theathena
asked on
Spontaneous rebooting
If I recall correctly, I've only seen this in Win95 machines but I really suspect power issues.
A machine could be in use or it could be idle - out of the blue it reboots. This phenomena is very intermittent. I've searched high and low for info. on this - including MSKB - and have only found info. specific to non-MS OSs (nothing @ MSKB).
The most recent machine I've been told this is occurring on is running Win95 - they are also experiencing parity errors. They swapped out the RAM - and the problems remain. The machine was fine in my shop but when installed on-site - w/in a couple of days they started experiencing these problems. I'm suspicious of their office wiring. I'll have the system back in my shop this week. If the system passes ok in-shop, I'll take it back to the site w/ a UPS in tow. I could be barking up the wrong tree. Any other advice as to how to nail this down would be greatly appreciated.
TIA
Bj
A machine could be in use or it could be idle - out of the blue it reboots. This phenomena is very intermittent. I've searched high and low for info. on this - including MSKB - and have only found info. specific to non-MS OSs (nothing @ MSKB).
The most recent machine I've been told this is occurring on is running Win95 - they are also experiencing parity errors. They swapped out the RAM - and the problems remain. The machine was fine in my shop but when installed on-site - w/in a couple of days they started experiencing these problems. I'm suspicious of their office wiring. I'll have the system back in my shop this week. If the system passes ok in-shop, I'll take it back to the site w/ a UPS in tow. I could be barking up the wrong tree. Any other advice as to how to nail this down would be greatly appreciated.
TIA
Bj
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Just a followup comment. Being in an office, is the problematic PC plugged into a circuit that is being shared with any high powered equipment. You might check current loads and see if there isn't an overloading condition causing line voltage to dip momentarily. If your going to install a UPS, it may help.
AJHanks
AJHanks
ASKER
Thanx, AJHanks! Great feedback in the PS specs. I set up my Asus Probe voltage monitoring utility to the specs you indicated and got an overvoltage on my +3.3vdc. Hmmm, will have to check that out.
I just remembered that the client transported the system I'd repaired for him back to his employer's offices strapped to the back of a bicycle! I later learned he did the same w/ a monitor! Not exactly terrific for maintaining connection or even component integrity.
I've done some more research - dug through my manuals - searched more online. Looks more and more like power to me. Either an internal or external power source problem or as you pointed out a connection problem. I have suspected power problems at those offices for some time.
I'm not sure how to figure out what devices/equipment are sharing the same circuit(s) w/o taking the users' equipment down. This is an office taking up half of a building floor. There are at least 15 computers there, 2 or 3 network printers (laser), and at least one copier.
Any suggestions re: method and equipment needed to investigate power routing and load?
TIA
Bj
I just remembered that the client transported the system I'd repaired for him back to his employer's offices strapped to the back of a bicycle! I later learned he did the same w/ a monitor! Not exactly terrific for maintaining connection or even component integrity.
I've done some more research - dug through my manuals - searched more online. Looks more and more like power to me. Either an internal or external power source problem or as you pointed out a connection problem. I have suspected power problems at those offices for some time.
I'm not sure how to figure out what devices/equipment are sharing the same circuit(s) w/o taking the users' equipment down. This is an office taking up half of a building floor. There are at least 15 computers there, 2 or 3 network printers (laser), and at least one copier.
Any suggestions re: method and equipment needed to investigate power routing and load?
TIA
Bj
Power circuit Tracers
There are test instruments you can use to locate a circuit, even multi circuits. A multi circuit, 2 circuit for example can cost about $830. Additional transmitters cost about $200 each. Check out http://mitchellinstrument.com
Grainger Supply, catalogue #390, page 803, has a selection of single circuit breaker locators. Their web site is www.grainger.com. You plug a transmitter into the outlet where the PC goes, then go to the circuit breaker box, passing the receiver over each breaker or fuse. It even checks for proper wiring sequence and GFCI. Cost is ranges from $35 to $400. Once you've located the circuit breaker (fuse) you can trip the breaker and easily identify all equipment that's affected. If you have a good DVM, one that has a min/mas recording or holding feature, you could monitor the outlet voltage. Most equipment is designed and built to tollerate some degree of line voltage loss. An overloaded circuit, especially if the wiring is not to code, will result in significant voltage drop. Note: these drops can be quite fast in real time, like a spike.
ATX Voltage
Just an added note. I had to replace the entire power supply when I had a problem with out of spec voltages. The dealer exchanged it gladly after I showed him my measurements. The PC worked fine afterwards. The replacement power supply voltages were well within the ATX specs too. By the way, I took measurements with a DVM.
Intermittant connections
I had to replace a mother board to fix the CPU socket problem. There's no doubt that there was a circuit tracing factacture. As you could guess, the tracings are so small and a fracture hard to detect. I used a 200X microscope, viewed each solder joint and couldn't find it. I don't think I have to tell you that thermal stress and mechanical stress can cause such damage. Your client carrying the PC box on the back of a bike is a no no. These boards are too thin for their size and really don't tollerate mechanical shock or flex very well. That's the trade off for low cost.
I think I've unloaded about as much as you can stand.
Good luck
AJHanks
There are test instruments you can use to locate a circuit, even multi circuits. A multi circuit, 2 circuit for example can cost about $830. Additional transmitters cost about $200 each. Check out http://mitchellinstrument.com
Grainger Supply, catalogue #390, page 803, has a selection of single circuit breaker locators. Their web site is www.grainger.com. You plug a transmitter into the outlet where the PC goes, then go to the circuit breaker box, passing the receiver over each breaker or fuse. It even checks for proper wiring sequence and GFCI. Cost is ranges from $35 to $400. Once you've located the circuit breaker (fuse) you can trip the breaker and easily identify all equipment that's affected. If you have a good DVM, one that has a min/mas recording or holding feature, you could monitor the outlet voltage. Most equipment is designed and built to tollerate some degree of line voltage loss. An overloaded circuit, especially if the wiring is not to code, will result in significant voltage drop. Note: these drops can be quite fast in real time, like a spike.
ATX Voltage
Just an added note. I had to replace the entire power supply when I had a problem with out of spec voltages. The dealer exchanged it gladly after I showed him my measurements. The PC worked fine afterwards. The replacement power supply voltages were well within the ATX specs too. By the way, I took measurements with a DVM.
Intermittant connections
I had to replace a mother board to fix the CPU socket problem. There's no doubt that there was a circuit tracing factacture. As you could guess, the tracings are so small and a fracture hard to detect. I used a 200X microscope, viewed each solder joint and couldn't find it. I don't think I have to tell you that thermal stress and mechanical stress can cause such damage. Your client carrying the PC box on the back of a bike is a no no. These boards are too thin for their size and really don't tollerate mechanical shock or flex very well. That's the trade off for low cost.
I think I've unloaded about as much as you can stand.
Good luck
AJHanks
ASKER
I'll check out those sites for test equipment you mentioned. I do have a good DVM so I can ck the AC. I've had my eye on a couple of power monitors at Jensen. Will have to wait. ($) As for repairing motherboards - I don't have access to the fancy repair and diagnostic equipment I had in the Navy. Repair of these multilayer motherboards is out of the question for me out here in the civilian world. Any repair - unless the problem is really odvious (cold solder joint, runs shorted by foreign object or solder overflow, etc.) isn't cost-effective to perform anymore. It's more cost-effective to narrow down the problem to the mobo and change it out. Kinda sad. I miss componenet level troubleshooting but it seems that's a thing of the past.
I really appreciate your feedback AJHanks. Will let ya know how it turns out.
Bj
I really appreciate your feedback AJHanks. Will let ya know how it turns out.
Bj
hi
i have also experianced this with an AT system running win 98, Cyrix MII333 and 64mb EDO - it was the internal PSU, a replacement solved all problems in the system. i then put this PSU on a test rig. after a couple of hours 'on', the system rebooted every 15 to 35 mins without fail. I traced the problem to a fracture in a track on the main PCB in the PSU near one of the small transformers, also a dry joint on same transformer pins ( linked over fracture and resoldered pin and all now ok ) as stated in the previous post pcs and pushbikes do not mix !!
i have also experianced this with an AT system running win 98, Cyrix MII333 and 64mb EDO - it was the internal PSU, a replacement solved all problems in the system. i then put this PSU on a test rig. after a couple of hours 'on', the system rebooted every 15 to 35 mins without fail. I traced the problem to a fracture in a track on the main PCB in the PSU near one of the small transformers, also a dry joint on same transformer pins ( linked over fracture and resoldered pin and all now ok ) as stated in the previous post pcs and pushbikes do not mix !!
This happened to me as well, although I never found the problem I did a thorough scan disk and a defrag, and the problem disappeared.
AJHanks