A 300-Watt USB plug?

[Yousuf Khan]

After Intel's nearly killed USB 3.0 in favour of its sciene project,
LightPeak, it looks like it's now going the other way with suggestions
for new features for USB 3.0+. It's considering making increasing amount
of power USB ports are able to deliver, by both increasing voltage and
amperage. This one seems like an immediatel useful science project.

Yousuf Khan

***

SemiAccurate :: Intel is considering some tweaks to the USB standard
"USB devices are going to need to be able to deliver more power, a lot
more power in fact and it looks like Intel wants to implement support
for up to 42V at 1.8A for a total peak power draw of 35W. However, this
appears to be just a first step with the presentation mentioning "future
expansion to 200-300W using new connectors". Now that might seem a
little bit excessive, but it would allow you to plug in just about
anything to a USB port and it would be able to either be powered by it
or charged by it. However, Intel also wants USB to be able to deliver
less than 5V, something that some devices require and can’t have today."
http://www.semiaccurate.com/2010/10/07/intel-considering-some-tweaks-usb-standard/

 

[Rod Speed]

After Intel's nearly killed USB 3.0 in favour of its sciene project,
LightPeak, it looks like it's now going the other way with suggestions
for new features for USB 3.0+. It's considering making increasing
amount of power USB ports are able to deliver, by both increasing
voltage and amperage. This one seems like an immediatel useful
science project.
Yousuf Khan

***

SemiAccurate :: Intel is considering some tweaks to the USB standard
"USB devices are going to need to be able to deliver more power, a lot
more power in fact and it looks like Intel wants to implement support
for up to 42V at 1.8A for a total peak power draw of 35W. However,
this appears to be just a first step with the presentation mentioning
"future expansion to 200-300W using new connectors". Now that might
seem a little bit excessive, but it would allow you to plug in just
about anything to a USB port and it would be able to either be
powered by it or charged by it.

Wont handle my microwave or my welder either.

Wont even handle the fan heater either.

 

[Intel]

Intel is considering some tweaks to the USB standard "USB devices
are going to need to be able to deliver more power, a lot
more power in fact. However, this appears to be just a first
step with the presentation mentioning "future expansion to 200-
300W using new connectors".

Finally - The USB-powered toaster has arrived!

What's next?

Maybe Intel wants me to be able to recharge my electric car from a USB
port?

 

[krw]

Finally - The USB-powered toaster has arrived!

What's next?

Maybe Intel wants me to be able to recharge my electric car from a USB
port?

On your laptop? Hey, you could plug the laptop into the cigarette lighter!

 

[Robert Redelmeier]

On your laptop? Hey, you could plug the laptop into the
cigarette lighter!

Ha.Ha More seriously, what V * A this is going to be
and how it will ever stay under the NEC low-voltage Class2
(30Vrms or 60Vdc, 100VA)?

Is Intel going to go with 60 Vdc @ 5A and try to get waivers
with GFIC balancing & fast overcurrent trips? That'll take
at least 2*18 AWG, not the nice flexible 24 AWG . Or are
they thinking something even nuttier, like 300Vdc @ 1A to
keep wiresize down?. What protections will convince NFPA & UL ?


-- Robert

 

[krw]

Ha.Ha More seriously, what V * A this is going to be
and how it will ever stay under the NEC low-voltage Class2
(30Vrms or 60Vdc, 100VA)?

Is Intel going to go with 60 Vdc @ 5A and try to get waivers
with GFIC balancing & fast overcurrent trips? That'll take
at least 2*18 AWG, not the nice flexible 24 AWG . Or are
they thinking something even nuttier, like 300Vdc @ 1A to
keep wiresize down?. What protections will convince NFPA & UL ?

I was wondering the same thing. ...and why would I need so much power. It's
*got* to be expensive. The whole idea of USB is that it's cheaper than dirt.

 

[Robert Redelmeier]

I was wondering the same thing. ...and why would I need
so much power. It's *got* to be expensive. The whole idea
of USB is that it's cheaper than dirt.

I can see why the power -- getting rid of not-dirt-cheap wallwarts.
How is a lot trickier. But I can see ways since actives
(semiconductor logic) are also cheaper than dirt. Maybe even
get around the hiV arc-flash.

-- Robert

 

[Tom Lake]

Wont handle my microwave or my welder either.

Wont even handle the fan heater either.

If my car's battery goes dead, maybe I can start my car with it?

Tom L

 

[daytripper]

I can see why the power -- getting rid of not-dirt-cheap wallwarts.
How is a lot trickier. But I can see ways since actives
(semiconductor logic) are also cheaper than dirt. Maybe even
get around the hiV arc-flash.

-- Robert

Actually, wallwarts *are* dirt cheap.

But, more to the point, when have you ever even heard of a "300 watt
wallwart"? Or anything even close to that capacity?

I'd guess 99.very_large_integer% of the wallwarts *ever sold* - at least in
the Reality I inhabit - are under 24 watts capacity...

/daytripper

 

[krw]

100VA is enough for that, IMO.

Actually, wallwarts *are* dirt cheap.

Cheap but not *dirt* cheap. We're paying about $5 each for 10W WWs in
five-thousand lots. Even putting a zero or two behind the quantity doesn't
get them down too much further.

But, more to the point, when have you ever even heard of a "300 watt
wallwart"? Or anything even close to that capacity?

Laser printer? ;-)

I'd guess 99.very_large_integer% of the wallwarts *ever sold* - at least in
the Reality I inhabit - are under 24 watts capacity...

Probably a lot less than that. Most go for things like mice and maybe USB
disk drives. My scanner has a big one, 36W.

Another problem I see is hubs. Are they all going to supply 300W? ...or does
the wiring (and planning) just get that much more complicated.

 

[Robert Redelmeier]

Actually, wallwarts *are* dirt cheap.

As Keith posts side-thread, not all that cheap, especially if you
are targetting a device which retails for $99. Then there is all
the packaging hassle and different versions required, especially
for the EU states. Not to mention RMA issues.

But, more to the point, when have you ever even heard of a
"300 watt wallwart"? Or anything even close to that capacity?

Granted no, but I'd include power bricks in "wallwarts", and many
of them are ~100W, admittedly for heavier devices like laptops.

I'd guess 99.very_large_integer% of the wallwarts *ever sold* -
at least in the Reality I inhabit - are under 24 watts capacity...

Agreed. A less ambitious goal for PoUSB would be more reasonable.
The current 500mA @ 5V is 'way light. It is rather silly that
2.5"HDs can be USB powered, but 3.5"HDs need an external source.
Why the stupid wallwart for speakers? 24W is quite reasonable.

The idea of powering a laser printer through the CPU is just
nuts. An inkjet, OK. I'm a little squeamish about centrally
powering the monitor even though my first machines worked this way.
But not a projector (that bulb needs cooling).

-- Robert

 

[krw]

As Keith posts side-thread, not all that cheap, especially if you
are targetting a device which retails for $99. Then there is all
the packaging hassle and different versions required, especially
for the EU states. Not to mention RMA issues.

Two points. Our (10W) wall warts have replaceable connectors for world-wide
use. That can be dealt with. Some devices are in the $50 range. 10% of
*retail* to the WW manufacturer is a *lot*.

Granted no, but I'd include power bricks in "wallwarts", and many
of them are ~100W, admittedly for heavier devices like laptops.

Laptops powered by USB... Now there is a picture. ;-) Who's the boss?

Agreed. A less ambitious goal for PoUSB would be more reasonable.
The current 500mA @ 5V is 'way light. It is rather silly that
2.5"HDs can be USB powered, but 3.5"HDs need an external source.
Why the stupid wallwart for speakers? 24W is quite reasonable.

Some 2.5" drives needed a PS, too. Or perhaps it's just that not all USB
ports really can supply .5A. Anyway, some of mine came with WWs.

The idea of powering a laser printer through the CPU is just
nuts. An inkjet, OK. I'm a little squeamish about centrally
powering the monitor even though my first machines worked this way.
But not a projector (that bulb needs cooling).

SMOP for the device driver. ;-)

 

[Percival P. Cassidy]

On 10/09/10 06:19 pm, Robert Redelmeier wrote:

The current 500mA @ 5V is 'way light. It is rather silly that
2.5"HDs can be USB powered, but 3.5"HDs need an external source.

Don't 3,5" HDs need 12v as well as 5v?

Perce

 

[Robert Redelmeier]

Don't 3,5" HDs need 12v as well as 5v?

Good point, but with enough current allowable on the 5v,
the 12v could be made by switching up. There are good
tech reasons, but it just exaggerates a category break.

-- Robert

 

[Arno]

On 10/09/10 06:19 pm, Robert Redelmeier wrote:

Don't 3,5" HDs need 12v as well as 5v?

They do and the reason is far heavier spindle and head assembly.

Arno