usb 2/3 power question

Hi All,

Paul will know this one off the top of his head.

If I plug a USB 2 peripheral into a USB 3.0/3.1 port on my
computer, do I still get the USB 3.0/3.1 power? Or
do I get throttles back to 500 ma?

Many thanks,
-T

T wrote:
Hi All,

Paul will know this one off the top of his head.

If I plug a USB 2 peripheral into a USB 3.0/3.1 port on my
computer, do I still get the USB 3.0/3.1 power? Or
do I get throttles back to 500 ma?

Many thanks,
-T

Well, it's a mystery for a standard USB3 connection.
There is still only one VBUS pin. VBUS would be the
limiting factor, as there are additional ground contacts.

There is a PD spec, but I don't even want to go there...
That's a huge spec, if I could find it.

Pin Signal name

1 VBUS
2 D-
3 D+
4 GND

5 StdA_SSRX-
6 StdA_SSRX+
7 GND_DRAIN
8 StdA_SSTX-
9 StdA_SSTX+

Shell Shield (GND)

*******

There's a whiff of a hint here.

https://www.makeuseof.com/tag/usb-30...ogy-explained/

"Power

The unit load has been increased to 150 mA
and a configured device can draw up to six unit loads,
which adds up to 900 mA. === OK, so a load of baloney so far

This exceeds USB 2.0 by 80% and leads to faster recharging
or powering of more than four devices from a single hub
[i.e. six unit loads or six keyboards or six mice as unit loads]

In addition, the minimum device operating voltage was dropped
from 4.4 V to 4 V, which saves energy. === Aha!
"

So the goal post that was moved, was cable drop. By making the
peripheral IC more tolerant of voltage drop in the cable, they
could arrange for a higher current draw spec.

*******

On a computer, there is "accounting" and in hardware there
are "policemen".

On a desktop, we see "policemen" like this:

1) A *pair* of USB2 ports is protected by a Polyfuse
marked "110" or 1.1 amps. If both devices drew 500mA,
the Polyfuse would be getting a bit soft. If only one
USB2 port is being used, then the desktop can (almost)
power a 2.5" hard drive (which draws 1.0 amps at startup).
It depends on the details of the hard drive itself,
as to whether the Polyfuse is going to get too soft,
before spinup finishes.

2) On a USB3 card, a *single* port is protected
with an "F200" Polyfuse. That tells us the policeman
on the desktop allows 2000mA versus the claimed 900mA.

These imply something about the ampacity of the USB pins. The
battery charging spec runs VBUS at 1.5 amps, implying the pin
itself (regardless of voltage drop) will handle that current
flow.

On a laptop, there is a tendency to run 8 pin DIP "current flow
monitors". Some of these, on a USB2 port, would promptly cut
out at ~500mA, even though the desktop is a bit sloppy in
this regard. The real requirement on current flow policemen,
is that "nothing burns" and the silicon current flow monitor
on a laptop, should have the threshold set higher.

A laptop cannot (safely) use Polyfuses, because of the lack
of airflow over the PCB. The Polyfuse may be in a compromised
thermal situation, making it open prematurely perhaps. The
silicon fuse method in the 8 pin DIP, eliminates this problem.

The result is, there is a danger a laptop will have different
policing behaviors than a desktop. The Intel recommendations
to designers, is to "open up" the policing value to higher
than the "accounting" value. Subject to not burning any
pins or tracks.

*******

OK, so what of the "Accounting" side ? The USB devices have
a config space, where they "declare" their load. My mouse for
example, is 98mA. The resolution of the info field was 2mA,
so this is "1 bit shy of 100mA value". The mouse doesn't really
draw 98mA. In effect, the mouse is declaring "I am a unit load".

So the "Accounting" is also loaded with baloney.

The machine should be working with the accounting numbers,
so at least a device declaring a huge loading, would be
rejected during negotiation time. A USB3 device declaring
"I am a rat ******* drawing 900mA", then plugged into a
USB2 port, you'd think that would be rejected from Accounting
principles.

However, I don't know how that part really works.
You'll have to use your imagination. While the bus
may have been expanded to higher current flow and
more unit loads, we don't know what the top negotiated
value is allowed on USB3. It must be specified somewhere.
You'd think for backward compatibility, there'd be a
common limit. How is a user to know their USB3 device
won't work if plugged into a USB2 port, because
of an "accounting failure" ? That would be bad design.

*******

OK, so what do we know.

1) There is a possibility of power being switched on
USB ports. I don't know how that's happening, as
some schematics have no series pass devices in the path.
Perhaps the device is told to go quiescent and that's
why its LED goes off.

2) Hardware has a "policeman" which is there to prevent
copper tracks from burning or USB pins from burning.
The USB pins themselves seem to be supporting rather
high values, like 1.5A. The fuse is set to at least
protect the pins, and the "F200" I found on my sample
USB card implies the connector pins are actually 2 amps.

3) The OS and software use "accounting principles".
No idea how the accounting for USB3 declarations,
changes when plugged to a USB2 port. The port certainly
"knows" which protocol standard was negotiated, so it
has the opportunity to reject excessive "declared" loads.
The devices won't really know they're at 4V due to
cable voltage drop, from the original 5V source.
USB peripherals don't generally have an ADC to detect
"cheesy cabling" with 28ga wire.

Sorry I can't give a simple answer.
You'll have to use some imagination I guess...

Summary: At the physical level, there is *plenty* of current.
But, will "accounting" switch something off logically ?
Dunno. Would a USB3 device be dumb enough to "declare"
it actually draws 900mA ?

Paul

T wrote:

Hi All,

Paul will know this one off the top of his head.

If I plug a USB 2 peripheral into a USB 3.0/3.1 port on my
computer, do I still get the USB 3.0/3.1 power? Or
do I get throttles back to 500 ma?

Many thanks,
-T

The signalling is backwards compatible. As for power, does the
unidentified USB2 device draw more than 500mA even when plugged into a
USB3 (900mA) port? A properly designed USB2 device should not attempt
to draw more power from the +5V VBUS line than the 500mA load from a
USB2 port; else, that USB2 device could burn out a USB2 port.

Whether more power will be drawn from the port is device dependent. You
didn't identify the device. I remember when external USB2 HDDs had a
Y-adapter to plug the drive into two USB2 ports. The power from one
USB2 port was sufficient after the disk was spinning but the surge
current on starting the disk exceeded what one USB2 port could supply,
so the Y-adapter gave the drive the power from two USB2 ports. As I
recall, the connection order was to connect the 2 USB2 cable ends to the
2 USB ports before connecting the merged USB2 cable end to the drive.
The other way around would have the drive trying to suck more power from
a USB2 port than it could handle before you managed to get the 2nd USB2
cable end plugged in. Thankfully those went away. A USB device should
NOT be drawing more power than the type of USB port to which it was
intended to connect.

A USB3 port has the /potential/ of charging faster a USB2 device but not
if the USB2 device throttles its load to a USB2 level for safety.
Getting a new computer to have USB3 ports does not mean your USB2 device
will charge or suck any more power than from a USB2 port. You need the
specs on the device to see what is its maximumm input current load. I
can change the capacity of a power supply to adds gobs of more power
potential but (at the same voltage) a resistor is still going to have
the same current draw. Not all USB device no matter what their USB spec
level will adapt to the USB port type to draw more or less power to
match the port type.

Power delivery capacity by the port does not necessitate increased power
draw by the device. You'll have to check what the device will draw.

On 12/4/18 6:32 AM, Paul wrote:
T wrote:
Hi All,

Paul will know this one off the top of his head.

If I plug a USB 2 peripheral into a USB 3.0/3.1 port on my
computer, do I still get the USB 3.0/3.1 power?Â* Or
do I get throttles back to 500 ma?

Many thanks,
-T

Well, it's a mystery for a standard USB3 connection.
There is still only one VBUS pin. VBUS would be the
limiting factor, as there are additional ground contacts.

There is a PD spec, but I don't even want to go there...
That's a huge spec, if I could find it.

PinÂ*Â*Â*Â* Signal name

1Â*Â*Â*Â*Â*Â* VBUS
2Â*Â*Â*Â* D-
3Â*Â*Â*Â* D+
4Â*Â*Â*Â* GND

5Â*Â*Â*Â* StdA_SSRX-
6Â*Â*Â*Â* StdA_SSRX+
7Â*Â*Â*Â* GND_DRAIN
8Â*Â*Â*Â* StdA_SSTX-
9Â*Â*Â*Â* StdA_SSTX+

ShellÂ*Â*Â*Â* Shield (GND)

*******

There's a whiff of a hint here.

https://www.makeuseof.com/tag/usb-30...ogy-explained/

Â* "Power

Â*Â* The unit load has been increased to 150 mA
Â*Â* and a configured device can draw up to six unit loads,
Â*Â* which adds up to 900 mA.Â*Â* === OK, so a load of baloney so far

Â*Â* This exceeds USB 2.0 by 80% and leads to faster recharging
Â*Â* or powering of more than four devices from a single hub
Â*Â* [i.e. six unit loads or six keyboards or six mice as unit loads]

Â*Â* In addition, the minimum device operating voltage was dropped
Â*Â* from 4.4 V to 4 V, which saves energy. === Aha!
Â* "

So the goal post that was moved, was cable drop. By making the
peripheral IC more tolerant of voltage drop in the cable, they
could arrange for a higher current draw spec.

*******

On a computer, there is "accounting" and in hardware there
are "policemen".

On a desktop, we see "policemen" like this:

Â*Â* 1) A *pair* of USB2 ports is protected by a Polyfuse
Â*Â*Â*Â*Â* marked "110" or 1.1 amps. If both devices drew 500mA,
Â*Â*Â*Â*Â* the Polyfuse would be getting a bit soft. If only one
Â*Â*Â*Â*Â* USB2 port is being used, then the desktop can (almost)
Â*Â*Â*Â*Â* power a 2.5" hard drive (which draws 1.0 amps at startup).
Â*Â*Â*Â*Â* It depends on the details of the hard drive itself,
Â*Â*Â*Â*Â* as to whether the Polyfuse is going to get too soft,
Â*Â*Â*Â*Â* before spinup finishes.

Â*Â* 2) On a USB3 card, a *single* port is protected
Â*Â*Â*Â*Â* with an "F200" Polyfuse. That tells us the policeman
Â*Â*Â*Â*Â* on the desktop allows 2000mA versus the claimed 900mA.

These imply something about the ampacity of the USB pins. The
battery charging spec runs VBUS at 1.5 amps, implying the pin
itself (regardless of voltage drop) will handle that current
flow.

On a laptop, there is a tendency to run 8 pin DIP "current flow
monitors". Some of these, on a USB2 port, would promptly cut
out at ~500mA, even though the desktop is a bit sloppy in
this regard. The real requirement on current flow policemen,
is that "nothing burns" and the silicon current flow monitor
on a laptop, should have the threshold set higher.

A laptop cannot (safely) use Polyfuses, because of the lack
of airflow over the PCB. The Polyfuse may be in a compromised
thermal situation, making it open prematurely perhaps. The
silicon fuse method in the 8 pin DIP, eliminates this problem.

The result is, there is a danger a laptop will have different
policing behaviors than a desktop. The Intel recommendations
to designers, is to "open up" the policing value to higher
than the "accounting" value. Subject to not burning any
pins or tracks.

*******

OK, so what of the "Accounting" side ? The USB devices have
a config space, where they "declare" their load. My mouse for
example, is 98mA. The resolution of the info field was 2mA,
so this is "1 bit shy of 100mA value". The mouse doesn't really
draw 98mA. In effect, the mouse is declaring "I am a unit load".

So the "Accounting" is also loaded with baloney.

The machine should be working with the accounting numbers,
so at least a device declaring a huge loading, would be
rejected during negotiation time. A USB3 device declaring
"I am a rat ******* drawing 900mA", then plugged into a
USB2 port, you'd think that would be rejected from Accounting
principles.

However, I don't know how that part really works.
You'll have to use your imagination. While the bus
may have been expanded to higher current flow and
more unit loads, we don't know what the top negotiated
value is allowed on USB3. It must be specified somewhere.
You'd think for backward compatibility, there'd be a
common limit. How is a user to know their USB3 device
won't work if plugged into a USB2 port, because
of an "accounting failure" ? That would be bad design.

*******

OK, so what do we know.

1) There is a possibility of power being switched on
Â*Â* USB ports. I don't know how that's happening, as
Â*Â* some schematics have no series pass devices in the path.
Â*Â* Perhaps the device is told to go quiescent and that's
Â*Â* why its LED goes off.

2) Hardware has a "policeman" which is there to prevent
Â*Â* copper tracks from burning or USB pins from burning.
Â*Â* The USB pins themselves seem to be supporting rather
Â*Â* high values, like 1.5A. The fuse is set to at least
Â*Â* protect the pins, and the "F200" I found on my sample
Â*Â* USB card implies the connector pins are actually 2 amps.

3) The OS and software use "accounting principles".
Â*Â* No idea how the accounting for USB3 declarations,
Â*Â* changes when plugged to a USB2 port. The port certainly
Â*Â* "knows" which protocol standard was negotiated, so it
Â*Â* has the opportunity to reject excessive "declared" loads.
Â*Â* The devices won't really know they're at 4V due to
Â*Â* cable voltage drop, from the original 5V source.
Â*Â* USB peripherals don't generally have an ADC to detect
Â*Â* "cheesy cabling" with 28ga wire.

Sorry I can't give a simple answer.
You'll have to use some imagination I guess...

Summary: At the physical level, there is *plenty* of current.
Â*Â*Â*Â*Â*Â*Â*Â* But, will "accounting" switch something off logically ?
Â*Â*Â*Â*Â*Â*Â*Â* Dunno. Would a USB3 device be dumb enough to "declare"
Â*Â*Â*Â*Â*Â*Â*Â* it actually draws 900mA ?

Â*Â* Paul

Thank you!

On 12/4/18 9:03 AM, VanguardLH wrote:
A properly designed USB2 device should not attempt
to draw more power from the +5V VBUS line than the 500mA load from a
USB2 port; else, that USB2 device could burn out a USB2 port.

"Properly Designed" is the issue. And a "Properly Designed"
USB 2 port would log you out if you drew too much current.
Unfortunately, "I WANT IT CHEAP" often overrides "Properly
Designed". I have seen too many USB 2 port burn out. Apple for
one, states in their iPOD manual not to plug its charger
into a computer.

In article , wrote:

A properly designed USB2 device should not attempt
to draw more power from the +5V VBUS line than the 500mA load from a
USB2 port; else, that USB2 device could burn out a USB2 port.

"Properly Designed" is the issue. And a "Properly Designed"
USB 2 port would log you out if you drew too much current.

no, a properly designed port will either source only its maximum
current (with the device possibly not working properly or charging at a
slower rate) or shuts down the port, possibly triggering an alert to
inform the user. it will not log anyone out.

Unfortunately, "I WANT IT CHEAP" often overrides "Properly
Designed". I have seen too many USB 2 port burn out.

yep. there's a lot of crap out there.

Apple for
one, states in their iPOD manual not to plug its charger
into a computer.

that's not possible since the charger and computer both have usb-a
ports, with usb cables having a usb-b plug on one end.

however, with usb-c, that no longer matters. you *can* connect a usb-c
charger to a computer, which will power and recharge the laptop, except
for hp, which requires hp usb power adapters.

On 12/4/18 10:00 AM, nospam wrote:

that's not possible since the charger and computer both have usb-a
ports, with usb cables having a usb-b plug on one end.

however, with usb-c, that no longer matters. you *can* connect a usb-c
charger to a computer, which will power and recharge the laptop, except
for hp, which requires hp usb power adapters.


Way older model than that. Neet to hear they are using
usb-c ports now.

In article , wrote:


that's not possible since the charger and computer both have usb-a
ports, with usb cables having a usb-b plug on one end.

however, with usb-c, that no longer matters. you *can* connect a usb-c
charger to a computer, which will power and recharge the laptop, except
for hp, which requires hp usb power adapters.

Way older model than that.

the model does not matter.

a usb charger *cannot* be plugged into a computer unless you use a
non-standard a-a cable. the reason why usb cables have different plugs
at either end is so you *can't* do that.

if you meant plugging the ipod into the computer to charge, that works
perfectly fine and fully supported by every ipod that supports usb
charging.

Neet to hear they are using
usb-c ports now.

if only the rest of the world would...

nospam wrote:

you*can* connect a usb-c
charger to a computer, which will power and recharge the laptop, except
for hp, which requires hp usb power adapters.

Similar for Huawei.

In fact I have two USB-C phones, one laptop, one tablet, each charges
with their own charger, but mix them up and it's less than a 50:50
chance, though as far as I can remember, they all charge with a 3rd
party charger.

In article , Andy Burns
wrote:


you*can* connect a usb-c
charger to a computer, which will power and recharge the laptop, except
for hp, which requires hp usb power adapters.

Similar for Huawei.

In fact I have two USB-C phones, one laptop, one tablet, each charges
with their own charger, but mix them up and it's less than a 50:50
chance, though as far as I can remember, they all charge with a 3rd
party charger.

https://images.techhive.com/images/a...laptop_chartin
g-100649896-orig.png

full article:
https://www.pcworld.com/article/3017...arging-univers
al-or-bust-we-plug-in-every-device-we-have-to-chase-the-dream.html

and then there's the issue with usb-c cables that can fry stuff:
https://bensonapprovedcom.wordpress.com

On 12/4/18 10:12 AM, nospam wrote:
if you meant plugging the ipod into the computer to charge, that works
perfectly fine and fully supported by every ipod that supports usb
charging.

That is what I was referring to. And iPOD not iPAD. The very first
model's manual told you not to charge it from a computer. All the
world loves fast charge and does not want to hear about the "Kaboom" factor.

I had an accountant how liked to go jogging to work off stress
during the work day (he worked for the customer that everyone
else kicked out the door. His job was VERY stressful). He
liked to listen to music when he jogged so his family bought
him an iPOD for his birthday. And of course he did not read
the manual. And his laptop had all his accounting data on it
and was the "Cheap" model (accountants). He burned out all his
USB port. Wanting to save money by not calling me, he sent
it in for repair. (I remove the hard drives when a customer
does this.) As no one repairs laptops anymore, he got himself
a brand new shiny laptop. He never was able to track down
his old hard drive. It was a total disaster.

It is great to hear Apple has fixed the issue now!

And we able back to "properly designed". His laptop should have
logged out the port, not burned out the motherboard. The iPOD
should not have been designed with a USB port if it was going to
do Fast Charge. And the customer should have asked me what
to do. Triple whammy. Stepping over dollars to save dimes.



Neet to hear they are using
usb-c ports now.

if only the rest of the world would...

1+.

I hated USB 1. Tolerated USB 2. Liked USB 3.0. I ADORE USB-C/3.1.

In article , wrote:

if you meant plugging the ipod into the computer to charge, that works
perfectly fine and fully supported by every ipod that supports usb
charging.

That is what I was referring to.

except you said connect charger to computer.

And iPOD not iPAD.

that doesn't matter. they're all usb devices, other than the first two
generations of the ipod, which were firewire and could not connect via
usb (which at the time was too slow).

The very first
model's manual told you not to charge it from a computer.

no it didn't.

the very first (and second) ipod model were firewire and definitely
charged from a computer without any issue. firewire could source a
*lot* of power, capable of bus-powering a 3.5" hard drive or multiple
2.5" hard drives.

the 3rd gen ipod introduced the 30 pin dock connector which supported
both usb and firewire, depending on which cable was connected. however,
it could only sync via usb (not charge), so a dual-connector hybrid
cable was included along with a firewire power adapter, with usb for
syncing and firewire for charging. it was a bit of a kludge.

the 4th gen ipod could both sync and charge from usb without needing a
hybrid cable, as could all versions of the ipod mini, nano and shuffle.
in fact, the first ipod shuffle looked like a usb stick.

All the
world loves fast charge and does not want to hear about the "Kaboom" factor.

there is no kaboom factor.

worst case, it will charge slowly.

I had an accountant how liked to go jogging to work off stress
during the work day (he worked for the customer that everyone
else kicked out the door. His job was VERY stressful). He
liked to listen to music when he jogged so his family bought
him an iPOD for his birthday. And of course he did not read
the manual. And his laptop had all his accounting data on it
and was the "Cheap" model (accountants). He burned out all his
USB port.

that wasn't due to the ipod, and quite likely because the laptop was
'cheap'.

if it was the ipod, there would have been widespread usb port failures,
and there was not.

Wanting to save money by not calling me, he sent
it in for repair. (I remove the hard drives when a customer
does this.) As no one repairs laptops anymore, he got himself
a brand new shiny laptop. He never was able to track down
his old hard drive. It was a total disaster.

restore from a backup. he did have a backup, didn't he?

if he did not, then *that* is the real problem, not the usb ports.

It is great to hear Apple has fixed the issue now!

it was never broken.

And we able back to "properly designed". His laptop should have
logged out the port, not burned out the motherboard.

usb ports aren't 'logged out' and it's not clear what caused the
failure.

The iPOD
should not have been designed with a USB port if it was going to
do Fast Charge.

it doesn't do a fast charge, however, it does support 500ma if the
host supports it.

And the customer should have asked me what
to do. Triple whammy. Stepping over dollars to save dimes.

nope.

On 12/5/18 9:17 AM, nospam wrote:
In article , wrote:

if you meant plugging the ipod into the computer to charge, that works
perfectly fine and fully supported by every ipod that supports usb
charging.

That is what I was referring to.

except you said connect charger to computer.

And iPOD not iPAD.

that doesn't matter. they're all usb devices, other than the first two
generations of the ipod, which were firewire and could not connect via
usb (which at the time was too slow).

The very first
model's manual told you not to charge it from a computer.

no it didn't.

the very first (and second) ipod model were firewire and definitely
charged from a computer without any issue. firewire could source a
*lot* of power, capable of bus-powering a 3.5" hard drive or multiple
2.5" hard drives.

the 3rd gen ipod introduced the 30 pin dock connector which supported
both usb and firewire, depending on which cable was connected. however,
it could only sync via usb (not charge), so a dual-connector hybrid
cable was included along with a firewire power adapter, with usb for
syncing and firewire for charging. it was a bit of a kludge.

the 4th gen ipod could both sync and charge from usb without needing a
hybrid cable, as could all versions of the ipod mini, nano and shuffle.
in fact, the first ipod shuffle looked like a usb stick.

All the
world loves fast charge and does not want to hear about the "Kaboom" factor.

there is no kaboom factor.

worst case, it will charge slowly.

I had an accountant how liked to go jogging to work off stress
during the work day (he worked for the customer that everyone
else kicked out the door. His job was VERY stressful). He
liked to listen to music when he jogged so his family bought
him an iPOD for his birthday. And of course he did not read
the manual. And his laptop had all his accounting data on it
and was the "Cheap" model (accountants). He burned out all his
USB port.

that wasn't due to the ipod, and quite likely because the laptop was
'cheap'.

if it was the ipod, there would have been widespread usb port failures,
and there was not.

Wanting to save money by not calling me, he sent
it in for repair. (I remove the hard drives when a customer
does this.) As no one repairs laptops anymore, he got himself
a brand new shiny laptop. He never was able to track down
his old hard drive. It was a total disaster.

restore from a backup. he did have a backup, didn't he?

if he did not, then *that* is the real problem, not the usb ports.

It is great to hear Apple has fixed the issue now!

it was never broken.

And we able back to "properly designed". His laptop should have
logged out the port, not burned out the motherboard.

usb ports aren't 'logged out' and it's not clear what caused the
failure.

The iPOD
should not have been designed with a USB port if it was going to
do Fast Charge.

it doesn't do a fast charge, however, it does support 500ma if the
host supports it.

And the customer should have asked me what
to do. Triple whammy. Stepping over dollars to save dimes.

nope.


Oh brother. I showed him in his manual.

In article , wrote:


Oh brother. I showed him in his manual.

post a link to the manual where it said not to charge an ipod from a
computer.

https://support.apple.com/en_US/manuals/ipod

On 12/5/18 9:37 AM, nospam wrote:
In article , wrote:


Oh brother. I showed him in his manual.

post a link to the manual where it said not to charge an ipod from a
computer.

https://support.apple.com/en_US/manuals/ipod


It was years ago. Where in the hell am I going to find
old manuals. You either believe me or you do not.