Hi,

It seems the specification allows USB devices to draw up to 100mA (for
a "low power" port) and 500mA (high power port).

Does anyone have any idea of the incidence of high power ports?
Especially on laptops.

For a design with USB control, I am trying to decide if it is worth
including an option to take all the power from USB, rather than a
wall-wart type adapter. It would be great just to have the one plug,
and also allow fully mobile operation. But if a large fraction of
laptops only have low power ports, it may be more trouble than it is
worth. (It occurs to me that in order to support plugging in an
external hub, all laptop ports *ought* to be high power).

Thanks,

-- 

John Devereux

On Aug 5, 7:29=EF=BF=BDam, John Devereux <jdREM...@THISdevereux.me.uk> wrot=
e:
> Hi,
>
> It seems the specification allows USB devices to draw up to 100mA (for
> a "low power" port) and 500mA (high power port).
>
> Does anyone have any idea of the incidence of high power ports?
> Especially on laptops.
>
> For a design with USB control, I am trying to decide if it is worth
> including an option to take all the power from USB, rather than a
> wall-wart type adapter. It would be great just to have the one plug,
> and also allow fully mobile operation. But if a large fraction of
> laptops only have low power ports, it may be more trouble than it is
> worth. (It occurs to me that in order to support plugging in an
> external hub, all laptop ports *ought* to be high power).
>
> Thanks,
>
> --
>
> John Devereux

This is from memory -- becuase everytime I think "USB", I have to go
get the books - which are at the office.

So, I'll just mention this:
There are hubs, ports and controllers.  I think controllers are all
high-power??  This is what would be on the laptop.  Hubs would be
downstream from this, and could be either - or can also be powered
hubs meaning they have a wall wart.   Ports are downstream from there
and are enumerated as either high or low power, standard, or high
speed, etc...  Depending on what's available to be delivered to them.

I might have the actual terminology wrong, but the jist is right.

But as to low-power controllers on laptops??  I don't know.  I
wouldn't think so.????
Someone here will know for sure.   Good luck with the project.

-mpm

mpm <m...@aol.com> writes:

> On Aug 5, 7:29�am, John Devereux <jdREM...@THISdevereux.me.uk> wrote:
>> Hi,
>>
>> It seems the specification allows USB devices to draw up to 100mA (for
>> a "low power" port) and 500mA (high power port).
>>
>> Does anyone have any idea of the incidence of high power ports?
>> Especially on laptops.
>>
>> For a design with USB control, I am trying to decide if it is worth
>> including an option to take all the power from USB, rather than a
>> wall-wart type adapter. It would be great just to have the one plug,
>> and also allow fully mobile operation. But if a large fraction of
>> laptops only have low power ports, it may be more trouble than it is
>> worth. (It occurs to me that in order to support plugging in an
>> external hub, all laptop ports *ought* to be high power).
>>
>> Thanks,
>>
>> --
>>
>> John Devereux
>
> This is from memory -- becuase everytime I think "USB", I have to go
> get the books - which are at the office.
>
> So, I'll just mention this:
> There are hubs, ports and controllers.  I think controllers are all
> high-power??  This is what would be on the laptop.

That is what I am hoping for... I can specify that the device must be
plugged directly into the laptop, or perhaps a hub if specifically
labelled "high-power". But I don't think a user is going to know if
his laptop port is high-power or not.


>  Hubs would be
> downstream from this, and could be either - or can also be powered
> hubs meaning they have a wall wart.   Ports are downstream from there
> and are enumerated as either high or low power, standard, or high
> speed, etc...  Depending on what's available to be delivered to them.
>
> I might have the actual terminology wrong, but the jist is right.
>
> But as to low-power controllers on laptops??  I don't know.  I
> wouldn't think so.????
> Someone here will know for sure.   Good luck with the project.

Thanks!

-- 

John Devereux

On Tue, 05 Aug 2008 12:29:45 +0100, John Devereux
<j...@THISdevereux.me.uk> wrote:

>
>Hi,
>
>It seems the specification allows USB devices to draw up to 100mA (for
>a "low power" port) and 500mA (high power port).
>
>Does anyone have any idea of the incidence of high power ports?
>Especially on laptops.
>
>For a design with USB control, I am trying to decide if it is worth
>including an option to take all the power from USB, rather than a
>wall-wart type adapter. It would be great just to have the one plug,
>and also allow fully mobile operation. But if a large fraction of
>laptops only have low power ports, it may be more trouble than it is
>worth. (It occurs to me that in order to support plugging in an
>external hub, all laptop ports *ought* to be high power).
>
>Thanks,
Key to remember is that the device is meant to _ask_ to have the extra
power. Technically it is not allowed just to draw it (though on many
machines this will be OK). So your device needs to be able to wake up,
and connect to the USB, without drawing more than 100mA. 
The host, then can allow or refuse the request.
One solution to more power, is to use two USB headers. This is
commonly done on some older USB powered hard drives, where the 'spin
up' current is borderline for the port.
You need to be fairly careful, when switching to using the full power,
that the inrush current is kept low. 

Best Wishes

Roger Hamlett <r...@ttelmah.demon.co.uk> writes:

> On Tue, 05 Aug 2008 12:29:45 +0100, John Devereux
> <j...@THISdevereux.me.uk> wrote:
>
>>
>>Hi,
>>
>>It seems the specification allows USB devices to draw up to 100mA (for
>>a "low power" port) and 500mA (high power port).
>>
>>Does anyone have any idea of the incidence of high power ports?
>>Especially on laptops.
>>
>>For a design with USB control, I am trying to decide if it is worth
>>including an option to take all the power from USB, rather than a
>>wall-wart type adapter. It would be great just to have the one plug,
>>and also allow fully mobile operation. But if a large fraction of
>>laptops only have low power ports, it may be more trouble than it is
>>worth. (It occurs to me that in order to support plugging in an
>>external hub, all laptop ports *ought* to be high power).
>>
>>Thanks,
> Key to remember is that the device is meant to _ask_ to have the extra
> power. Technically it is not allowed just to draw it (though on many
> machines this will be OK). So your device needs to be able to wake up,
> and connect to the USB, without drawing more than 100mA. 
> The host, then can allow or refuse the request.

I was looking at using the FT232R, which seems to be able to negotiate
this for me. <http://www.ftdichip.com/Products/FT232R.htm>;.

> One solution to more power, is to use two USB headers. This is
> commonly done on some older USB powered hard drives, where the 'spin
> up' current is borderline for the port.

No, that's nasty :)

> You need to be fairly careful, when switching to using the full power,
> that the inrush current is kept low. 

Thanks, this is mentioned in the FT232R datasheet too (and there is a
control pin and a nice little circuit to deal with it).

-- 

John Devereux

"John Devereux" <j...@THISdevereux.me.uk> wrote in message 
news:8...@cordelia.devereux.me.uk...
> Does anyone have any idea of the incidence of high power ports?
> Especially on laptops.

It's safe to say they're ubiquitous.  I've never seen a low-power (only) USB 
port that's integral to the laptop (as opposed to being on, e.g., a PCMCIA USB 
card or an unpowered hub), although I'm sure someone can point to an example 
where this was done.  (On the other hand, I've definitely seen a few laptops 
where they plus another proprietary connector next to a USB socket that 
provides even *more* power than USB alone can for, e.g., older DVD burners/big 
hard drivers -- said peripherals then use a "combination" plug that 
simultaneously contacts the USB socket and the proprietary power socket.)

> For a design with USB control, I am trying to decide if it is worth
> including an option to take all the power from USB, rather than a
> wall-wart type adapter.

I'd vote for, "yes, absolutely."  If you don't need more than 2.5W, it's 
almost criminal not to allow power to be drawn from USB alone. :-)

---Joel

"Roger Hamlett" <r...@ttelmah.demon.co.uk> wrote in message 
news:j...@4ax.com...
> Key to remember is that the device is meant to _ask_ to have the extra
> power.

Yes, although as I imagine we're all aware many peripherals don't bother and 
few USB controllers actively restrict current to less than 500mA... hence some 
devices cheat and do draw >100mA even before asking permission.

> One solution to more power, is to use two USB headers. This is
> commonly done on some older USB powered hard drives, where the 'spin
> up' current is borderline for the port.

I've *never* seen one of these that supports the full "ask first" protocol on 
*both* connectors. :-)

---Joel

"Joel Koltner" <z...@yahoo.com> writes:

> "John Devereux" <j...@THISdevereux.me.uk> wrote in message 
> news:8...@cordelia.devereux.me.uk...
>> Does anyone have any idea of the incidence of high power ports?
>> Especially on laptops.
>
> It's safe to say they're ubiquitous.  I've never seen a low-power (only) USB 
> port that's integral to the laptop (as opposed to being on, e.g., a PCMCIA USB 
> card or an unpowered hub), although I'm sure someone can point to an example 
> where this was done.  (On the other hand, I've definitely seen a few laptops 
> where they plus another proprietary connector next to a USB socket that 
> provides even *more* power than USB alone can for, e.g., older DVD burners/big 
> hard drivers -- said peripherals then use a "combination" plug that 
> simultaneously contacts the USB socket and the proprietary power socket.)
>
>> For a design with USB control, I am trying to decide if it is worth
>> including an option to take all the power from USB, rather than a
>> wall-wart type adapter.
>
> I'd vote for, "yes, absolutely."  If you don't need more than 2.5W, it's 
> almost criminal not to allow power to be drawn from USB alone. :-)

Fantastic, thanks Joel :)

-- 

John Devereux

In article <8...@cordelia.devereux.me.uk>, 
j...@THISdevereux.me.uk says...
> 
> Hi,
> 
> It seems the specification allows USB devices to draw up to 100mA (for
> a "low power" port) and 500mA (high power port).
> 
> Does anyone have any idea of the incidence of high power ports?
> Especially on laptops.

Out ThinkPads run external drives without wall warts, so I'm sure 
they're "high" power.  Don't have a spec that sez so though.

> For a design with USB control, I am trying to decide if it is worth
> including an option to take all the power from USB, rather than a
> wall-wart type adapter. It would be great just to have the one plug,
> and also allow fully mobile operation. But if a large fraction of
> laptops only have low power ports, it may be more trouble than it is
> worth. (It occurs to me that in order to support plugging in an
> external hub, all laptop ports *ought* to be high power).
> 
> Thanks,
> 
> 

-- 
Keith

On Aug 5, 9:11=A0pm, krw <k...@att.bizzzzzzzzzz> wrote:
> Out ThinkPads run external drives without wall warts, so I'm sure
> they're "high" power. =A0Don't have a spec that sez so though.


I don't know of any laptop drive that would actually power up with
only 100ma of +5V.  Most need more than half a watt just idling, and
typically need around 2W while doing actual I/O to the disk.