Discussion Computer crashes at the same time every hour due to glitched power strip ?

[coolgod]

Hi I just debugged a problem that was giving me immense headaches, I wanted to share my experience to help possible people in the future. I recently started using an old PC I built 10 years ago again and I discovered it gave random BSODs with different codes each time about once a day. The computer was running windows 10, it had an i5-4670 CPU, 1 stick of 8GB ram, 1 SSD, 1 HDD, 1 small graphics card and a 500W PSU, the components were all from brands which generally had a reputation for good quality.

I updated the BIOS/drivers but couldn't figure out the problem. I removed the graphics card and started to use the integrated GPU, and it seemed like the problem got better? It came back again so I reinstalled windows 10. The random BSODs were gone but now the computer random freezes with solid flashes of random colors on my screen then automatically shutdown in about a minute. I thought it was a graphics card driver issue but after installing many drivers and configuring different settings nothing changed.

I was convinced it was a hardware issue so I tried to get event logs and memory dumps. No matter what settings I configured, I could not get a memory dump or event log to analyze the crash. I thought it was a memory issue so I ran the memory check tool in windows and HCI memtest, both of which came up clean. Then one day the crashes seemed to get more frequent, and I noticed a pattern, it seemed like the crashes all happened at XX:48 time regardless of the computer runtime and running applications. A new type of crash also occured, a pure black screen then the same 1 minute delay auto shutoff.

I thought at first my PSU was dying and maybe there was some sort of repetitive ultra short voltage sag event on the local power grid that triggered these crashes. I dismissed this idea since there would probably be too many complaints from others living in my city also. It was then I realized the problem was probably my power strip. Even though my monitor, modem, router, speaker and phone charger were also plugged into this power strip, I haven't noticed any problems with them. I changed the power strip and my computer crashes have all disappeared.

The culprit was a programmable energy saving power strip, even though I never used the clock program features, I always used it as a basic power strip. The power strip also has a coin battery inside, probably used to keep programmed clock settings alive in memory. I think this cell battery died, which isn't obvious since the power strip display is still powered by AC. I suspect
somehow this dead cell battery caused the circuits to initiate a voltage sag event at first randomly then later at XX:48 time.

The power strip model in question is Woods-049690, I found an old walmart link here. https://www.walmart.com/ip/New-Wood...9?wmlspartner=wlpa&selectedSellerId=101024090

I will replace the coin battery and see if original power strip still give crashes. The moral of the story is first don't plug your PC into fancy power strips. Second, if you are getting random crashes (BSOD/flashing colors on screen/black screen), or oddly repetitive crashes, try replacing your power strip.

 

[DaleH]

Hi I just debugged a problem that was giving me immense headaches, I wanted to share my experience to help possible people in the future. I recently started using an old PC I built 10 years ago again and I discovered it gave random BSODs with different codes each time about once a day. The computer was running windows 10, it had an i5-4670 CPU, 1 stick of 8GB ram, 1 SSD, 1 HDD, 1 small graphics card and a 500W PSU, the components were all from brands which generally had a reputation for good quality.

I updated the BIOS/drivers but couldn't figure out the problem. I removed the graphics card and started to use the integrated GPU, and it seemed like the problem got better? It came back again so I reinstalled windows 10. The random BSODs were gone but now the computer random freezes with solid flashes of random colors on my screen then automatically shutdown in about a minute. I thought it was a graphics card driver issue but after installing many drivers and configuring different settings nothing changed.

I was convinced it was a hardware issue so I tried to get event logs and memory dumps. No matter what settings I configured, I could not get a memory dump or event log to analyze the crash. I thought it was a memory issue so I ran the memory check tool in windows and HCI memtest, both of which came up clean. Then one day the crashes seemed to get more frequent, and I noticed a pattern, it seemed like the crashes all happened at XX:48 time regardless of the computer runtime and running applications. A new type of crash also occured, a pure black screen then the same 1 minute delay auto shutoff.

I thought at first my PSU was dying and maybe there was some sort of repetitive ultra short voltage sag event on the local power grid that triggered these crashes. I dismissed this idea since there would probably be too many complaints from others living in my city also. It was then I realized the problem was probably my power strip. Even though my monitor, modem, router, speaker and phone charger were also plugged into this power strip, I haven't noticed any problems with them. I changed the power strip and my computer crashes have all disappeared.

The culprit was a programmable energy saving power strip, even though I never used the clock program features, I always used it as a basic power strip. The power strip also has a coin battery inside, probably used to keep programmed clock settings alive in memory. I think this cell battery died, which isn't obvious since the power strip display is still powered by AC. I suspect
somehow this dead cell battery caused the circuits to initiate a voltage sag event at first randomly then later at XX:48 time.

The power strip model in question is Woods-049690, I found an old walmart link here. https://www.walmart.com/ip/New-Wood...9?wmlspartner=wlpa&selectedSellerId=101024090

I will replace the coin battery and see if original power strip still give crashes. The moral of the story is first don't plug your PC into fancy power strips. Second, if you are getting random crashes (BSOD/flashing colors on screen/black screen), or oddly repetitive crashes, try replacing your power strip.

Good troubleshooting! This is good to know.

 

[coolgod]

Wanted to give an update, so it turns out switching the power strip or its batteries didn't actually solve my problem. I managed to get my computer to crash with a new power strip.

I figured out my computer crashed whenever a phone charger was plugged in the same power strip as my computer while charging a phone, maybe other devices also? I could replicate this behaviour with different power strip, different USB charger, different USB cable and different phones also.

It appears this strange behaviour has been documented before also.

Extremely strange behavior: charging phone causes PC to crash...?

This isn't much of an issue I need help resolving...but more of a curious behavior I've been seeing with my fancy new Samsung Galaxy S4. I first tried charging the new phone using the USB port on my Windows 8 PC. Since my PC is located on the ground, it seemed convenient to place the phone on...

forums.tomsguide.com

I am bewildered at this odd behaviour.

 

[Misgar]

I'd suggest buying a new phone charger, from the manufacturer of the phone. Is it a high power charger, or an old fashioned trickle charger?

I don't know of any phone chargers that exceed a nominal 120W and that's peanuts compared to the power requirements of a big computer. A small PSU should not "crowbar" the mains supply.

Without dissecting the PSU and subjecting it to a number of tests in the lab, it's difficult to diagnose the fault at a distance.

If the charger came with an expensive high-end phone, you'd expect the manufacturer to provide a high quality charger. If, however, the charger is a cheap no-name unit bought as a replacement, it could be a fire hazard waiting to die.

I've dissected cheap (fake) battery chargers and compared them with OEM units. The fake PSUs are badly designed, use poor quality components and often lack safety protection circuits. But, they often cost one tenth the price of the OEM unit. I wonder why?

 

[coolgod]

I'd suggest buying a new phone charger, from the manufacturer of the phone. Is it a high power charger, or an old fashioned trickle charger?

I don't know of any phone chargers that exceed a nominal 120W and that's peanuts compared to the power requirements of a big computer. A small PSU should not "crowbar" the mains supply.

Without dissecting the PSU and subjecting it to a number of tests in the lab, it's difficult to diagnose the fault at a distance.

If the charger came with an expensive high-end phone, you'd expect the manufacturer to provide a high quality charger. If, however, the charger is a cheap no-name unit bought as a replacement, it could be a fire hazard waiting to die.

I've dissected cheap (fake) battery chargers and compared them with OEM units. The fake PSUs are badly designed, use poor quality components and often lack safety protection circuits. But, they often cost one tenth the price of the OEM unit. I wonder why?

I replicated this problem with a good high quality OEM phone charger and a generic USB charger that came with a bluetooth speaker. Both chargers are under 25W.

Do you think the interference takes the path of the power supply? I thought the interference was some sort of EM waves that travelled through air. I think I can run some test to figure out.

 

[Misgar]

No, you get "conducted" interference (along wires) as well as "radiated" interference (through the air) from most electronic equipment.

Switched mode power supplies and computers generate large amounts of both conducted and radiated interference. This interference can affect nearby equipment.

Computers are also susceptible to interference from external sources. If you place a mobile phone next to a computer, there is a small chance of the phone affecting the computer, e.g. strange noises from the computer speakers when the phone contacts the local base station.

Try physically separating the phone charger and the phone from the computer. The old "inverse square" law operates here as far as radiated interference is concerned. Doubling the separation results in the radiated interference reducing to one quarter. Three times the distance one ninth, four times one sixteenth, five times one twenty-fifth.

Conducted interference travels much further before it is attenuated. If possible, place the phone and charger at the opposite end of the house and see if the interference reduces.