GND TX1+/- VBUS CC1 D+/- SUB1 VBUS RX2-/+ GND <-- Used

GND RX1+/- VBUS SUB2 D-/+ CC2 VBUS TX2-/+ GND <-- "spare(tm)"

And that's assuming all the wires are actually *in* the cable, and the
manufacturer didn't just cheap out and common things "vertically" (e.g.
TX2+ is common to RX2+, like in "Amazon Basics" cables).

--
|_|O|_|
|_|_|O| Github: https://github.com/dpurgert
|O|O|O| PGP: DDAB 23FB 19FA 7D85 1CC1 E067 6D65 70E5 4CE7 2860