Odd Problem with Processor Power Management

[sechs]

I recently upgraded the motherboard on my main computer, which allows me to finally take advantage of the PowerNow feature on my Opterons. I have duly set the power scheme in Windows to "minimal power management," and now the AMD power monitor shows the processors going to the one and only lower p-state when the load is low enough.

The problem that I've run into is an odd one.

I use a Logitech G7 cordless mouse. I also have a Dell LCD monitor with a powered generic USB hub.

If I plug the mouse receiver into a USB port on the monitor, when the processors go to the lower p-state, the mouse gets all jumpy. If the processors stay in the lower power state long enough, the mouse stops working all together. Unplugging and replugging the receiver temporarily fixes the problem.

If I plug the mouse receiver into any of the USB ports on the motherboard or on my case from the motherboard header, there's no problem.

The only other USB devices that I have on this machine are the UPS, plugged into the motherboard, and the battery charger for the mouse, plugged into the monitor. Unplugging the battery charger does not make a difference.

Any thoughts on what's going on and how to fix it?

 

[Bozo]

Sounds like the low power state is lowering the power to the USB ports too. This coupled with the voltage drop through the cable running from the computer to the monitor could be causing the problem.
Posibly a shorter or better quality cable would help.
Maybe a BIOS setting to only lower the power to the CPU??

Bozo :joker:

 

[Mercutio]

The monitor's USB ports are affected by the power state of the monitor itself. I've found - to my dismay - that plugging my phone in to charge on a lot of monitors that have USB powers only works if the monitor is on and displaying something. Not in standby and not off. On a cell phone this is pretty easy to verify, since there's a handy power meter that is showing what's what.

I guess you could switch your display to the "blank screen" screesaver and turn off all power management for it, but then you'll still have to leave your monitor on in order for things to behave. You're probably just better off plugging your mouse in to one of the ports on your desktop.

 

[sechs]

Sure. If the monitor is asleep, then the USB ports are asleep. But I'm actually using the computer in my normal course of business, and this is happening. I'll be trying to use Firefox on the monitor, and the mouse movement will start to get whacked-out.

If it were a power problem, then I'd expect it to also happen on the motherboard ports. As I understand it, the hub is powered, so that's not in question. And if I put the receiver on the end of my ten foot extension cord, it works fine on the motherboard ports.

I also tried turning off power management for the root hub to which the monitor is connected. No dice.

 

[LunarMist]

The monitor's USB ports are affected by the power state of the monitor itself.

Why would a monitor have a USB port? I've never heard of that before.

 

[Mercutio]

Dell Ultrasharp displays all have USB ports. They're kind of handy for things like webcams, USB phones, thumb drives et al.

 

[sechs]

Can anybody find anything on the specs for USB hub on these Dell monitors?

I just want to confirm that this should be powered and USB 2.0.

 

[sechs]

Can anybody find anything on the specs for USB hub on these Dell monitors?

I just want to confirm that this should be powered and USB 2.0.

 

[Fushigi]

Can anybody find anything on the specs for USB hub on these Dell monitors?

I just want to confirm that this should be powered and USB 2.0.

From the manual of my 2407:

"This monitor supports High-Speed Certified USB 2.0 interface." and goes on to list the 480Mbps speed.

"The monitor's USB interface works only when the monitor is on or in power save mode, If you switch the monitor off and then on, attached peripherals may take a few seconds to resume normal functionality."

It doesn't explicitly mention that the USB hub is powered, but it seems to be implied.

 

[Fushigi]

I should also mention that while it says it is USB 2, I haven't done any performance testing. Vista, however, wouldn't set a thumb drive for ReadyBoost while in the 2407; the same thumb drive is ReadyBoost-able plugged directly in the PC. That could be my laptop docking station, though, as I don't recall testing it directly from the laptop to the 2407.

 

[sechs]

Thanks for the info.

Obviously, the mouse isn't USB 2.0, but I might suspect some funny business with Windows if it was a 1.1 hub.

The whole thing is just awfully perplexing.