Theory as to dying AAO's...

According to intels datasheets on the 200 series of Atom processors, the Diamondille chips have a C6 "very deep sleep state"... basically meaning that the chip all but shuts down.

What ive noticed with my One is it would go to the black screen/green power LED on a couple of conditions:
1. Whenever Windows XP would go into hibernation mode, backing up the sleep state theory
2. Whenever I tried to turn off the wifi card with the .3119 BIOS... which again is a lower power state

My guess is either 1: The .3114 BIOS is bugged implementing the C6 state, or 2: It doesnt know how to handle XP's hibernation or 3: A combination of these factors... however since its happened with the Linpus one two im leaning towards option 1.

I downgraded to .3109 and have disabled all hibernation and so far so good...

BTW i have an AAO 110, 1gig Kingston RAM in the slot for 1.5 total, and a 60gig toshiba HDD installed, running Windows XP Pro.

 

Hi there.

So far so good in here (One Linpus 8/512mb), but I'm thinking about modding mine (same One but in Blue) with 60Gb HDD and 1.5gb, but I dont know the exact HDD and RAM models I should buy.

Could you point me out what to buy?

 

Maybe it's a bios bug. Hmm. If I remember correctly mine died like this. While I was using Windows I decided that I should reinstall form USB and pressed the power button but then I realized it will hibernate(because of my settings) so I keep it pressing for 5 sec and turned off the AA1. After turnin it on I inserted the USB and pressed the F12 key. The screen went off and I never saw it runnign again until I flashed the bios. So if BIOS flashing helps what could be the problem then except the BIOS himself

PS: Don't use it in bed. I managed to overheat mine twice as the holes choke and I believe that's not good :mrgreen:

 

I sit mine on a book, to help the airflow!

 

Intel's revised specification for the Atom

http://download.intel.com/design/chipse ... 319536.pdf

gives a number of design couple of issues which might have a bearing on this. Their notes imply that OEMs can do a workround through the BIOS. My assumption is that Acer have not implemented this correctly.

For instance:

AAE34 Processor May Not Wake Up from an Inactive State When an
Enhanced Intel SpeedStep® Technology Transition is Pending
Problem: Due to this erratum, the processor may hang in rare scenarios when it is in an inactive
state and there is an Enhanced Intel SpeedStep Technology transition pending.
Implication: The processor may hang and be unable to resume execution. A processor reset will be
needed to restart processor execution. Intel has not observed this erratum with any
commercially available software.
Workaround: It is possible for the BIOS to contain a workaround to this erratum.

I assume that the type Z530 processor (1.6MHz version) is included in the Aspire One but the issues listed (over 30) apply to all versions of the Atom. There are a whole host of bugs listed and I have picked this one as it seems to be most relevant. However it is clear that the Atom is still a very buggy chip and there is still the possibility that the problem has not yet come to light.

If it is a processor bug then this would imply that this is a generic problem although different manufacturers (OEMs) may have implemented work arounds in different ways so it does not follow that all Netbooks using the Atom are at risk. However, there is probably a problem with the BIOS in the Aspire One (certainly v0.3114 but also possibly 0.3109).

This implies that ALL Aspire One's are at risk of this problem until Acer can introduce can produce a revised BIOS that can take care of this or Intel do a revision of the Atom that takes care of these problems (which I would think would be several months away).