Can a power supply only work sometimes?

[loavesfishescomp]

Computer: Dell XPS Gen 2 (with the long 460w power supply on the bottom)

Went to client's house to diagnose. Computer was attempting to boot every 1.5 seconds on its own, then shut down. To elaborate, the power supply fans would spin up briefly then shut down, over and over again. Pressing the power button had no change.

What I did on site:
-Replaced Ram to one stick. No change.
-Removed CMOS battery and it would attempt to fire up still, but more frequently, about once every second. Replaced with good battery and it booted up fine. Buttoned up the computer, and the same problem occurred again.

Took the computer back to the shop. It ran fine?! Did a stress test. Passed. Tested the voltages during stress test via Everest. Fine voltages across the board. Left the computer on all night, came in the next morning, and still running fine!
Couldn't replicate the problem. Gave it back to them to test it, as it was running fine here. Gave client a new power cord, and asked them to try it in a different outlet. They said they did. Computer ran fine for one night. Next morning, the computer had the problem (repetitive 1.5 second boot).

Again the computer is in my shop waiting for a new power supply in the mail. And it's running fine?!

I am guessing power supply. Any other suggestions? Please help me. Ty.

 

[B Trevathan]

Sounds like a short somewhere to me. If it is a short then starting and stopping over and over again can mess the hard drive up.

Did you test the switch with a multimeter?

Have you tried running with just the bare minimum plugged in, like just having the power, hard drive, video card and keyboard plugged into the motherboard and nothing else?

Did you do any testing on the PSU? Did you plug the PSU into a PS tester?

Try draining the flea power by unplugging everything from the computer like power, monitors, printers, USB devices, etc. and hold the power button in for 30 secs or so, then plug everything back in and test it.

Do you see any physical damage on the motherboard? Look for bad capacitors, are any of them swelled even a little, sometimes caps blow from their bottoms and its hard to tell that they have popped except for just a little swelling on their tops.

Do you get any beeps? What about any diagnostic lights codes, does it ever stay on long enough to give you any?

 

[hondablaster]

BOTH a failing motherboard or PSU can usually exhibit this same type of symptom. Its my semi professional experience, I'm going to lean toward PSU about 80% or above. I'm not sure if your PSU on the Dell has a 20-24 proprietary connector or not but if its a standard ATX connector you could have been able to plug in a handy reliable PSU and see the problem stop immediately.

Its kinda hard considering the PSU/mobo has time to cool down shrink/expand on the ride to your shop. Best time to pre-nab this type of issue is on site if you can help it. Sometimes its easier said than done. Likewise Dell seems to like propriety stuff.

Ive heard stories it was the power of the house that caused the issue. I'm not an electrician so I can't say any specific details but it is a possibility. My friend is a Journey Man electrician and he said be careful with homes that have the old school groundless connectors (2 prong) they tend to wear out electronics because of something do with lacking a ground and having noise......

My 2 cents.

 

[hyfidel]

Why would it be a short or a faulty PSU if it works fine in the shop? Sounds like dirty power or a bad USB peripheral/cable (keyboard, mouse, etc.). Try it with a keyboard and mouse from the shop, and unplug everything else. If that doesn't do the trick, try plugging in on another circuit. Customer may need a battery backup to smooth out the power.

 

[iisjman07]

PSUs can be 'flaky' where they appear fine one second and not the next

 

[4ycr]

I had a motherboard that did this but it would not stay powered on long enough to test anything.

 

[Alan22]

This could be a faulty power switch.

I recently had a Dell Inpiron 530 that would randomly power on/off like it was possesed. I replaced the power switch and it's been fine. I installed the bad switch in another computer and it experienced the same power on/off issues.

 

[kotarel]

I had that kind of issue once where the CPU wasn't placed in properly (probably moisture) taking it out and replacing it in fixed the problem.

My 2c.

 

[BDPCR]

chip creep, bad psu , bad switch although it does work well in his shop, or mobo. My bet is bad power and a resulting dying PSU, make sure that client buys a line conditioner.

 

[loavesfishescomp]

Great information from all. Thank you.

I will reseat the chip, put in a new (dell proprietary) power supply (great deal on eBay for 19.00 NEW here for your future needs), looking bent USB pins (brilliant), bad caps again, but not the electricity in the house, and possibly a UPS to smooth things out.

I will post back if and when I find the solution. Cheers.

PS: Curious. Why would removing the clock battery make the power cycle twice as frequent?

 

[hondablaster]

PS: Curious. Why would removing the clock battery make the power cycle twice as frequent?

Its more than a clock battery it holds all the settings of the BIOS. Think of that battery as a power source for the worlds smallest piece of RAM. Its CMOS.

Without it, your BIOS would not know how big the HDDs are or how many. Which drive is primary, boot order etc, how fast it should run etc etc. This is why is rapidly reboots, its in a reconfiguration cycle probably, with no battery there is no data so autoconfig cycle repeats.

http://bugclub.org/beginners/memory/cmosram.html

 

[loavesfishescomp]

Update

Update

Came into the office this morning and the computer was doing the repetitive power cycle. Unplugged it for 40 minutes, after plugging it back in, it boots fine. Odd. (It was shut off last night)

Will post back.

 

[B Trevathan]

The coin shaped battery that supplies backup power to the CMOS is just a backup battery. Think of it like the batteries that you would put in a radio that you have in your house, you have the radio plugged into the wall and you enter in all the stations you want into the preset buttons, if you then unplug the radio and you don't have batteries in the radio you lose all the preset stations, but if you have batteries in when you unplug then the preset station will still be there.

If that battery is dead you will still be able to boot your system, it will just not remember any of the settings, for example if the battery is dead and you unplug your computer and then plug it back in and start it up your computers time may say 12:00 am. most drives will load because most BIOS default settings are for auto detection.

Came into the office this morning and the computer was doing the repetitive power cycle. Unplugged it for 40 minutes, after plugging it back in, it boots fine. Odd. (It was shut off last night)

I guess this rules out the power going to the computer as being dirty, bad, noise or groundless.
Did you say that you have already put in the new PSU?
How many of the suggestion have you tried that we gave you?

 

[loavesfishescomp]

I do understand about the BIOS being reset, but why the absence of a BIOS battery would increase it's repetition frequency [to twice as frequent] baffles me. Excuse my ignorance.

I have not inspected the MOBO for leaking capacitors, or reseated the chip yet (the USB connections seem fine)...Will do that tomorrow, as the power supply should be here in the next 3 days.

The thing that (also) makes no logical sense to me is this: If the chip was heating/cooling and therefore not seated right, it would seem that this would not cause this thing to fire-up every 1.5 seconds (with a turn-on duration of .3 of a second) without even touching the power button. The power button has absolutely no effect on the system when it is in this cycle. It seems like there is something allowing power to go to the motherboard intermittently, and I don't know enough to guess if it's the MOBO or the Power-supply.

I will update tomorrow after I do all of your suggestions. & Thank you.

 

[red12049]

Its more than a clock battery it holds all the settings of the BIOS. Think of that battery as a power source for the worlds smallest piece of RAM. Its CMOS.

Without it, your BIOS would not know how big the HDDs are or how many. Which drive is primary, boot order etc, how fast it should run etc etc. This is why is rapidly reboots, its in a reconfiguration cycle probably, with no battery there is no data so autoconfig cycle repeats.

http://bugclub.org/beginners/memory/cmosram.html

Honda,

You've been around here for a fair piece now. You should have at least an idea why the forum was in such upheaval a few months ago, and many members left and/or were banned.

This post is simply not factual. Even the page you linked to does not support your theory, not to mention that it is copyright 1992. A few important things have changed since then.

Common sense indicates that, if this were normal behavior, it would be impossible to reset the CMOS if the battery ever did die, even if you replaced it.

Please folks, confirm your hypotheses before posting.

Sorry if I sound like an old curmudgeon, sometimes it just leaks out.

Rick

 

[loavesfishescomp]

Update

Update

-All capacitors look fine.
-Per your suggestions, I accessed the CPU and found it wasn't properly seated?! I rectified it and will do further testing.

Will post back.

 

[B Trevathan]

It seems like there is something allowing power to go to the motherboard intermittently

There should be power going to the motherboard if the computer is plugged in. It doesn't matter if the computer is on or off, when it is plugged in to the wall outlet there should be +5 Volts going to the motherboard from the power supply through the purple wire on the main power connector (P1). This is why its a good idea to unplug your computer when you work on the inside of it, The purple wire provides +5 Volts standby / +5VSB (on Dells it listed as +5VFP for flea power) at 2.0 Amps and less than 1 Amp can kill you! Use a multimeter and tell me if you have 5 Volts on the purple wire.

The power button has absolutely no effect on the system when it is in this cycle.

Its real easy to test the power switch, just unplug it and hook it to a multimeter set to read continuity and press the button, if you're using a analog multimeter you will see the needle move way over to the right when you press the button.

I don't know enough to guess if it's the MOBO or the Power-supply.

I recommend testing the PSU with a power supply tester but if you don't have one you can still test the power supply by, unplugging the computer from the wall outlet and then unplug all the power connections inside of the computer and on the power supplies main power connector use a wire or paper clip to short the green wire to one of the black wires (Grounds) and carefully lay it aside where it won't touch anything and plug the computer back into the wall outlet, if the fans on your power supply spins then your PSU should be good. The green wire is the Power On wire, when this green wire is grounded to a black wire it tells the the PSU that power should be delivered to the power rails (Turns your computer on, it is like pressing the power switch). When it is not grounded it tells the PSU not to deliver current to the rails (Your computer is off).

 

[Larry Sabo]

The purple wire provides +5 Volts standby / +5VSB (on Dells it listed as +5VFP for flea power) at 2.0 Amps and less than 1 Amp can kill you!

I'm not sure how many amps it takes to kill one, but 5v certainly isn't going to drive it. Out of curiosity, I just stuck my DVM probes in my mouth and got around 200K ohms, which means 5v would drive all of 30 microamps. So even if he used his tongue to test that 5v line, it's certainly not going to kill him.

 

[hondablaster]

Honda,

You've been around here for a fair piece now. You should have at least an idea why the forum was in such upheaval a few months ago, and many members left and/or were banned.

This post is simply not factual. Even the page you linked to does not support your theory, not to mention that it is copyright 1992. A few important things have changed since then.

Common sense indicates that, if this were normal behavior, it would be impossible to reset the CMOS if the battery ever did die, even if you replaced it.

Please folks, confirm your hypotheses before posting.

Sorry if I sound like an old curmudgeon, sometimes it just leaks out.

Rick

You are wrong and Im not sure what your point is. And you didnt post a link or state why I was wrong. He asked why it would reboot faster. I gave an accurate hypothesis. Have you ever stuck a CMOS jumper in clear CMOS mode. Has it not rebooted repeatedly? (depending on mobo) What are you talking about.

http://www.tomshardware.com/forum/251082-30-battery-motherboard-matter

Is this ACG?

 

[loavesfishescomp]

Honda
Whenever I've jumped the CMOS jumper to clear the BIOS, it simply boots up and tells me it's cleared, then I shut it off, move the jumper and all's good. If I understand right, removing a battery will not immediately clear the BIOS-- you must wait 5 minutes for it to clear... so if you are correlating a cleared BIOS with an increase in boot-up frequency, then I didn't understand your explanation of why it was occurring and your theory behind it. Please do so because I need to understand better.

And because the chip was reseated (as it was askew), I gave it back to the client to use their computer, and to see if it solved the problem. If it occurs again, then It'll be back in the shop and I will test the 5V purple wire and post back in 24 hours... why the intermittent problem baffles me. The only time it occurs is if the computer is plugged in for 6+ hours with the computer off, and it then you can hear the fans spin up briefly and the power light on the front of the computer light up in sequence with the fans.

Cheers to all.