Netbook freezes randomly

Update on my repair service from Acer:

Just called them today, and they told me they are replacing my hard drive.
I am glad that they are replacing something new but I still have a feeling the HARD DRIVE is not the problem.

I guess I will have to wait again and see what happens next. It will take another 5-7 days to complete the repair as I was told.

Gosh, almost 2 weeks w/o a brand new netbook. Acer SUCKS. Sorry, but I didn't expect to go through this much HASSLE for
buying a BRAND NEW ACER netbook.

 

Concerning the "power-related-problem" theory: I don't think power consumption has anything to do with the problem. My 751 used to freeze (and on very rare occasions still does...) when using Microsoft Word 2007, hereby specifically when scrolling using the scrollbar; a situation that can hardly be considered to use a lot of power.

I'm still fairly certain that the problem is located within the chipset's memory-controller, which in combination with the graphics adapter seems to be the cause of the freezing (thus problems can occur whenever the graphics adapter uses a certain amount of Ram or when a lot of traffic is caused, e.g. by the webcam).

Have any of you tried disabling DEP (as I suggested a view pages earlier)?

 

Hi DrNo,

I have a few thoughts. I have looked at my copy of Office 2003 (because that is what I have) just to remind myself, but the scrolling behavior of Word 2003 is as I remember it. When you scroll with Word 2003, you will note a hard drive hit when scrolling from the top of the document to the bottom, but you will see a huge hard drive hit (at least in comparison to the file size you would expect you are working with) if you try to scroll from the bottom to the top. I have seen it in every version of Word and I can't image why they have this characteristic, but it is there. I would respectfully suggest that the scrolling that you are doing is having a much bigger impact on the hard drive and system resources than you might image and that it does not contradict what I have suggested.

I followed your work with the DEP with great interest and tracked it back to the original Intel documents and explanations. As I understand it, this feature is examining the destination or source of data and determining if that address is an appropriate place for that kind of data. If data seems to be going somewhere inappropriate, the system stops. By switching this feature off, you have enjoyed much longer times between freeze-ups, if in fact you see any at all these days. It seemed to me that if the HDD was running on fumes (power-wise) then it seemed reasonable that it will send garbage data every once in a while. If the DEP is running as designed, then some of the time it will flag this data as bad and stop the system. By switching off the DEP function you are allowing the system to pass the garbage data into the normal data stream, but note that if it is data and not program instructions, you may not even notice the slight corruption in a file. If it were an image field, it would surprise me if you noticed a single bad pixel in a single frame of information. Notice too that while the HDD delivered bad data, the data on the drive has not been corrupted, so the next access may show no damage whatsoever.

Finally, your comments on the US15W/GMA500 chip would be much more convincing to me if the Asus 1101HA, which uses the same chip, was also seeing the same system crashes that we are, but it doesn't seem that that is the case. Certainly their Forum doesn't suggest any such major issues. They use the same memory and the same HDD's we do. The heavy traffic use patterns that you describe as being more failure prone are consistent with a high power demand application (especially the camera example) and that is the specific time when I would predict a crash.

Regards

 

I am also in doubt of the power theory. If I open the webcam by going to My Computer - Webcam, I have not yet had a freeze (but I have not tested this for very long periods). If I use the webcam in Yahoo Messenger, it will freeze after a few seconds. After the DEP mod, it can now last something like 3o minutes. Using the internal 3G and Bluetooth does not seem to trigger a freeze. I am running BatteryBar to try to see a pattern powerdrain/freezing. Though I have just started doing this, it does not look like the freezing is related to the discharge rate.

EDIT: I let Avira do a scan which took close to 30 minutes. During this time I let WMP play music at full volume and I watched myself on the webcam that I open through My Computer. The internal 3G was also active. CPU jumping around 60 to 70% mostly, fan working hard, and screen at the brightest. Discharge rate was somewhere from 10 to 12 mW (I am on battery). During this time the cam did not hang a second.
When the scan was finished, I closed all applications and 3G, and opened webcam in Yahoo. Not wanting to wait maybe 30 minutes, I "provoke" by clicking on the statisics in the webcam window and the pc froze after a few seconds. Discharge rate 6 to 7 mW.

 

Congrats Cmeide,

It looks like you have found a very ill-behaved application.

It would be best to realize that I wasn't suggesting that every freeze-up incident is power related, clearly as you have demonstrated, there are apps that are just born-to-die. Nor was I saying that every 751 has a defective power distribution system and will inevitably freeze up. It would be better to say that there are machines that fail repeatedly for no obvious reason and that some of them probably have power related issues.

With regards your measurements of total input power dissipation, I am of the opinion that it is in the Output of the onboard power supply /regulators and the distribution of power through the wiring traces around the Printed Circuit board where problems arise. I am pretty confident that both the power pack in the line cord and the battery pack are supplying enough power to the onboard power supply/regulators to run the computer. It's what happens to the power as it goes through the onboard power supply/regulators and gets split up and routed around the PCB wiring where I think the corruption takes place. I am especially suspicious of the fan power leads and the hard drive leads since hard drives and fans seem to be the hardware that Acer replaces most often. Alternatively, it may be in fact the fan or the HDD that is actively malfunctioning and replacing them is exactly what is needed. Or it may simply be that the fan and HDD are the only things (other than the mobo) that they Can replace and you got to do something for the customer. Time will tell.

Good work with your experiment, it's just the kind of work that we need to nail down all of the issues that cause freeze-ups.

"In God we trust, all others bring data."

 

Good to see you guys making progress! My 751 doesn´t have any freeze issues, but i never use the webcam. I´d like to check this, if it helps you. Can you please post a little guide how to setup? Maybe in an extra thread here, i think lots of people are willing to help/test.

Greetings,
Felix

 

Each one their own atitude.

Anyway, to get back to facts. My 751 is like most others here with the freezing problem. It can be fine for days without any problems and it can have days when it will freeze several times an hour. I have only used it with XP and nearly every time it freezes, it will just freeze as it is. Only the LCD backlight on/off function will work. 2 times I have experienced a black screen.
Every time holding the power button a few seconds will make the computer reboot.
I have exchanged my HDD for a larger unit, but this has made no difference.

I have never experienced any pc before not capable of running Yahoo Messenger with webcam and I think it would we well known if that application had a major problem. I have tried 2 different versions on the 751, none of them being beta. Running the webcam in Yahoo, it will from anything ranging from 3 seconds to maybe 30 minutes, freeze.

Finding the power theory interesting, I try putting load on the HDD, fan, 3G, Wlan, speakers, and cam all at the same time to see if this will make the computer freeze. So far I have only tested a few hours but it has not made any difference.

I posted my inial results yesterday so that others could pick up the thread and try it out. Then after some time we can decide if there is a pattern to it or my 751 has a special behavior.

BatteryBar tells me excactly how much power each unit draws as I enable/disable them. I find this a very useful tool. It also tells me that the webcam draws very little power compared to many other units in the computer.

And as I tried to point out, the webcam will run just fine if used on its own. Its only in Yahoo (and for others in Skype) I experience any problems. This tells me there likely is a hardware conflict triggered by software somewhere.
I hope others can run some tests on this to see if it behaves the same.

Now I am very eager to hear others experience and opinions on this.

 

I just checked the XP - "Event Log" and found one interesting entry. Upon requesting more information about it, this came up:


----------------------------------------------------------------------------------------
Details
Product: Windows Operating System
ID: 1
Source: ACPIEC
Version: 5.2
Symbolic Name: ACPIEC_ERR_WATCHDOG
Message: %1: The embedded controller (EC) hardware didn't respond within the timeout period. This may indicate an error in the EC hardware or firmware, or possibly a poorly designed BIOS which accesses the EC in an unsafe manner. The EC driver will retry the failed transaction if possible.

Explanation
The embedded controller (EC) hardware did not respond within the time limit. The EC driver will retry the failed transaction. If the EC continues to fail, the operating system might stop responding, or “hang.”

Possible causes include:

An error in the EC hardware or firmware.
A poorly designed basic input output system (BIOS) that accesses the EC incorrectly.
----------------------------------------------------------------------------------------------------

So there are, as it seems, two possible outcomes:

1.) There is indeed a critical hardware malfunction (due to whatever) that randomly causes the system to freeze.

2.) The hardware is ok, but either the firmware or the BIOS (or both?) are, quote, "poorly designed".

Maybe those of you who know how to access the Event Log can check if yours also displays this specific error; the source is called "acpiec", type "error", event "1".

 

Hi guys,

I have 751h with freezing problem too. I never use webcam. I have all drivers and BIOS up-to date.
Netbook freezes from time to time, even when no programs opened (except backgrounded).
It never freeze in safe mode.
100% it will freeze while playing video (324p)
once it fall to the blue screen - I made a photo

May be it will help

 

I'm pissed!!! After being without my compute for 3 weeks, Acer sends it back to me already in a "frozen" state! Their note says they updated the BIOS and it "passed their checks". What the heck!?

I wasted 3 weeks and $20 in shipping just for them to reflash the BIOS, which everyone on this forum knows doesn't help!? They are useless!

 

You should've told the reps that updating the BIOS won't solve anything.
I was persistent in telling them I've tried everything and I told them it might be a hardware problem plus I added the webcam issue.

When I checked the status of my repair, they said they are replacing the hard drive.

You can send it back for repair again... insist that your netbook still has the same problem and they will shoulder the shipping costs.

I know it really sucks buying a NEW netbook which you can't even use. If not for the 11.6 screen, I wouldn't have bought an acer.

 

************************************************************************************************
Hi DrNo,

ACPI is a hybrid driver that is to be found both within Windows and another part of it in the BIOS. It handles monitoring and control of certain power functions including fan control, bus timers, traffic cop on the data bus and noting if something is showing up that is not expected. It appears (I am still tracking it down) not to be functional in SAFE mode. Since 2002 many of the major computer manufacturers have had incidents with ACPI that required a BIOS patch, this list includes Acer. With the exception of Windows XP SP2, that would send off a blizzard of event notices and noticeably slow the system, a problem that Microsoft did fix; it looks like the computer manufacturer problems correspond to changes of the cpu and bridge chip set (i.e, a major new BIOS creation). In other words exactly what we see since ours are the first example of the Poulsboro chip set that Acer has ever seen.

And for those who are interested, yes, an ACPI failure might well exhibit the same symptoms as a power system problem since the fan for sure and the powering of the hard drive (I think, I am still checking) are under the aegis of the ACPI. And to anticipate a question, if the US15W/GMA 500 chip was fundamentally defective it wouldn't be working in the Asus 1101HA.

 

@ Nick_in_Wash: Interesting theory you got there... this could also be the explanation for there being no real pattern: Insignificant power leaks may not be causing the actual problem, but if occurring simultaneously with a ( otherwise probably also negligible) failure in the ACPI, this could very likely lead to a severe malfunction. Having both technical failures at the same time may happen, but it does not necessarily have to happen. Like rolling two dice: There's a certain chance to get snake eyes (double 1; just as an example - the chance is at about 2,7%, as far as I remember), but you could also end up rolling a hundred times with a different outcome.

Long story short: There is a way to completely disable ACPI. I will give it a shot and conduct some long term testing during the next couple of days.

Regards.

DrNo

 

Notice that in my initial analysis I gave the drivers a pass because I assumed that drivers that Acer used would be the same drivers that Asus used. I forgot that the ACPI is a large driver that Acer would have to carve from scratch and that as such if they goofed it, all of the problems that we see are explainable just on that basis. ACPI is a Power Controller module, If it malfunctions anything it touches is likely to look like a power system failure even it it isn't.

It is like the story of the man who fall down hard and then sits up and examines himself. He feels his legs and they feel broken, he feels his arm and it feels broken, he feels his ribs and they feel broken. Finally he looks at his hand and notices that it is broken. If you hand is broken, then everything you touch feels broken, even if it isn't.

Please be careful if you are disabling the ACPI, I notice that several of the ways to do it do not allow you to turn it back on once you have it off, unless you reinstall Windows XP (boo)

I am coming to realize that the fact that the system works well in SAFE mode really does indite at least one of the system drivers.

Regards,

 

This theory really sounds interesting.
Are there any graphics drivers that work for GMA 500 in MPS?
I seem to be stuck in 640x480 and 4 bit color.

 

Reading through this thread - yes, mine freezes randomly as well - I thought I'd contribute the following: My freezes have only occurred, without exception, either when recovering from screen saver or sleep, or when typing in a web browser, IE7 or Firefox. Again, no exceptions so far.

 

Lock up almost always appears to happened when scrolling a webpage with the touchpad (virtual scroll region on the right). I moved to using a mouse thinking that might help but alas today it froze scrolling with the wheel on the USB mouse. I uninstalled Chrome also thinking that was the source but Safari locked up also and IE8 locked while editing a GMail reply. Almost all of these issues (along with the comments about the webcam lock ups) point to a graphics driver and/or chipset problem because that's the one common SW component. I would think a power controller issue would happen under a strain or other exercise of the power system as opposed to a non-expensive webpage update (content was simple and random). Note that scrolling and video would exercise the raster blit functionality of the video chip and poorly constructed drivers could result in a lockup.

Note the lockup is a frozen screen usually showing the scrolling cursor and the lamp control works.

 

Have not found it in searching the past 12 pages, but has anyone tried a new solid state hard drive (SSD)? If power consumption is to blame, one would think the considerably lower load of an SSD (0.5w standby, 2w read/write) would eliminate the problem/theory.

I was just about to order a Super Tallent 32GB SSD from Newegg, but they sold out. A colleague installed the same SSD in his AO110, and it made a HUGE difference in performance... <10sec boot time for XP Pro.

I didn't want to void the warranty, but given that I'm outside the 30 day return period w/Newegg, and sending the unit to Acer doesn't seem to help anyway, I figured I've got nothing to lose.

 

It's unclear whether it is a power problem directly or a problem with the BIOS or other low-level code. At least one person has swapped out hard drives (presumably temporarily) and it didn't help. Another guy says they installed a new Hard Drive in his system and that it still freezes (see a few messages up).

You are probably wasting your money on the SSD if you are only buying it for the sake of fixing the freezing issue. But if you were going to buy it anyway, that's another story.