Dumb Question About USB Standby Power

[edy]

Hi folks,

I have some PC's at work and noticed one day an employee left one of those "LED glowing" USB charging cables plugged into the computer, like this:



The strange thing was, it was glowing yet the computer was OFF. It was still plugged in to the mains though.

My question is, do the USB ports on a desktop/tower PC normally supposed to continue delivering some power to the USB ports when the system shut down completely? There doesn't appear to be enough current to power anything or charge anything, but may be enough current bleed to still keep the tiny LED lights powered up on the glowing charging cable.

Is there anything in the specs regarding standby current on USB ports when the PC is shut down (not sleep mode but completely shut down)? Thanks.

- Edy

[wblock]

Newer motherboards have BIOS/UEFI options to keep USB power supplied even when the computer is off.  I'm not sure how that is set up with the computer power supply.  5V standby is not enough current, but do they leave the entire power supply on?

[Chalcogenide]

Even my older (first gen i3) Acer laptop did the same when in standby, even if it did not specifically have any BIOS option to keep the USB ports powered: it just did. I think that during the standby the 5V is kept ON because you may wake the PC by moving a mouse or by typing on a keyboard.

[Monkeh]

Yes, +5VSB is enough power. Modern power supplies allow 10-15W normally.

[edy]

My computer (mini-tower ATX chassis) is SHUTDOWN completely. It has a soft-switch on the front panel which is getting some current so it can sense when I press the button, but otherwise I would expect all ports to have no power when the main supply is off.   

[tszaboo]

It is because no-one bothers following the USB specifications. You cannot light an LED when it is in standby, that power is out of the limits (1mA? something like that). The device should take care of that.

[wraper]

My computer (mini-tower ATX chassis) is SHUTDOWN completely. It has a soft-switch on the front panel which is getting some current so it can sense when I press the button, but otherwise I would expect all ports to have no

ATX power supply is never completely OFF. Standby 5V is always there, no matter what. My current power supply can output up to 3A over 5Vsb, previous one was rated for 6A.

power when the main supply is off.   

Why should you expect the ports being OFF? In most cases they are not OFF at all. Many motherboards can even emulate phone charger while being off, so the phone can be charged at the max speed.
http://www.gigabyte.com/MicroSite/185/on-off-charge.htm

[edy]

It is an old ATX chassis, circa 2009 or so. I have a picture of it attached as I am not in the office at the moment so I can't tell you the brand/model.

[fubar.gr]

ATX power supplies have a small, mostly independent, power supply for the 5VSB. The 5VSB rail is always on as long as the PSU is plugged in and the switch at the back (if there is one) is switched on.

And since 5VSB comes from essentially a separate power supply, it is not uncommon for otherwise dead PSUs to still be able to supply 5 volts on the purple wire.

Now, on most motherboards you can select (via a jumper) the voltage source of the front panel USB ports. Either the main power supply 5 Volt rail, or the 5VSB rail.

This arrangement isn't new at all. I remember my 2003 computer had that option too

[rrinker]

 My MB even has an option to output more than standard USB power on a few of the ports so you can get full charging rate for a smartphone - you just have to load their special driver and the designated ports will provide something like 1.1 amps. I think the driver even has a measuring ability - not sure, I only tried it once just to see it work, since I have a charge cable hooked up for my devices anyway right on my desk and with the advent of wifi sync I never plug them in to a computer.

[gslick]

This is a brief subset of what the Universal Serial Bus Specification Revision 2.0 has to say about suspend current from a USB device point of view:

http://www.usb.org/developers/docs/usb20_docs/

7.2.3 Power Control During Suspend/Resume

Suspend current is a function of unit load allocation. All USB devices initially default to low-power. Low-power devices or high-power devices operating at low-power are limited to 500 uA of suspend current. If the device is configured for high-power and enabled as a remote wakeup source, it may draw up to 2.5 mA during suspend. When computing suspend current, the current from VBUS through the bus pull-up and pull-down resistors must be included. Configured bus-powered hubs may also consume a maximum of 2.5 mA, with 500 uA allocated to each available external port and the remainder available to the hub and its internal functions. If a hub is not configured, it is operating as a low-power device and must limit its suspend current to 500 uA.

While in the Suspend state, a device may briefly draw more than the average current. The amplitude of the current spike cannot exceed the device power allocation 100 mA (or 500 mA). A maximum of 1.0 second is allowed for an averaging interval. The average current cannot exceed the average suspend current limit (ICCSH or ICCSL, see Table 7-7) during any 1.0-second interval (TSUSAVG1). The profile of the current spike is restricted so the transient response of the power supply (which may be an efficient, low-capacity, trickle power supply) is not overwhelmed. The rising edge of the current spike must be no more than 100 mA/us. Downstream facing ports must be able to absorb the 500 mA peak current spike and meet the voltage droop requirements defined for inrush current during dynamic attach (see Section 7.2.4.1). Figure 7-48 illustrates a typical example profile for an averaging interval. If the supply to the pull-up resistor on D+/D- is derived from VBUS, then the suspend current will never go to zero because the pull-up and pull-down resistors will always draw power.

[Jeroen3]

That is about proper standby behavior, when initialized.
When uninitialized, you have 100mA (1 unit). But no motherboard actually cares and the polyfuse or overcurrent protector is rated for 500.
Laptops are a bit more picky, especially since a lot of cheap laptops only use 1 root hub for all ports. I'm not sure if they still do that for usb 3.0.

Some motherboards have an software option to keep (some of) the USB's powered. This doesn't count for the data pins, unless they have some SuperCharger technology.
When USB devices want to charge, they ask for the power units. As they should. When they are assigned the power units, and the pc turns off, most gadgets are able to keep charging.
If the device actively monitors the USB state, which they should, they stop charging and won't continue until the operating system is awake again. Unless they use less than 1 power unit, which they are guaranteed when uninitialized.

USB devices shall limit the power they consume from VBUS to one unit load or less until configured.

The LED's in the cable are allowed to take 100mA. Since a cable won't enumerate as a device.
As you'd understand, charging your smartphone with 100mA is going to take a long time.