Question PC 5V starts off at 5.07, then rises slowly.

[TheSonikku]

My PC’s been having a LOT of power issues lately and now there’s yet another one:
- Using a voltage stabilizer, my PC turns on, and after a few minutes, the 5V voltage spikes and keeps fluctuating between 5.07 and 7.30, constantly.

- Using only a power strip, it didn’t suddenly spike to 7, however it starts off at 5.07V and has been slowly rising, currently at 5,82V. I haven’t tried playing or doing anything intensive yet.

I bought a replacement PSU and both of the power supplies had similar problems (albeit this is the first time I tried using a power strip instead of a stabilizer). I sent it to a technician to check and it ran fine for 2 days with both PSUs without crashing (ASUS Anti-Surge was turned on, so it’d have crashed if it went beyond 6V), but it crashed in less than an hour at my place.

The stabilizer is brand new; as for the specs:

• MOBO: ASUS M5A78L-M/USB3
• GPU: Gigabyte Phoenix GTX1660
• RAM: HyperX GDDR3 8GB RAM (2x)
• CPU: AMD FX8320 Octo-Core
• PSUs: EVGA 850BQ 80Plus Bronze (850W) (Current) and ATX Spark PCYES 75+ (pretty much offbrand, this was the one I bought recently)
• I have quite a bit of USB stuff plugged in, including a LED mouse, keyboard and a drawing tablet (no screen).

As an added story: My PC‘s been crashing randomly before for quite a while but only recently was it because of overvoltage; I finally properly grounded it but the overvoltage problem is still here. I’m putting this in MOBO discussion because I’m not sure if it’s the MOBO voltage sensor gone rogue or if it could really be a problem in my house.

Also, I can’t check with a multimeter yet.

 

[Grobe]

Are you known with other electrical equipment in the house having issues?

Have you an opportunity to measure the voltages directly (using a multimeter) ?

 

[TheSonikku]

Are you known with other electrical equipment in the house having issues?

Have you an opportunity to measure the voltages directly (using a multimeter) ?

Haven’t had the opportunity to use a multimeter; No, no other equipment. Although, for example, the screen goes black for a moment sometimes when people use other electrical stuff in the house. (e.g. fan or microwave)

 

[TheSonikku]

Haven’t had the opportunity to use a multimeter; No, no other equipment. Although, for example, the screen goes black for a moment sometimes when people use other electrical stuff in the house. (e.g. fan or microwave)

ALSO, My PC was stuck at displaying 5.76-5.91V for the longest time, then suddenly it spiked up to 7.2 and has risen to 7.65 so far. (PC’s still on, I was in the middle of making a backup of stuff and it’s still backing that stuff up)

 

[TheSonikku]

ALSO, My PC was stuck at displaying 5.76-5.91V for the longest time, then suddenly it spiked up to 7.2 and has risen to 7.65 so far. (PC’s still on, I was in the middle of making a backup of stuff and it’s still backing that stuff up)

Kept PC still on, went this entire time at 7.65V, suddenly it went back down to 5.9V after about 5 hours. Hm.

 

[MeanMachine41]

I never advocate using a budget bronze rated PSU as usually something of quality is sacrificed to keep prices low.
In your case one important review indicates" strictly average voltage stability."

A PSU should remain within + or - 5% on all rails. OFC a poor household supply would not help.

 

[TheSonikku]

I never advocate using a budget bronze rated PSU as usually something of quality is sacrificed to keep prices low.
In your case one important review indicates" strictly average voltage stability."

A PSU should remain within + or - 5% on all rails. OFC a poor household supply would not help.

Honestly, I agree wholeheartedly but at the same time the voltage fluctuations seem so random that it doesn’t seem like a PSU problem. Especially since at the other technicians’ place, it seemed to work fine for 2 straight days. I’ll only be able to check with a multimeter on tuesday, too...

 

[MeanMachine41]

Regardless of the cause those spikes can kill Motherboard components and peripherals. Voltage rectifiers help stabilize but your spikes are way over the top.
I would cease using your system till you check your power at the wall socket. The PSU can work for two strait days but is that under load?

 

[DSzymborski]

Motherboard voltage sensors are typically garbage. Honestly, if your spikes were that high you probably wouldn't have a PC right now.

It certainly doesn't help that you're running a 125W CPU in a really low-end AM3+ motherboard, which is a notorious terrible combination and avoidance is always urged. I'm not surprised you're getting some odd behavior given that your CPU is literally trying to murder your VRMs. I'd knock the clock down 800 MHz to make it an 8320E, a CPU you could actually safely run, and see if anything changes in the behavior.

The technician information isn't useful without knowing what the technician actually did. There are very good and very poor technicians. If the technician diagnosed your PC simply by running it at his house, then he's probably towards the latter end of the scale.

As for that second power supply, send it to a recycling center. This should never be used in anything ever.

 

[TheSonikku]

Motherboard voltage sensors are typically garbage. Honestly, if your spikes were that high you probably wouldn't have a PC right now.

It certainly doesn't help that you're running a 125W CPU in a really low-end AM3+ motherboard, which is a notorious terrible combination and avoidance is always urged. I'm not surprised you're getting some odd behavior given that your CPU is literally trying to murder your VRMs. I'd knock the clock down 800 MHz to make it an 8320E, a CPU you could actually safely run, and see if anything changes in the behavior.

The technician information isn't useful without knowing what the technician actually did. There are very good and very poor technicians. If the technician diagnosed your PC simply by running it at his house, then he's probably towards the latter end of the scale.

As for that second power supply, send it to a recycling center. This should never be used in anything ever.

I’ve been using the CPU-MOBO combination for well over 5 years now (I actually once even had the MOBO replaced due to other complications, but it was still the same one), and only recently after installing a new cooler did stuff like this start to happen.

While I do agree the other PSU is pretty bad too, it’s still a remotely working PSU; I just want to keep it for the sake of being able to quickly swap to test if my other PSUs ever have a problem.

Gonna have to try figuring out how to underclock this CPU exactly, now...

 

[TheSonikku]

Regardless of the cause those spikes can kill Motherboard components and peripherals. Voltage rectifiers help stabilize but your spikes are way over the top.
I would cease using your system till you check your power at the wall socket. The PSU can work for two strait days but is that under load?

No, it wasn’t under load; but at my place the MOBO-indicated voltages barely last 1-2 hours without being under load before they go above 6...

 

[TheSonikku]

I’ve been using the CPU-MOBO combination for well over 5 years now (I actually once even had the MOBO replaced due to other complications, but it was still the same one), and only recently after installing a new cooler did stuff like this start to happen.

While I do agree the other PSU is pretty bad too, it’s still a remotely working PSU; I just want to keep it for the sake of being able to quickly swap to test if my other PSUs ever have a problem.

Gonna have to try figuring out how to underclock this CPU exactly, now...

ACTUALLY I just realized I made a mistake; I have an FX8320, not 8350.

 

[TheSonikku]

Update: 12V also now displays it as 10.48V.

 

[Grobe]

Update: 12V also now displays it as 10.48V.

Probably bad PSU.

 

[TheSonikku]

Probably bad PSU.

Even though both PSUs had the same voltage symptoms, ignoring the fact the other brand new one is off-brand?

 

[Grobe]

Probably - not same as obviously. There are two most probably possibilities :

• Failure on multiple voltage sensors, makes it as less possible. However - there may be poor design that if certain parts fails, it affect more than one measurement. That can only be answered if given access to the full pcb schematic and spending time to study it.
• Failure on PSU. When you say 5V goes up and 12V goes down, that can be caused by either a psu failure - or some other component that exceed the maximum ampere rating of the psu. This is because the 12V and 5V is normally shared by the same internal transformer, yet only one feedback channel (normally 12V). That gives the two possibilities - either the PSU feedback is faulty, or some component on 12V is faulty and make a current rush that exceed the psu max rating.

Or - there is an error in an internal A/D converter that make the measurement possible to be represented as numbers. However, this is very rare and probably not the cause.

 

[TheSonikku]

Probably - not same as obviously. There are two most probably possibilities :

• Failure on multiple voltage sensors, makes it as less possible. However - there may be poor design that if certain parts fails, it affect more than one measurement. That can only be answered if given access to the full pcb schematic and spending time to study it.
• Failure on PSU. When you say 5V goes up and 12V goes down, that can be caused by either a psu failure - or some other component that exceed the maximum ampere rating of the psu. This is because the 12V and 5V is normally shared by the same internal transformer, yet only one feedback channel (normally 12V). That gives the two possibilities - either the PSU feedback is faulty, or some component on 12V is faulty and make a current rush that exceed the psu max rating.

Or - there is an error in an internal A/D converter that make the measurement possible to be represented as numbers. However, this is very rare and probably not the cause.

As for the components, could a CPU fan possibly cause that? I didn’t have such a problem before the CPU fan was installed.

 

[jasonf2]

I have worked in industrial environments with extremely dirty power. Most real power issues are transient, harmonic related, or sag/swell issues. If it is an input power issue the only real cure at a reasonable price is a full online reconstructive UPS and it still isn't cheap. What you are describing doesn't sound like input power issues unless it is a ground issue. With your fluctuation on multiple power rails the first thing I would do is verify a good ground and try a different PSU. Others will argue with me on sizing, but I suggest going slightly over sized and get a good one (I suggest titanium). Also if you just replaced a fan and things started going wonky replace the fan with a new one. Fans have a nasty habit, especially on poor power supplies, of over pulling power and lowering rail voltages. Digressing back to the ground thing. While a ground cannot be used for a current carrier it is used for a 0 volt reference and if it is disconnected or very marginal it will cause the voltage on a power supply to float around without a good reference.

 

[TheSonikku]

I have worked in industrial environments with extremely dirty power. Most real power issues are transient, harmonic related, or sag/swell issues. If it is an input power issue the only real cure at a reasonable price is a full online reconstructive UPS and it still isn't cheap. What you are describing doesn't sound like input power issues unless it is a ground issue. With your fluctuation on multiple power rails the first thing I would do is verify a good ground and try a different PSU. Others will argue with me on sizing, but I suggest going slightly over sized and get a good one (I suggest titanium). Also if you just replaced a fan and things started going wonky replace the fan with a new one. Fans have a nasty habit, especially on poor power supplies, of over pulling power and lowering rail voltages. Digressing back to the ground thing. While a ground cannot be used for a current carrier it is used for a 0 volt reference and if it is disconnected or very marginal it will cause the voltage on a power supply to float around without a good reference.

Do you think disconnecting the fan, opening up the case and blowing the inside with a bigger fan would be fine to test if the voltages stabilize?

Also, again, my screen shuts off for a second whenever stuff in my room is turned on/off. Maybe it is grounding issues?

I was using the PC the entire day today and the voltage was locked at 7.65V for +5V and 10.48 for +12V. Or, again, so it says.

 

[jasonf2]

Do you think disconnecting the fan, opening up the case and blowing the inside with a bigger fan would be fine to test if the voltages stabilize?

First off disconnecting the CPU cooler fan will cause your machine not to post, and for a good reason. So you would have to bypass the RPM monitor in bios to get it to boot. While you might get away with it it is also going to at the very least overheat and throttle your cpu and the process of booting a pc is almost as intensive as running a video game for just a short time. Personally I would wait until you have a new fan and lean towards the safe side. I don't think you are probably going to hurt anything as long as the heat sink is still attached but I also wouldn't risk it and at best it is going to run poorly as it thermal throttles.

 

[TheSonikku]

First off disconnecting the CPU cooler fan will cause your machine not to post, and for a good reason. So you would have to bypass the RPM monitor in bios to get it to boot. While you might get away with it it is also going to at the very least overheat and throttle your cpu and the process of booting a pc is almost as intensive as running a video game for just a short time. Personally I would wait until you have a new fan and lean towards the safe side. I don't think you are probably going to hurt anything as long as the heat sink is still attached but I also wouldn't risk it and at best it is going to run poorly as it thermal throttles.

Well, I did have it running for a week with only the case open and a big fan. (last CPU fan died, stopped spinning) and i bought a cooler master one. I couldn’t do anything other than open at most 2 programs at once while constantly monitoring the temperatures so they didn’t go over 80C, but idle it stayed at around 72C. I didn’t have any crashing problems before that (well, aside from crashing due to high temperature). Also, could you check my last post (I edited it) to see if anything maybe influences what I should do?

EDIT: Also, would messing with the fan speed possibly change anything in the voltages if they’re related?

 

[Grobe]

EDIT: Also, would messing with the fan speed possibly change anything in the voltages if they’re related?

Only if the fan is faulty or if the rpm is set too low.

 

[jasonf2]

Well, I did have it running for a week with only the case open and a big fan. (last CPU fan died, stopped spinning) and i bought a cooler master one. I couldn’t do anything other than open at most 2 programs at once while constantly monitoring the temperatures so they didn’t go over 80C, but idle it stayed at around 72C. I didn’t have any crashing problems before that (well, aside from crashing due to high temperature). Also, could you check my last post (I edited it) to see if anything maybe influences what I should do?

EDIT: Also, would messing with the fan speed possibly change anything in the voltages if they’re related?

Your core voltages should be stable no matter what you are doing with the computer. All of my prior recommendations still stand. You can try the fan and see what happens, but do so at your own risk.

 

[TheSonikku]

Update: Used the PC for around 2 weeks with no problem, voltages were reading randomly between 5.07-7.6V and 10.48-12.1V; Usually it started off at 5-5.1 and rose to 7, but sometimes it started at 7.

Now, 2 weeks later, PC crashed again. Complete power-out, and now the PC turns on but no beeping; leaving it on makes it turn off after a while. Stuff connected seems to be fine, though.

 

[TheSonikku]

Update: PC wasn’t turning on this morning; tried to turn it on in the afternoon and it didn’t load but said “Overclock failed”; Went into the BIOS and set underclock from -2% to 0%; PC’s voltages for the rest of the day were between 5.07-5.16V and 12.1V on the program and the PC didn’t turn off at all.