Invalid wrote:
> In message <1189502720.449588.61050.DeleteThis@19g2000hsx.googlegroups.com>, dazzle
> <dazzle.DeleteThis@edazzle.net> writes
>> On 10 Sep, 17:15, Invalid <n....DeleteThis@wisty.plus.com.invalid> wrote:
>>> In message <1189432099.491399.59....DeleteThis@22g2000hsm.googlegroups.com>, dazzle
>>> <daz....DeleteThis@edazzle.net> writes
>>>
>>>
>>>
>>> >I have a GA-7VA Rev. 1.0 mobo with the F12 Bios and I have USB 2.0
>>> >problems.
>>>
>>> >It only seems to affect my wireless network adapter, currently a
>>> >Netgear W121T, after the adapter has been plugged in for a while,
>>> >almost all day then the USB socket it is plugged into stops working.
>>> >When I restart Windows it just hangs and I have to do a reset then it
>>> >takes a while for the USB to come back up and detect things are
>>> >plugged into it. It doesn't matter which slot I have the Netgear
>>> >plugged into the USB slot stops working - even if I plug another USB
>>> >device into it, nothing happens.
>>>
>>> >The last time this happened I disabled USB 2 in the Bios, then no USB
>>> >worked at all. Windows picked up the USB slots but nothing happened
>>> >when I plugged a USB device in. I re-enabled USB 2 in the Bios and
>>> >everything is working fine again.
>>>
>>> >Has anyone else had problems with this motherboard and USB 2, with or
>>> >without wireless network adapters.
>>>
>>> >BTW I am using Windows Vista Home Basic.
>>>
>>> >Any idea, help appreciated.
>>>
>>> I can't help explicitly, as I don't have that board, but it sounds like
>>> the Netgear adapter is close to the limit on power use from the USB
>>> Port.
>>>
>>> The USB ports have current limiting protection built in. If the Netgear
>>> exceeds (or gets very close to) the allowable current drain (a signal
>>> transient might push the demand over the limit, or as it happens after a
>>> while it might simply be thermal changes that trip the protection) the
>>> port shuts down to protect itself.
>>>
>>> I don't know how the protection system works in your case, but it sounds
>>> as if it takes time to recover after an overload - hence nothing works.
>>>
>>> Does the USB Port come back if you shut down, power off and restart the
>>> computer? If you plug the Netgear adapter into a different port does it
>>> work?
>>>
>>> I suspect you either need an external power source for the adapter (if
>>> it can use one) or a decent powered hub that will provide enough power
>>> for the adapter
>>>
>>> --
>>> Peter R Cook- Hide quoted text -
>>>
>>> - Show quoted text -
>>
>> If I plug another USB device into the same port as the wireless then
>> it doesn't work. If I plug the wireless into another USB port then it
>> does work. Usually after a shutdown/reboot the USB port is fine. It
>> only seems to affect the wireless adapter. I had a similar problem
>> with a Belkin wireless USB adapter as well.
>
> OK that confirms its definitely the port current limit being exceeded.
>>
>> The USB ports have 500mA, each, power available to them, but the
>> wireless adapter reports as using 0mA. My USB keyboard uses 100mA.
>
> This looks peculiar, I don't know how the current is measured. It may
> simply be something that the device itself reports when asked. Most of
> the USB dongles I can find on the web are in the High 400's (e.g. 460ma,
> 480ma). The protective devices appear to have some form of thermal
> mechanism (see
> http://www.usb.org/developers/whitepapers/power_delivery_motherboards.pdf
> ).
>
> It could be that over time that part of the MB is getting hot - you
> could try improving the cooling
>>
>> I'll look into the self-powered USB hub as I hadn't thought of that.
>> Thanks for the help.
>>
> Pleasure
If the wireless USB device is plugging directly into the motherboard, you can
use Microsoft UVCView.
http://www.microsoft.com/whdc/device/stream/vidcap/UVCViewdwn.mspx
Using UVCView, you get to look at the configuration data the USB device is
offering the system. For example, when I connect a USB mouse, there is a
data field with the recommended current consumption. My mouse reports it
expects to use 98mA. That is not a measured value. The data field has a
resolution of 2mA, so the allowed values will always be an even number.
The manufacturer selected 98mA strategically, so they could be under the
100mA limit for usage on bus powered hubs.
The max that field could contain, should be 500mA. No manufacturer can
afford to be honest, and claim their recommended current is higher than
the known spec limit. So a clever maker, could set the field to 498mA.
In terms of enforcing the limit, there are a couple possibilities. The
software could add the configuration values together and decide a limit
is being exceeded. For example, if you had a passive (bus powered) USB
hub, and put a bunch of 500mA devices on it, there could be some consequences
then.
The motherboard has overcurrent limiting, but that is not necessarily to
strictly enforce the 500mA limit. That is to protect the motherboard from
damage. My motherboard has a Polyfuse, which is a thermally based fuse.
One fuse is used per two USB ports. Which means, if my USB mouse went crazy
and started drawing 1.0 amperes of current, the Polyfuse would not blow.
And yet, that port would then be 2x the spec limit for consumption. The
rating printed on the top of the fuse, is probably a bit higher than 1 amp,
so as to not police the exact spec limit for current consumption. That way,
two 498mA devices could run indefinitely, without popping the fuse.
It is possible there are measurement devices, that could measure the consumption
and shut a port down. There is an OC# terminal on some USB 2x5 headers, and one
USB doc I looked at, seemed to suggest that was hooked to the power rail, and
was simply noting when the power dies on the port. (For example, if there was a
Polyfuse on the USB front panel circuit board, and it opened.) A device that
digitally enforces current consumption, and sends a logic signal on the OC# pin,
could function as a policeman. But since desktop motherboards at least, are
pretty cost sensitive, you aren't likely to see something like that on a
desktop. Maybe something like that is popular on a laptop, where power policing
could be more important.
I'm only drawing these conclusions from looking at a limited set of
documents. I haven't read the entire 500 page spec, and I'm not sure
it actually tells you the philosophy behind power management in any case.
The ultimate reason for limiting the current, is to prevent the tiny contacts
on the USB connector, from being burned. Above that limit, the copper wires
in the motherboard, may also be limited as to how much current they can
carry, without frying. Current carrying capability is determined by track width,
and sometimes power tracks for things, aren't set that big. With four layer
circuit boards, there are limited routing resources for power.
Paul
>> Stay informed about: GA-7VA Rev. 1.0 F12 Bios and USB 2.0 
FAQ
Profile
Private Messages
Log in
