• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Corsair 850HX voltage readings

strep3241

strep3241

Senior member
I believe this has been happening since day one. I have used a couple different programs to monitor the voltage readings and they are the same. The +12 volt rail shows 3.5 at idle and jumps up to 12 or 13 under load. The +5 VCCH volt rail shows 3.6 all the time.

I have been having some small problems mainly due to overclocking and I wonder if this could be the cause. Is it possible this is a false reading? If so, why does the +12 volt rail jump up under load?

Now the bios shows the correct readings.

I could install this in another computer to test it out. But what if I do this and it still shows the same readings? Does that confirm that the PSU is bad or going bad?

I would like to test it with a volt meter before I try it in another computer. How do you do this? Do I have to go through the 24 pin connector or can I go use one of the molex connectors?
 
Your computer wouldn't even turn on if the 12V rail was at 3.5V, so yes it is a false reading. Try a multimeter. All the programs you use obtain reading from the same chip on the motherboard, all the programs you use are inaccurate.
 
Someone else told me that as well.

I used a volt meter with a paper clip shorting the two pins and I got normal readings. But that is with just the case fans running and nothing else.

If this is a false reading, why does the +12 volt reading jump up when under load? I would think it would stay the same if it was a false reading.

Do you think it would be worth it to try and install this PSU in another computer? What if I get the same readings?
 
HOOfan 1 said:
Your computer wouldn't even turn on if the 12V rail was at 3.5V, so yes it is a false reading.
Click to expand...

Isn't that the truth. That'd kinda be like getting a car to run on gas that has only 1/3 the octane rating.....something like 29. How anyone can think a computer can function on 1/3 its required voltage is beyond me.



memoryram said:
Someone else told me that as well.

If this is a false reading, why does the +12 volt reading jump up when under load? I would think it would stay the same if it was a false reading.
Click to expand...

Maybe the software is written badly and doesn't really detect the voltages properly until an amperage level is reached.....or could be a myriad of other things causing the obviously incorrect readings.



Rifterut said:
Im getting real sick of saying software is inaccurate use a multi meter/volt meter.

OP software programs are not accurate, the ONLY way to get correct results is through manually using a meter.
Click to expand...


Yep, you can say that until you're blue in the face, yet you see these questions weekly. And even when give the correct answers, you still get "Are you sure 'cause I heard...." in return.

Maddening, isn't it?
 
Meghan54 said:
Isn't that the truth. That'd kinda be like getting a car to run on gas that has only 1/3 the octane rating.....something like 29. How anyone can think a computer can function on 1/3 its required voltage is beyond me.





Maybe the software is written badly and doesn't really detect the voltages properly until an amperage level is reached.....or could be a myriad of other things causing the obviously incorrect readings.






Yep, you can say that until you're blue in the face, yet you see these questions weekly. And even when give the correct answers, you still get "Are you sure 'cause I heard...." in return.

Maddening, isn't it?
Click to expand...

Yep...my brother's girlfriend's sister's hairdresser knows a guy who heard from his brother's wife's uncle who got drunk with a guy whose brother was once told by a guy who read it in some magazine...
 
I checked the voltage at idle using the 4 pin molex connector. I received the normal readings. So I guess it is safe to say that is a false reading.

What about the +5 VCCH reading, how can I check that one and what is that?
 
Meghan54 said:
Yep, you can say that until you're blue in the face, yet you see these questions weekly. And even when give the correct answers, you still get "Are you sure 'cause I heard...." in return.

Maddening, isn't it?
Click to expand...

Yeah, especially when there's a sticky on it.

Seems to be a number of subjects of frequent worry.

"Why does Windows only report 3.25GB RAM?"

"Why does my SSD benchmark low?"

"Why does the owner's manual claim I can go 7k miles w/o oil change when everyone knows you should change at 3k miles?"

The latest one is:

"OMG Sandy Bridge overclocks like crap! I can't get BCLK past 105MHz!"

gotendbz1 said:
memoryram are you testing or benchmark. what the reason for this indepth testing ?

is the psu acting funny?
Click to expand...

My guess (100% unsubstantiated) is that he checked it just because the monitoring software could, then saw what was obviously wildly out of spec numbers, and thus became worried.
 
memoryram said:
I believe this has been happening since day one.
Click to expand...

Are you actually having issues? Or, are you wanting to perform these checks due to the software readings alone?

HOOfan 1 said:
Your computer wouldn't even turn on if the 12V rail was at 3.5V, so yes it is a false reading. Try a multimeter. All the programs you use obtain reading from the same chip on the motherboard, all the programs you use are inaccurate.
Click to expand...

Well said.

Rifterut said:
OP software programs are not accurate, the ONLY way to get correct results is through manually using a meter.
Click to expand...

Well said again.

Meghan54 said:
Maybe the software is written badly and doesn't really detect the voltages properly until an amperage level is reached.....or could be a myriad of other things causing the obviously incorrect readings.
Click to expand...

A lot of the software is poorly written. Additionally, have you ever noticed that sometimes temp readings will change after a BIOS flash? The software readings depend on info received from whatever sensor they are programmed to read. These sensors have to be calibrated and this is done in the BIOS. So, if the calibration is off any little bit, the software can report VERY inaccurate temps.

I had a guy at our forums once that debated me for several weeks, threatened class action lawsuits, bought a Yellowbeard voodoo doll and stuck pins in it, etc etc. He was absolutely enraged beyond reason that I told him that his software was wrong. He insisted that it was our product that was at fault. He refused any help I offered for the first week or 2 including updating his BIOS since temps weren't noted in the BIOS release notes.

After about 2-3 weeks of his tirades, he finally grudgingly admitted that he had an older version of the monitoring program that had not been updated for his MOBO. Updating the software resolved most of his computer (not mental) issues. Also, his readings changed AGAIN after a BIOS flash. But, he still never apologized for his abuse he dished out and still clung to the idea that we were somehow responsible.
 
Pisses me off how everyone somehow believes a reading based off a cheap software sensor somehow substitutes for a real multimeter reading for a psu. STOP READING SOFTWARE VOLTAGE.
 
why does this piss everyone off? there is no excuse... simply tell the person to use a multimeter... if you are pissed off it is you who has allowed it to do so
 
I have been having issues mainly when overclocking. I can't seem to get it stable. I thought the PSU might have been the problem since the voltage readings were low. Well I ruled the PSU out since I checked voltages manually and they are reading fine.

It may not be hardware problem, it could be that I don't have the settings right. Even at a mild overclock of 3.2gz, I can't seem to get it stable. I can pass IBT and Linx but sometimes when I leave it for a while and come back to it, it has restarted with a message saying "windows has recovered from an unexpected shutdown" or something similar. I am almost positive it has to do with the memory settings. I have tried to run it at 1066mhz or close to that. When I try to run it at 1600mhz, it doesn't want to run at all.
 
Last edited:
@ tomoyo and Rifterut - how then did you both get your v-core readings, which you both have in your sig and if you's used a multi-meter for that, could you please explain to the OP and myself exactly how you do this

tomoyo = "Main: Core2 E6600 @ 3.0ghz(1.2v core)w/SI-128,"

Rifterut = "i7 930 4200Mhz, 1.35v"

@ OP - have you ever used OCCT with any of your hardware for testing stability?
 
Last edited:
BTRY B 529th FA BN said:
@ tomoyo and Rifterut - how then did you both get your v-core readings, which you both have in your sig and if you's used a multi-meter for that, could you please explain to the OP and myself exactly how you do this

tomoyo = "Main: Core2 E6600 @ 3.0ghz(1.2v core)w/SI-128,"

Rifterut = "i7 930 4200Mhz, 1.35v"

@ OP - have you ever used OCCT with any of your hardware for testing stability?
Click to expand...

Perhaps you are reading a differnt thread? this thread is about PSU voltage rails not CPU voltage. And as far as i know its impossible to use a multi meter on most modern boards with heatsinks over the VRM's because they block access to the surface mount electronics you would need access to in order to take a reading with a meter directly off the board(which is possible if you had the blueprints for the board and the heatsink(s) were not in the way).

Removing the VRM heatsinks for access while running a 180-200bclk under load overclocked on a chipset designed for 133bclk is NOT something i would reccomend to anyone, and if not overclocked under load whats the point in taking the reading. Also without a electronics background it would be very easy to damage the small components you will need to measure from for CPU voltage. This is totally 100% not in the same class as sticking a probe into a molex connector for PSU readings.

I have taken readings directly off motherboards in the past but this was long before the bios/software was written to support voltage readings and long before motherboards were pushed hard enough to require giant heatsinks on the chipsets and VRM's. Think back in the days where you adjusted everything with jumpers, FSB, Multi, Ram divider, CPU voltage, RAM voltage.

here is a link to an example of a old motherboards voltage sensing locations on a old X48 DFI board, note how most of those CPU voltage locations would be under heatsinks on modern boards.

http://www.images.clunk.org.uk/reviews/dfiltx48t2r/voltage-points/new/IMG_2044voltagepointskey.png
 
Last edited by a moderator:
BTRY B 529th FA BN said:
Let me ask again. How then did you get your v-core readings.
Click to expand...

Through the bios as well as CPUz and the asus utility(pc probe, which BTW seems to be the least accurate). They all agree + or - 5%. Well they may not be 100% accurate they are the only feasable methods of obtaining them. I really wish there was a way to monitor off the board like the old days but in this day and age of huge CPU/mobo heatsinks its just not possible.
 
Uh v-core is a setting you control in the bios. It's not 100% precise either, but it's something you are actively choosing and actually controlled by the mobo itself. This is completely different from picking up voltage information from an external device such as a psu using a mobo voltage indicator that almost all software reads incorrectly, let alone the fact that it's generally very imprecise even when read correctly. My v-core number is the setting in the bios, not the actual v-core itself in practice, I have no interest in trying to represent that.
 
memoryram said:
My 920 stepping is D0. Oops, that is the revision, the stepping is 5.

And sorry if I made some people mad. I just glanced right over the sticky.
Click to expand...

I think your system is an example why overclocking is never guaranteed. I've seen a number of examples of motherboards and cpus that just don't overclock right and do terribly at stability even with a very small overclock. In some combinations it just doesn't work. For when you have a good psu already, it's clearly some other issue. If you really want to oc, you need to try some other mobos/cpus and also possibly unplug some devices to isolate the issue.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top