Jump to content
Corsair Community

Spontaneous reboots with VX450W


fll-freak

Recommended Posts

I have had a VX450W installed in a Linux file server for about two years running 24/7 without apparent problem. Starting around middle December I started to experience a few reboots. Nothing in any of the logs would indicate a problem. The reboots starting happening a closer intervals (7, 4, 4, 3, 1, 4, 1, .7, .7 days). The system is not a power monster. A simple dual core MB, 4G mem, 3 SATA drives, 1 optical, no fancy video just what is on the mobo.

 

Hardware and software have not changed for at least 4 months. I have removed various parts and seem to be down to the mobo and PSU. Not easy to swap either of them out, so I decided to probe the power. Rail voltages as measured on an unused Molex all look good with just 50mv ripple.

 

I then put an oscilloscope on the Power_Good signal the PSU generates for the mobo to start up or reboot. Although I have not had a crash yet (remember Murphy?) I have seen the occasional jitter on the line. I have attached one such screen shot. It shows an glitch of about 60 usec long with an amplitude of +- 1V. The glitch does not seem to correlate to any system activity such as drive activity. Glitch happens every few hours. PSU is connected to house power via an APC UPS (ES-550).

 

Would this suggest a PSU failure or could the glitch be a product of something else in the system calling for a sudden slug of current, or a glitch from the UPS?

 

 

Comments welcome.

 

1/13/2012: I have had a few more glitches. One started to approach the TTL low voltage threshold but I have not had any reboots. The PC is running with its case open to allow for the scope probes. Tomorrow I will run some test leads out of the case and run it closed up. Of course, the few pf of capacitance of the scope probe might just be enough to dampen the glitch to bellow TTL threshold. You got to love the Heisenberg uncertainty principle.

 

-Skye Sweeney

Jitter.bmp

Link to comment
Share on other sites

I do have another computer (several in fact) that I could remove a PSU from to swap in for the VX450W. I have a few more experiments I want to run before performing this fairly invasive surgery. The first is that although I have not added any internal hardware or modified any system HW, in the last few months I have added a USB to RS-485 adapter to collect data from a remote sensor. It is remotely possible this device may be at fault. I am in the middle of the "add it/remove it" test. But with a 7 day possible wait for a failure, this will take some time.

 

Hence my question about the glitch/buzz on the Power_Good signal. Is this something that is seen on properly functioning PSUs? Has Corsair ever monitored the signal integrity of this signal for a few days? How is that signal generated? Is there a micro controller monitoring the power, or is it all done with analog circuitry?

 

I will report back after this last test and unless I learn something interesting, after I swap out PSUs.

Link to comment
Share on other sites

Since the computer was not rebooting, I took the opportunity on Saturday to run a full backup of the data drives (57 DVDs). Still no reboots. I disconnected the oscilloscope monitoring the POWER_GOOD line and ran all day Sunday with no reboots. At this point it looks the failure is at best intermittent with possibly a 7+ day repetition rate. If the failure is not gone for good and I can reproduce it again, I will swap out the PSU.

 

I did notice that I can force a similar glitch on the POWER_GOOD line if I run a paper shredder in another room but on an adjacent power circuit. I am wondering now if my issue may not be bad house power. I will look to purchasing a power line scrubber to put in front of the PSU and UPS.

 

Thanks for the help.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...