Computers feeling the cold!

[dgsmith]

We've had what I assume is a hibernation problem with a particular batch of PCs for some time (like, 2 years) and it really is getting frustrating now. I'll come in of a morning and find a handful are powered up, but are in an indefinate sleep/hibernation mode and not even a reboot will bring them out - you have to physically cut the power to the machine and plug back in to bring these back.

 

We have tried updating the BIOS and even the drastic action of physically swapping the entire classroom with another suite of machines but the same thing happens in either room.

 

Rather annoying as you see the power light on the chassis lit up so at a glance you think it is on, although the LED light on the mouse is not lit up and it is effectively "off". These can not be woken by WOL or keyboard input or by off/on (other than the physical power cut as noted above).

 

The computers where custom built by a company some here will know of, although as the components inside are very industry standard I don't think it's fair to blame the company who assembled them (as it could be a simple issue). Needless to say, it's very frustrating. They boot up and and can be used through the day, so it seems some just decide to go to hibernation for no reason when the majority of the others are fine (and which ones it happens to is random).

[GREED]

No power save settings or anything stupid like that? That was a solution I had for a batch the power saving setting on the BIOS caused issues.

[BatchFile]

You haven't got a dodgey batch of PSUs have you? I've seen that kind of behaviour caused by that in the past...

 

MIght be interesting to change the PSU in a couple and see if it fixes them??

[dgsmith]

Looked through the BIOS on a few countless times (I do it every few months when it gets very annoying) but never found anything obvious in there. One machine could be fine for some time, this will happen, then be fine again. There isn't any consistancy.

 

Interesting note about the PSUs - they're under warranty but how would we justify getting some swapped to test this?

[GREED]

Oh I just thought of something, check what USB connections you have. Hear me out:

 

We had a batch of PCs that worked fine, but upon sleeping would not wake up due to a combination of USB attached extensions and devices. Took ages to diagnose and nothing was the same, although some USB extensions were often the cause...

[dgsmith]

Oh I just thought of something, check what USB connections you have. Hear me out:

 

We had a batch of PCs that worked fine, but upon sleeping would not wake up due to a combination of USB attached extensions and devices. Took ages to diagnose and nothing was the same, although some USB extensions were often the cause...

 

USB mice and USB keyboards only, with no other USB cables/connections. Some are PS/2 mice though it happens on the PS/2 and USB connected machines.

[GREED]

Out of ideas!!!

[_Bob_]

Check your capacitors. I've seen machines do this when they have dodgy caps.

http://www.edugeek.net/forums/hardware/3313-leaking-capacitors.html

 

[AngryTechnician]

I recently dealt with two machines side by side, supposedly identical spec, one of which exhibited the above symptoms while the other did not. After a lot of cursing, a careful examination revealed that the only differing components were the RAM sticks. The working one had DDR400 in it, while the other had DDR333. The DDR333 didn't report any errors in Memtest, and the machine otherwise worked fine, but swapping it for DDR400 that was identical to the working machine immediately solved the problem.

 

I'm not sure whether it was the speed change that helped, or some other minor specification difference - but it did demonstrate that some machines can be unreasonably picky about what memory you use in them. If you have some spare lying around, it may be worth swapping the RAM in one of your problem machines and seeing what happens.

[dgsmith]

Check your capacitors. I've seen machines do this when they have dodgy caps.

http://www.edugeek.net/forums/hardware/3313-leaking-capacitors.html

 

 

Would this effect normal operation of the computer though, should this be the case? I suppose when we get a few moments we could check a few, though the operation of the PC is unaffected once it is in normal use. I suppose it's one of the few things we haven't checked yet, and not many other options open other than trying to get some PSUs swapped..

 

I recently dealt with two machines side by side, supposedly identical spec, one of which exhibited the above symptoms while the other did not. After a lot of cursing, a careful examination revealed that the only differing components were the RAM sticks. The working one had DDR400 in it, while the other had DDR333. The DDR333 didn't report any errors in Memtest, and the machine otherwise worked fine, but swapping it for DDR400 that was identical to the working machine immediately solved the problem.

 

I'm not sure whether it was the speed change that helped, or some other minor specification difference - but it did demonstrate that some machines can be unreasonably picky about what memory you use in them. If you have some spare lying around, it may be worth swapping the RAM in one of your problem machines and seeing what happens.

 

Just checked several on cpu-z (one that experienced the issue this morning and some others), and they all seem to have the exact same RAM, right down the manu, part numbers and speed. Bummer, as I thought there may have been some in this then!

[elsiegee40]

My PC at school does this from time-to-time - the MoBo battery is failing (must replace it sometime ) and I think it's caused by that.

 

My father-in-law's PC would only start when parked next to a radiator... in this case I have no idea what's caused the problem and replaced the PC.

[featured_spectre]

CMOS Battery is not trickle charging when it is switched on. I had this fault a while back with a PC in the admin office.

[dgsmith]

Ok, but this has been a problem since the PCs were new, not a developing fault over time. Would this simply be a case of getting the batteries replaced or is it the mobos that aren't sending proepr charge? I must say there has never been a problem with the system time on the computer resetting or anything like that though.

[synaesthesia]

If "cmos" batteries trickle charged there is certainly a fault - you'll explode them!

[GREED]

I had a whole batch of PCs that had this fault (CMOS battery) that had to be swapped out, never worked properly from new. Various errors (although not the one mentioned above).

[carvjo]

Possibly a build quality issue - big brand PC's are assembled in 'clean' environments, often by machines. The danger of contamination from dust, finger grease, static shock etc. is minimised. Joe Bloggs PC assembler may well use good components and take reasonable care while bolting PCs together in his garage/workshop/kitchen - but it's just not the same!

 

Go back to the PC Assembler and ask them to sort it if - failing that blow the dust out and run degreaser on the memory / processor / battery contacts - (give them a good spray with Aero Klene 50 or similar)

 

I know it's too late now - but I that's why I tend to buy Dell with 3yr warranties - any problems in that time (rare) - they send an engineer next day, swap the components and it's fixed..

[difinity]

We had a suite of computers, where randomly some would go to sleep. After doing some calculations on power consumption i guessed the PSU wasn't up to the job. We had the suppliers swop all of the power suplies and the problem went away.

 

They showed exactly the same problems, only a restart would cure it, and there was no light on the USB mouse.

[Poco]

personally i would take 5 machines that are giving this behaviour (this will mean they are out for some time)

 

Now i have my 5 i would change 1 component in each and wait to see what happens. if no results mark down what was changed and change it back. then try a different set of components. if after you have changed all of the components you still have it i would then be looking at software sides such as the CMOS side of things and maybe buy 3 batteries and change them. the real long and short of your situation is that there is no quick answer and realistically if you dont have parts stocked (which most people dont) enough to do the testing there will be no way of solving it without spending. i think the cmos battery might be a good call although it is a bit strange it is intermittent for such a period of time.