For the detectives out there..

[Zourous]

We've had a classroom full (30) of Fujitsu Esprimo E3521 desktops for coming up to 2 years now (1 year of warranty left). For the first year they were mostly ok, from memory. Now there seems to be a problem with 5 or 6 that intermittanly won't boot in the mornings. We've resorted to checking this room every morning and swapping out (for another model) the ones that decide not to boot. When they don't boot, the power light comes on, the fans whirl up and stay on constantly with no output to the screen, and Fujitsu support haven't worked it out either. Two or three have been sent off and returned with no fault found. Here are the clues \ things I've tried:

 

It's not the motherboard or power supply, Fujitsu have swapped one or both at the same time on different PC's and the problem comes back.

 

It's not the cpu, heatsink or fan. Swapped one from a working PC when the fault occured, same fault.

 

It's not to do with the power source \ power leads. We've moved the PC's around within the room and the same ones are re-offending.

 

It's not the memory or any other connectable drive

 

I'm thinking it is something to do with a heat issue because the offending PC's if left unplugged from the power will work fine the next morning, but then ones that are turned off but are still plugged into the power can still pose a problem in the morning, and I can't see how these would be getting hot if they aren't actually on??

 

I'm going to try redoing the thermal paste when some arrives but wondered if anyone could throw anything else into the ring because I'm running out of ideas now.

[Hefferzzzz]

Rats. Or Mice. Or Gremlins.

[Willott]

Could it be to do with a power state they're going into on shutdown/hibernate? Can you test how long you have to leave unplugged for the machine to start playing again?

[Achandler]

This might be a real long shot, but it has happen to me. We has some Lenovos (in a language lab) that had the headphone jack and USB sockets next to each other, and they kept not turning on.

 

If a student accidentally put the jack in the USB socket, even just touching the end, the computer would turn off (if on) and refuse to boot for a random amount of time, there was no consistent fix.

[plexer]

Could it be to do with a power state they're going into on shutdown/hibernate? Can you test how long you have to leave unplugged for the machine to start playing again?

 

I had that with a Dell OptiPlex once it went into power save you had to pull the plug to get it to come back up, disabled powersave and it's never done it since.

 

Ben

[Hulbert99]

I had the same with some Veritons. Did the above fix of pulling the power for 30 seconds and it would then boot.

[Zourous]

Pulling the power short term won't work. It takes roughly around half a day before I can get the offending PC to boot again, which is why it makes it such a pain.

[mac_shinobi]

Get an electrician to test all mains sockets / circuits etc in that room make sure its not related to that ?

 

Update the BIOS on all the PC's ?

 

As above disconnect all equipment from the pc's that is not required ie only have USB Keyboard and mouse attached, anything else USB wise disconnect

 

Also ref over heating I presume after they have been powered off for an hour or so you have used an aerosol can or something similar to blow out all dust from the power supply / heat sink / fans etc ?

Edited May 20, 2013 by mac_shinobi

[Grey-gear]

How warm is the room they are used in? I found a smialar problem with Dell OptiPlex but the room they where in was like a oven. We tried diffenerent things in the room but in the end never really got to the bottom of it. We did have a aircon in the room and it helped for a bit. You could ask users to turn off the computers when they are done.

 

Sorry I could help.

[mac_shinobi]

How warm is the room they are used in? I found a smialar problem with Dell OptiPlex but the room they where in was like a oven. We tried diffenerent things in the room but in the end never really got to the bottom of it. We did have a aircon in the room and it helped for a bit. You could ask users to turn off the computers when they are done.

 

Sorry I could help.

 

If that is the case maybe it needs some ethanol or the likes to clean off the old thermal paste from the heatsink and cpu and re do the thermal paste yourself using something like arctic silver 2 or the likes and putting the heat sink back on etc ( obviously not too much thermal paste ).

 

The other option is a replacement cpu and ram ( as had some Lenovos ) where they were using cheap ram and also the cpu needed replacing as it ran the hard drive diagnostics fine no issues but always kept rebooting on the RAM diagnostics which on other machines only took a minute or two but tried again on the faulty pc and they swapped the cpu ( processor ) and the pc works fine now.

 

Maybe a bad batch or the cpu has become faulty for whatever reason ?

[philtomo-25]

We have a batch of E3510`s which have very similar problems to the one you have. In our case it was the 3 capacitors on the Motherboard which was the problem.

The error we had was the whirling sound upon boot up and also the power completely cutting out within about 30 secs of boot up. Changing the capacitors did the trick for us and makes the PC usable again.

You can tell if its the capacitors by checking the top of them to see if they are swollen, if so then you need to change them. you can pick them up on line (bag of 10/20) for little as £10 online.

 

Phil

Edited May 20, 2013 by philtomo-25
typo

[mac_shinobi]

We have a batch of E3510`s which have very similar problems to the one you have. In our case it was the 3 capacitors on the Motherboard which was the problem.

The error we had was the whirling sound upon boot up and also the power completely cutting out within about 30 secs of boot up. Changing the capacitors did the trick for us and makes the PC usable again.

You can tell if its the capacitors by checking the top of them to see if they are swollen, if so then you need to change them. you can pick them up on line (bag of 10/20) for little as £10 online.

 

Phil

 

Didn't really think of that as if they still have warranty and have had the motherboards swapped already then doing that ( not sure here ) but may invalidate the warranty possibly ??

 

Also depends how good or bad the person is at soldering as not sure if you can damage components etc by over heating them or getting solder running between the pcb tracks etc etc ??

[mac_shinobi]

Just had a user knock the base unit over here ( different make of pc ) but re seated the ram and it powered on again ( had similar issues to what you mention ref power and whirring fans ) but after reseating the ram it powered on fine again ? Just a thought

[sdc]

We have a batch of E3510`s which have very similar problems to the one you have. In our case it was the 3 capacitors on the Motherboard which was the problem.

 

 

This would be my guess, too. Had a similar situation with some monitors doing exactly this; it was the three capacitors that regulated the power input. Once they'd completely discharged (over a good few hours) they were OK again, but once replaced the problem was solved. Some soldering dexterity may well be required on a motherboard; however it's perfectly possible to do if you're careful enough and have steady enough hands!!

[philtomo-25]

we have also had that aswell, but usually with a slightly different model. when the RAM is re-seated then pc works fine.

Ours are several years out of waranty so it was either scrap them or give it whirl with the capacitors seemed to have done the trick for us.

Our method was to cut the capaitors in half and them pull them off leaving the legs attached then soldering the news caps on.

[sdc]

Our method was to cut the capaitors in half and them pull them off leaving the legs attached then soldering the news caps on.

Agreed, this is the simplest and safest method in my experience too.

[mac_shinobi]

however it's perfectly possible to do if you're careful enough and have steady enough hands!!

 

Not me which is prolly why I don't work in a bomb disposal squad lol

[Zourous]

Thanks for the ideas guys but it isn't the MB's. Fujitsu have swapped these out and sometimes it happens the next day. Unless they are swapping bad for bad. They did swap a board one time and we got defective nic in return. Also it's not the ram, i tried that early on. Ram or no ram it does the same thing. Any how i'm going to try some more tests to see how i get on.

[mac_shinobi]

Thanks for the ideas guys but it isn't the MB's. Fujitsu have swapped these out and sometimes it happens the next day. Unless they are swapping bad for bad. They did swap a board one time and we got defective nic in return. Also it's not the ram, i tried that early on. Ram or no ram it does the same thing. Any how i'm going to try some more tests to see how i get on.

 

Looks like you answered ya own question then. Swapping bad for bad. Maybe the motherboards they have use cheap nasty capacitors !!

[Zourous]

Just to recap, once I find a PC in the morning that doesn't boot, I unplug it and in a few hours to half a day it will power on like normal again

 

I'm starting think it is something to do with the case. Something is affecting the MB. I had look inside the case and noticed the PCI slot blanking plates are quite close to the MB (see pic). There is probably 1mm gap between the edge of the MB and the case, and I can get a piece of paper inbetween them. It could be that others are actually touching the MB?? I did remove the blanking plates from 3 of our biggest offenders last night, but this morning 2 of them were at it again, or could the damage have already been done previously. Also I don't know if this is a factor or not, but sunlight shines directly onto the cases in this room. The PC's affected do seem to be in the line of direct sunlight but I could be talking complete crap. I have moved the PC's around the room and it seems to be the same ones.

[Zourous]

Just to recap, once I find a PC in the morning that doesn't boot, I unplug it and in a few hours to half a day it will power on like normal again

 

I'm starting think it is something to do with the case. Something is affecting the MB. I had look inside the case and noticed the PCI slot blanking plates are quite close to the MB (see pic)[ATTACH=CONFIG]18348[/ATTACH]. There is probably 1mm gap between the edge of the MB and the case, and I can get a piece of paper inbetween them. It could be that others are actually touching the MB?? I did remove the blanking plates from 3 of our biggest offenders last night, but this morning 2 of them were at it again, or could the damage have already been done previously. Also I don't know if this is a factor or not, but sunlight shines directly onto the cases in this room. The PC's affected do seem to be in the line of direct sunlight but I could be talking complete crap. I have moved the PC's around the room and it seems to be the same ones.

[brewerd]

We have a similar issue with some of our RM machines, our fix was to flash the BIOS for the latest version and then replace the power supply, as there is a fault with some intel BIOSes that interfere with power up cycles and can knacker power supplies and cause the intermittent power on issues.

 

I'd say its worth trying on a couple of your problem machines. If you can't get them on to flash the BIOS unplug all USB cables reset the power and try again, as this has also helped me turn on trouble machines in the past.

[Marci]

NEC ML530's did the same... usually pop the lid off, flip the clear CMOS jumper, and they'd fire straight up again. Some would terminally corrupt their BIOS, so swapping the bios chip for a preprogrammed spare became most convenient quick fix.

[philtomo-25]

i would serioulsy be looking at this being a capacitor issue, do you have one that is out of warranty? if so it might be worth swapping the 3 capacitors and seeing if you have anymore problems, this seems VERY similar to our issues with these pcs

[Zourous]

I've looked at the capacitors and none of them are bulging or anything other than flat. Also we've had 6 MB's replaced and since then 5 have reoffended and yes Fujitsu have tried the PSU + MB at the same time and it still happens. The bios is up to date, so nothing to change there. It is like something is affecting the MB. I left the 3 usual suspects on (after school) yesterday (they do a normal shut down later on) and this morning on power-up 2 are not booting. Some time later today, they will decide that I can boot them again.