HWinfo64 shows abnormal readings for PSU +12V rail

[The N]

I have noticed that PSU (SilverStone Strider SST 650W) readings in HWinfo64 has been abnormal lately. Today, while playing game (Ghost of Tsushima) HWinfo is showing reading for +12V at 12.048V (max) and 11.328V (min) as seen in the picture, later it further dipped to 11.148V (min) after an hour of gaming, which is super abnormal.

I understand that software readings, specifically, PSU related are not entirely accurate. And in my case, readings are way off which compelling me to buy a new PSU; I have a few options in mind to get from. However, I am not sure if PSU is faulty, or HWinfo is way off on PSU's voltage segment as I can play any game on max settings (that my GPU allowed me to), or stress PSU by using CPU/GPU load testing programs with no sign of failure/crash/shutdown whatsoever. At the time of writing, I am playing Ghost of Tsushima.

Need opinions/guide on subject matter, please.

My PC Specs:
12600K at stock boost
Noctua NH-U14S
MSI Z690 Edge
DDR4 32GB RAM
Asus Strix RTX 3060Ti 8GB
2x SSD
2x HDD
SilverStone Strider SST 650W (ST65F-PT) PSU In Question
MSI Optix MAG Series 27" 1440p Monitor

 

[ShrimpBrime]

Some will say that you have a 10% swing +/-, so you're close to calling the PSU junk.

If you have a multi-meter, probe the CPU EPS connector to verify the readings. Don't assume the software is 100% correct.

 

[freeagent]

Normally I would say it is literally time to replace it but.. it is working pretty hard, and you're 3.3 and 5v looks pretty tight.

I can do 650w no problem with my 5900X in my system.

 

[DR4G00N]

It's likely just a bad 12V connection in the 24-pin, try re-seating it or move the cable slightly and it'll likely come back up. Judging by how the 12V In for the CPU VRM is pretty solid ~12.1-12.2V I'm guessing this is the case.

This happens with my PSU from it being old and worn, if the 24-pin is not in exactly the right spot the 3.3V will droop below 3.2-3.1V and start causing issues.

 

[Macro Device]

you're

How come they allow moderators to do that..?

It's likely just a bad 12V connection in the 24-pin

Yeah, also had anomalies in my client's rig because the guy was too afraid to jank it and didn't push the cable deep enough.

 

[freeagent]

It's likely just a bad 12V connection in the 24-pin, try re-seating it or move the cable slightly and it'll likely come back up. Judging by how the 12V In for the CPU VRM is pretty solid ~12.1-12.2V I'm guessing this is the case.

This happens with my PSU from it being old and worn, if the 24-pin is not in exactly the right spot the 3.3V will droop below 3.2-3.1V and start causing issues.

My X58 system was really bad for that.. well worn.

 

[SPDIF]

You can't really trust that software reading, you should measure it with a REAL Multimeter to be sure.

 

[R-T-B]

I've never taken any stock in software readings here. If it isn't crashing, its probably within ATX spec, and not worth worrying about.

 

[eidairaman1]

Some will say that you have a 10% swing +/-, so you're close to calling the PSU junk.

If you have a multi-meter, probe the CPU EPS connector to verify the readings. Don't assume the software is 100% correct.

ATX SPEC tolerance is + or - 5% so 12V rail minimum is 11.4V, max is 12.6. 5 is 4.75, 5.25. 3.3 is 3.135, 3.465.

Id be checking the ATX, EPS 12V, pcie etc.

 

[AsRock]

Keep a eye on it, and if you start getting system shutdowns then you might of already found the issue.

 

[eidairaman1]

Keep a eye on it, and if you start getting system shutdowns then you might of already found the issue.

11.3 is below the atx spec

 

[AsRock]

11.3 is below the atx spec

Yup, and last time i had such readings Seasonic ask me to send it in, how ever i was having shutdowns too.

It could be just a false reading, although you could all so tel HWinfo64 to save to file and keep a eye on the 3.3v, 5v and 12v rails and see if that see/s some thing if a crash does happen.

 

[PassportBro]

Here's my experience.

Way back in like 2009 I bought a PSU from a local store. It was a Hiper HPU-5K880 880W PSU. Big mistake.

The PSU eventually blew up while inside my rig. Before it did so I did notice abnormal 12V readings in HW Monitor.

12V readings usually hovered around 11.3-11.7V as per the screenshots I have saved.

Replace immediately.

 

[ShrimpBrime]

ATX SPEC tolerance is + or - 5% so 12V rail minimum is 11.4V, max is 12.6. 5 is 4.75, 5.25. 3.3 is 3.135, 3.465.

Id be checking the ATX, EPS 12V, pcie etc.

Right 10% +/- meaning +5% and -5% from 12v.

Wonder if the EPS cable reseat did the trick or if there was anything done otherwise. Hope OP comes back let us know.

 

[freeagent]

I have injured 5 or 6 PSU's, I thought this was injured too, not yet.. but it will be.

Software can be trusted if your board is halfway decent.

 

[dirtyferret]

I've never taken any stock in software readings here. If it isn't crashing, its probably within ATX spec, and not worth worrying about.

this,
on the old JG forums, software was never considered an accurate measurement and I had a few PSU (from well known OEMs & brands) that showed they were out of spec yet ran fine. If you want an accurate measurement you need a multimeter power supply tester.

 

[SPDIF]

It's easy to measure under load with a normal multi-meter and it stays the best way to know what is wrong. If possible you can even try another PSU from with you know is good.
That software reading you should take with a big grain of salt, sometimes you can have more then 10 percent deviation. It's really up to the quality Mobo you have.

The best way stays a good multi-meter! You have special PSU tester's out there, but that is more for a computer shop.

 

[dirtyferret]

I have injured 5 or 6 PSU's

How many times have I told you to stop using a baseball bat to fix your PC

 

[The N]

Sorry guys, I was at office as per my routine. So thought to share details once I get back home and check thoroughly what suggested by all of you.
I appreciate the responses from all of you.

So I unplugged PSU cables from MB and while I was probing the 24-pin Cable with multimeter that I got, It gave me,
+12V = 12.34
+5V = 5.14
+3.3V = 3.39.

Just to make sure the stock voltages are fine.

Upon replugging cables (24-pin, PCIe and CPU EPS connectors on MB), I made sure the PSU connectors have a good contact with the base connectors on MB as what @DR4G00N @ShrimpBrime have suggested up there. GUESS WHAT! This does the trick for me for real. Voltages are much better now on +12V rail or I say, drastically improved. See pic

Well, I had no idea if re-seating could create such a huge difference.

Thanks a lot, guys. I think I can stay with this PSU until GPU upgrade (RTX 4070 Ti/Super).

It's likely just a bad 12V connection in the 24-pin, try re-seating it or move the cable slightly and it'll likely come back up. Judging by how the 12V In for the CPU VRM is pretty solid ~12.1-12.2V I'm guessing this is the case.

This happens with my PSU from it being old and worn, if the 24-pin is not in exactly the right spot the 3.3V will droop below 3.2-3.1V and start causing issues.

 

[SPDIF]

That means a bad contact with made a resistor so voltages go down, this resistor is heating up the bad contact and is the reason of burning connectors, aka 4090 video card's.

You should clean the contacts better with IPA and then connect them again.

 

[ShrimpBrime]

Sorry guys, I was at office as per my routine. So thought to share details once I get back home and check thoroughly what suggested by all of you.
I appreciate the responses from all of you.

So I unplugged PSU cables from MB and while I was probing the 24-pin Cable with multimeter that I got, It gave me,
+12V = 12.34
+5V = 5.14
+3.3V = 3.39.

Just to make sure the stock voltages are fine.

Upon replugging cables (24-pin, PCIe and CPU EPS connectors on MB), I made sure the PSU connectors have a good contact with the base connectors on MB as what @DR4G00N @ShrimpBrime have suggested up there. GUESS WHAT! This does the trick for me for real. Voltages are much better now on +12V rail or I say, drastically improved. See pic

View attachment 366478

Well, I had no idea if re-seating could create such a huge difference.

Thanks a lot, guys. I think I can stay with this PSU until GPU upgrade (RTX 4070 Ti/Super).

I gotta give the credit where due. That was all @DR4G00N. I only suggested to check with a multi-meter.
Thanks dude!!

 

[The N]

I gotta give the credit where due. That was all @DR4G00N. I only suggested to check with a multi-meter.
Thanks dude!!

Indeed, it was originated by @DR4G00N, thanks man. Sometimes little things create big difference.

 

[ShrimpBrime]

Indeed, it was originated by @DR4G00N, thanks man. Sometimes little things create big difference.

It's not always the case. I had an old 12 years PSU giving the same readings across several boards. I decided to junk it before it did become a problem. Multi-meter helped confirm it. Get a load and that 12v rail would droop. Actually within that 10% swing, but I don't like seeing less than 12v ever. At least not from quality power supplies anyways.

Glad it's fixed though!!

 

[The N]

That means a bad contact with made a resistor so voltages go down, this resistor is heating up the bad contact and is the reason of burning connectors, aka 4090 video card's.

You should clean the contacts better with IPA and then connect them again.

Yep, bad contact was the issue here but I just replugged with little force and it worked. I guess IPA cleaning not needed at this stage.

It's not always the case. I had an old 12 years PSU giving the same readings across several boards. I decided to junk it before it did become a problem. Multi-meter helped confirm it. Get a load and that 12v rail would droop. Actually within that 10% swing, but I don't like seeing less than 12v ever. At least not from quality power supplies anyways.

Glad it's fixed though!!

I am still unsure regarding multi-meter testing at load. I have that basic two-wired multi-meter bought from local electric shop. Can I check voltages when system is running?

 

[ShrimpBrime]

I am still unsure regarding multi-meter testing at load. I have that basic two-wired multi-meter bought from local electric shop. Can I check voltages when system is running?

Yes of course you can. Set the multi-meter to DC and probe positive negative on the CPU EPS connector which is easier than probing the 24pin motherboard connector. If it's a single rail PSU, it won't matter where you probe.