Can you help diagnose crashing problem with my new x220?

[hes]

Somewhat on a whim, I bought an x220 from a guy on Craigslist today. Lesson learned, never again. Or at least next time I won't be so trusting.

The machine worked okay when I met him to buy it. Surprise, when I get home I find that it seems to crash after it's been running a bit, say 20 to 30 minutes. I think the problem is tied to operating temperature, but I'm not positive yet.

Symptoms:
while working screen goes blank and power goes off. I think sometimes (always?) the keyboard light flashes and maybe other indicator lights flash at time of the crash. When under battery power the machine appears to loop power on and off, indicated by the power button light going on and off. (My guess is that each attempted restart crashes immediately b/c temperature of some part is too high.) I can stop this power-looping by removing the battery. If I wait an hour or two I can start the machine and it will start okay, but after 5 to 30 minutes will crash again (with length of time determined, i think, by how cool machine is when i start it).

Situation is similar when battery is removed and I use AC adapter, except I think after the crash there is no power looping, it just turns off.

Everything else about the machine seems normal. it's an i7, 8GB RAM, 128GB SSD.

I've had an x200 for a few years that I was looking forward to replacing with an x220 with better screen and a little better battery life. But it seems I wasn't careful enough in picking one out. I'm afraid it's a motherboard issue, which may be beyond my ability and/or willingness to repair. Do any of the experts here have a better idea of what problem (and fix) may be? Thanks!

[GomJabbar]

Running hot due to a bad fan or dirty system is not that uncommon. Typically the machine just suddenly shuts off. There is software that will monitor temps for you, such as Notebook Hardware Control (NHC). 100 deg. C is typically the shutdown temp for the CPU.

http://www.pbus-167.com/

EDIT: ThinkPad Fan Control looks a little more up to date than NHC.
http://sourceforge.net/projects/tp4xfan ... rce=navbar

From your description (loop power), your laptop may be going into Standby. Try pressing the Fn key and see if the machine wakes up.

[hes]

Running hot due to a bad fan or dirty system is not that uncommon. Typically the machine just suddenly shuts off. There is software that will monitor temps for you, such as Notebook Hardware Control (NHC). 100 deg. C is typically the shutdown temp for the CPU.

Thanks, good idea, but as it turns out I don't think heat is the issue. The fan runs smooth and quiet. I used TPFanControl and temps stay low: 38C at idle with fan at 0, 34C at idle with fan at 2000rpm. I tested a couple of youtube videos and temps never got above low fifties.

EDIT: Is it common for system to run hotter on battery? Numbers above were from ac-power only. Starting up today with battery it's showing temp of 72C with bios-controlled rpm of 4000. This is while only thing I'm doing is having Chrome window open to view webpage and type this post. Also, I switched TPFanControl from Bios-mode to Smart-mode and this happened: rpm went down to 2000 and temp went down from 72C to 62C. At that point, fan went to 0rpm and temp went back up to 70s. Right now it's under smart-mode with rpm at 3500 and temp at 77C.

EDIT2: I just checked Task Manager for first time. It shows 25% CPU usage, but listed processes show barely any usage at all. If I open Resource Monitor I see taskhost.exe that is showing a constant 25% cpu usage. I'm not sure what that means, but it seems strange.

From your description (loop power), your laptop may be going into Standby. Try pressing the Fn key and see if the machine wakes up.

I'm not sure about this. Do you say this because I say the power button was blinking? It's possible, and I will check this, but I think in any case it's trying to do a restart before it goes to suspend (because the keyboard light blink and blink of indicator lights is same as what it does at startup, and sometimes it gets to 'Thinkpad press blue thinkvantage button' startup message screen before crashing again). For now I'm guessing you're right, though, and that it is ending up suspended; I just didn't realize that you had to press Fn-only to resume from suspend.

Last night I had it running on ac power, no battery installed, and had it going an hour or so before it crashed. I was able to immediately start it up again. I put power management all on 'never' (i.e, screen dim, screen blank, and sleep all on 'never') and I went to bed, wondering if it would last overnight. I have no idea how long it lasted, but when I checked this morning the machine is turned off. Upon finding the machine turned off I decided to start it up with battery only and see what happens. I took these steps (1) unplugged ac adapter cable from back of computer, (2) inserted battery, and (3) pressed power button. Strangely, the machine started up and showed 'resuming Windows' message, and properly resumed to state with same apps open as I had running when machine was under ac power-only last night. I thought the machine required ac- or battery-power to retain suspended state, but there were at least 10 to 15 seconds between me unplugging ac-cable and inserting battery in back of the machine. Does that sound right?

Also, I did boot once from a usb-stick with live version of Linux Mint. It ran for half-an-hour or so before it crashed; identical to crashing symptoms on Windows, as far as I could tell.

I'll be checking today for anything else I can find that might help diagnose problem.

[RealBlackStuff]

If you are running W8 or W8.1, switch off the fast-boot routine: http://www.eightforums.com/tutorials/63 ... s-8-a.html.
That routine (when ON) makes the lappy go into hibernate.

[hes]

If you are running W8 or W8.1, switch off the fast-boot routine: http://www.eightforums.com/tutorials/63 ... s-8-a.html.
That routine (when ON) makes the lappy go into hibernate.

Thanks. It's not W8; it's running Windows 7 Ultimate. Also, I had same thing happen in Linux from live-boot of Linux Mint usb-stick, for whatever that's worth.

[axur-delmeria]

1. You should monitor the CPU clock speed along with the temperature.
I'm currently using Open Hardware Monitor http://openhardwaremonitor.org/ on my X220.

2. Regarding temperatures and power sources...

A. 65w AC adapter without battery = X220 is locked into low CPU speed.
B. battery only = no speed lock, max CPU speed depends on the power profile.
C. 65w AC adapter with battery = same case as B.
D. 90w AC adapter with or without battery = same case as B.

3. Clean the fan and heatsink. Since the X220 is around 3 years old by now, the thermal paste may have dried up. Consult the HMM if there should be thermal pads on certain parts of the heatsink (for cooling other components like chokes or power circuitry).

[crazeazn]

1. You should monitor the CPU clock speed along with the temperature.
I'm currently using Open Hardware Monitor http://openhardwaremonitor.org/ on my X220.

2. Regarding temperatures and power sources...

A. 65w AC adapter without battery = X220 is locked into low CPU speed.
B. battery only = no speed lock, max CPU speed depends on the power profile.
C. 65w AC adapter with battery = same case as B.
D. 90w AC adapter with or without battery = same case as B.

3. Clean the fan and heatsink. Since the X220 is around 3 years old by now, the thermal paste may have dried up. Consult the HMM if there should be thermal pads on certain parts of the heatsink (for cooling other components like chokes or power circuitry).

I certainly think its number 3.

[hes]

3. Clean the fan and heatsink. Since the X220 is around 3 years old by now, the thermal paste may have dried up. Consult the HMM if there should be thermal pads on certain parts of the heatsink (for cooling other components like chokes or power circuitry).

I certainly think its number 3.

I'm experiencing the crashes when CPU is at 55C. Can that have anything to do with fan/heatsink?

Sorry, I also think I may have provided some misinformation. When crashing on battery I think the first couple times the machine was attempting to restart and I somehow inadvertently put it into sleep mode. (Something I did pressing and/or holding power button?) Then I'd see the blinking power button (which doesn't exist on my x200) and I was mistakenly thinking that it was trying to restart but failing, that it had something to do with overheating.

I'm still getting the crashes, mostly after 10 minutes to 2 hours of use, but when i don't touch anything the machine is able to restart, and I get message of unexpected shutdown and choice of how to start Windows, and Windows starts up normally for another 10 minutes to 2 hours. At least that's how it seems to me now; it's hard to test when I can't replicate the crashes on demand.

[Neil]

Test other components, if overheating doesn't seem to be an issue.
Run a few passes of MemTest86+ just to make sure there aren't any errors with the RAM.
Run a hard drive test to scan for bad sectors, or any other anomalies to do with the hard drive.
Also, try running it without the battery installed, and see if it still crashes.
If all those don't return any info about the culprit, come back and maybe we can think of some more things to test.

[crazeazn]

the i7 def. generates more heat than the i5 so I would check/clean that first then do what neil says. You might need new thermal paste for the cpu. I usually run my spray blower on my heatsinks much to my disgust.

[GomJabbar]

Your temps look fine to me. My guess at this point is a bad BGA connection or other motherboard connection issue. It is possible to reflow bad BGA solder connections, but that is hit or miss and not so easy to do.

You likely need a new motherboard IMO, although possibly it could be a memory issue as someone else stated above.

[hes]

You likely need a new motherboard IMO, although possibly it could be a memory issue as someone else stated above.

Thanks, that's kind of how it's feeling to me. I ran the Windows Memory Diagnostic Tool a couple of times, both times finding no errors. I haven't run MemTest86.

And it turns out I wasn't mistaken about the blinking power button. I had another crash similar to the first one I had yesterday. After the crash the power button started blinking, blinking on then off roughly every five seconds. Pressing power button would not start things up again, nor would Fn key, or anything else. Strangely, bottom of computer was warm, but I don't think CPU temperature was high (at least it wasn't at the point of the crash). After I wait an hour or two I am able to start the machine up again.

Assuming this is a motherboard issue, any recommendations on where I should buy a replacement (I assume some reputable seller on ebay)? Also, are there any tutorials out there on how to replace x220 motherboard and is it something a (generally competent) newbie can do? I've installed plenty of hard drives and memory modules, but never taken a laptop computer apart.

Thanks again.

[crazeazn]

You likely need a new motherboard IMO, although possibly it could be a memory issue as someone else stated above.

Thanks, that's kind of how it's feeling to me. I ran the Windows Memory Diagnostic Tool a couple of times, both times finding no errors. I haven't run MemTest86.

And it turns out I wasn't mistaken about the blinking power button. I had another crash similar to the first one I had yesterday. After the crash the power button started blinking, blinking on then off roughly every five seconds. Pressing power button would not start things up again, nor would Fn key, or anything else. Strangely, bottom of computer was warm, but I don't think CPU temperature was high (at least it wasn't at the point of the crash). After I wait an hour or two I am able to start the machine up again.

Assuming this is a motherboard issue, any recommendations on where I should buy a replacement (I assume some reputable seller on ebay)? Also, are there any tutorials out there on how to replace x220 motherboard and is it something a (generally competent) newbie can do? I've installed plenty of hard drives and memory modules, but never taken a laptop computer apart.

Thanks again.

I think if you can do hard drives and memory modules you should be okay. I usually map out which screws go where and the order I took ribbons out so I dont lose track.. Parts are smaller and slightly more fragile on a laptop so as long as you're careful you should be okay.

[rkawakami]

You should download a free copy of the Hardware Maintenance Manual for your X220 from here:

http://download.lenovo.com/ibmdl/pub/pc ... 739_04.pdf

It will give you all of the directions needed to take the system apart (and hopefully put it back together without leaving any "extra" screws).

[GomJabbar]

From the link Ray provided above for the HMM, mentioned are some error codes that might be shown after the POST of the BIOS boot. There are numeric error codes, beep error codes, and error messages that may be displayed. See if any of those apply in your case.

[hes]

Thank you guys for all your help.

I'm pretty sure now that the crashing happens only when ac-adapter is plugged in. I'm testing more now, but I don't think I've ever crashed while running battery-only. Would that just confirm that it's probably the tricky/difficult to fix BGA or motherboard connection issue GomJabbar/DKB mentioned? Or potentially something different (simpler)?

[Neil]

Have you tried a different AC adapter? Could be as simple as a flakey adapter.

[hes]

Have you tried a different AC adapter? Could be as simple as a flakey adapter.

Yeah, I'm 99% sure I've ruled out adapters, both by using a known-good adapter on bad machine (still crashes) and using bad machine's adapter on a good x200 (no crashes).

[dr_st]

I'm pretty sure now that the crashing happens only when ac-adapter is plugged in. I'm testing more now, but I don't think I've ever crashed while running battery-only.

The default power plan locks the processor to the lowest speed when on battery. It is possible that that, and not the presence of AC, makes the difference.

To rule it out I would make sure the laptop uses identical power settings for plugged in and for battery. For example, setting everything to "Maximum Performance". This way - if it starts crashing also on battery - it is probably related to component stress. If it still crashes only on AC, it probably has to do with the power circuitry.

It may not matter much, since it may still require a replacement motherboard, but it may help diagnosis.