Computer turning on and off instantly

Twinkielol

Distinguished
Feb 16, 2016
79
0
18,640
Well... my newish custom build just died on me. I opened her up made sure to discharge static and then zip tied my psu's power cables together to make room and clean it up, turned it on half a second later it shuts off and keeps doing it until i unplug it.

Psu is an evga 850w supernova g2. Almost two months old.
Mobo is a b85m-g43
Cpu is an i5 3.2ghz quad 4460 i think.

Dont tell me the obvious stuff like check plugs and temps. Its not it.

Bad psu? Rma it?
 
Solution
is there a click when it dies? based on your description there is a short, and the over current prevention circuits are shutting it off. check for any squashed cables or one that are in tight areas. Also it can't hurt to try to reconnect the cables, you could always find the fault that way. If that fails take it out of the case and try the paperclip test method and see if it runs. try to put a fan on the molex cables as those power supplies should always have some sort of load on them. If you do not feel safe using the paperclip method, you can try to buy a PSU tester for cheap,or just get an RMA and hope it is the Power supply.
 
Was it doing it BEFORE you zip tied the PSU cables? If not, then you may have slightly unplugged one of the connections that might not have been fully seated before you tidied up the cables. Nothing else could account for it unless possibly you knocked something else loose like one of the memory modules or the CPU fan connector from the fan header. If the fan connection doesn't see an RPM signal, it will behave exactly like what you are seeing.

Is the CPU fan spinning up while it is on, before it shuts down? Pretty doubtful that the PSU would just go bad between when it was working and after you tied the cables up. Despite saying NOT to mention "plugs", I'd seriously triple check every connection from the PSU or elsewhere that might have been knocked free, check RAM is fully seated and that the graphics card is as well.

If all else fails, I'd bench it.

http://www.tomshardware.com/faq/id-1753671/bench-troubleshooting.html
 

Twinkielol

Distinguished
Feb 16, 2016
79
0
18,640
Trust me. I went through and triple checked all plugs are seated and plugged in to the fullest! Everything lights up my leds and whatnot. I do hear a click... huh. My psu came with a tester i can try it out :) if its the psu it wouldnt work with the tester right? I didnt zip tie too tight... a short as in touching wires orrr?

Edt: if my mobo is bad im totally chill. I hate it anyways and want an excuse to get a better one XD. Itz too cheap to be worth rmaing. The psu on the other hand....
 


pretty much when two wires of opposite polarity touch, or they both touch a common conductive medium ( a side panel lets say) Also it could be a single wire touching the case fram in which case it could end up grounding out making a short. If the PSU fails on the tester it is almost certainly dead.
 


"bad power" what is that? And I would not just start taking parts out of a system and powering it up. perhaps trying a single ram stick at a time, or no GPU, but just pulling stuff out is not a good way to test.
 


I would try, Zip ties can put a lot of pressure on cables. if they are pinched tightly the odds of there being a short at the zip tie spot or around a bend go up. either way when you test it you will have to take it out of the case most likely. if your eally need the zip ties you can try using a ball point pen to slide them backwards.
 

Twinkielol

Distinguished
Feb 16, 2016
79
0
18,640
Good news! It seems to have worked itself out thus far. Must have been static. It kept resetting instantly waited a while got longer in between. I took everything apart tested the parts individually it was definitely my mobo as the psu functioned fine on the solo test and just plugging in the cpu sata and mobo was still causing it to go funky. So i decided to reset the cmos as some people said it worked for them. After that it would turn on for 20 seconds reset and then stay on. I found a beep speaker and plugged it in. Told me no ram and thats when i realized the restarts were now because of the ram i removed when testing mobo. Seems like the cmos did it. Although things were goofed up after. Not sure if it was cmos for sure or timing.

My techie friend said that he had the same issue once. He let his pc sit all night and it worked fine the next day.
 


This is ridiculous. 100% not based in truth/fact. There is NO such thing as "bad power" getting "stuck" in the motherboard. Unless this is a case of getting lost in the translation due to another language being your primary language, it makes absolutely no sense. There IS such a thing as "bad" or "dirty" power, but it does not get "stuck" in the motherboard or PSU. BA unselected.
 


I second that.
it would appear that the motherboard did indeed get messed up somehow, but static either would have killed it or done nothing.
 


I'd pretty much guarantee this was fixed by you removing and reinstalling/plugging in something. Sometimes, even when they ARE plugged in exactly right, one of the pins in the connector gets tweaked at a weird angle or just doesn't make a good connection for whatever reason. I have also seen, many times, cable managed wires that were being "pulled" just enough to break the connection in the socket, or cause an intermittent connection.

When managing cables/wires, it's very important to allow slack to remain, at least in the last inch or two just before the plug/socket, so that the connection doesn't get tweaked and none of the individual wires get pulled to where inside their respective little slot in the connector, they do not break connections.

Of course it's possible to have been something like a memory module that wasn't seated well enough, which is very easy to do, even experienced builders sometimes have this happen, but the stress on one of the connection points is the most likely reason. And for future reference, that "tester" that comes with the PSU only tells you one thing, that the unit is actually powering on. It does not tell you if there are problems with any specific rail, minor shorts, grounding issues, failed capacitors or a variety of other shortcomings that might be influenced by other factors or are intermittent. It only tells you that the unit does in fact have power and can turn on.

I've seen plenty of units that tested fine with the jump tester or paperclip but were faulty nonetheless.

After you corrected the original problem, you are most likely right in that the cmos needed to be reset as it saw a hardware change that didn't correspond to the tables that had already been configured. Sometimes hardware changes just won't automatically configure and a reset is necessary.

In any case, NOT a bad PSU, like I said in the beginning. Glad you got it worked out anyhow. Good luck.
 
Solution

Twinkielol

Distinguished
Feb 16, 2016
79
0
18,640
Yea I don't know... I've just had a few people say static buildup can happen without actual damage and to let it dissipate. I'm no tech so I can't say who's right or wrong. However I do properly ground and discharge before I touch. And regardless that wasn't the cause in my case. Even though I originally was afraid that I had fried my Mobo on accident or shorted out the wires. But everything in that department checked out fine, no bends, cuts, or knicks in the wires at all.

All I know is it kept resetting instantly in a never ending loop. I rechecked the plug's connections 4 times with no success. Then I gave up and I took it all apart to check one piece at a time I personally go from the ground up. Power at least turned on it's a start anyways. So then I plugged in the mobo but had issues as I forgot to plug in the cpu/sata power which I didn't unplug from the mobo originally because my sata cable is a jerk and I got tired of dealing with it. So woops my bad there I am forgetful. Got the corresponding power cables hooked up this time around but it was still doing the same thing. Just more time in between restarts now. Frustrated I watched a YouTube video on it. Guy resetting his CMOS did it. I tried it. It started only restarting once then staying on for good. Initially I thought the cmos didn't work because it would still restart but the fact that it would stay on permanently after that caught my attention as it wasn't doing that before the CMOS reset. So I grabbed my beep speaker and it only beeped no RAM and restarted. I realized resetting the CMOS worked, it now only reset because the ram. I was relieved to say the least. I saw another post on here about the same thing, a guy messed with his cables and it happened to him. However the answers weren't useful on that one.

I think you're right. Cables got tweaked had to be fixed and then a reset was needed for it to be recognized. The longer time in between resets was because the issue was fixed but it now needed a CMOS reset. I was basically at the same step as the guy in the video at that point. I'm just being more thorough in this final post on my steps for anyone else who's had issues because my last one was kind of confusing. I do apologize for any redundancy.

Thanks for the help, guys! Appreciate it a ton. I really appreciate that you guys give thorough details on what it could have been whereas some posters just say it's fried RMA it without actually thinking it through or trying anything. Or they'll say it's heat related when all fans and temperatures are cool and functioning correctly.