[silverrabbit]

I have a puzzling problem. When running Speccy or HWMonitor, the voltage shown for the +12V buss is around 8.9 to 9.0 volts. On my other PCs, the buss is running at around +12 +/- 0.1V. I thought it might be the PS (One year-old Cooler Master M2 Silent Pro 850W), so I replaced it with an FSP Raider Silent 450W (the currently installed PS), and the symptom remains the same. I have attached the HWMonitor report for the PC.

Has anyone encountered this before or understand this?

Many thanks...
Silverrabbit

 

[Aus_Karlos]

Welcome to TSF,

Software voltage monitoring is never accurate as you can tell. If your PC was really running at 9v on the 12v rail it would of more than likely crashed or not even booted in the first place.
The voltage senors on most motherboards are an afterthought and are usually very cheap chips.
Usually the BIOS has better readings but they still can be off by a decent factor. The only way to know for sure is to use a voltage multi-meter set to DC and connect it to the 4 pin power connector.
With the four pins there are multiple voltages so you can test them all if need be.
The black wire is always the ground/neutral
The Red is +5v
The Yellow is +12v
The Orange is +3.3v

Make sure the multimeter is set to DC voltage which is a V with a solid and/or dotted line underneath. Place the black lead from the meter into the pin hole of the black wire first, this will make sure if the red lead touches anything else it wont short out the PSU.

Good little guide is here.
Power Supply FAQs

 

[silverrabbit]

Thank you, I'll try that tomorrow on the oscilloscope. I'll share my results here.

 

[Bill_Bright]

I generally blame the sensors or the chipset coding interpreting the data from the sensors before blaming the hardware monitoring software. The sensors are read by the motherboard/chipset and the software simply gets that information from there.

The mere fact both Speccy and HwMonitor are showing the same 9V confirms that it is not a software issue.

What does it say the voltage is in the BIOS Setup Menu?

I would not call the sensors "afterthoughts", but they certainly are low-tech and cheap - that's cheap as in inexpensive, as opposed to poor quality or likely to fail or fall apart prematurely.

Your power supplies are not putting out 9V or your motherboard just would not function. The ATX Form Factor standard allows for just ±5% tolerance on the +12VDC rail meaning the +12V must not drop below 11.4VDC or go above 12.6V. Motherboard regulator circuits can compensate somewhat, but I doubt they could for a 25% out-of-tolerance condition. Certainly not without some extreme stress and thus excessive heat situations.

And hard drive motors would have a very difficult time spinning up too - if able to at all.

So your PSUs are putting out close to 12V or else your computer would not be booting up (and you confirmed this by trying 2 different supplies).

I suspect this is due to a simple coding error in your BIOS firmware used to calculate/interpret the readings from the thermal diode sensor. The sensor could be bad, but typically a bad sensor gives no reading, or the reading jumps all over the place and is not consistently wrong by the same amount.

I recommend you check your motherboard's website for a BIOS update. If none, and your temps are all okay, I would not worry about it.

Oh, and just FTR, that's bus, not buss.

 

[silverrabbit]

AUS_Karlos & Bill_Bright: Thank you to both of you for the information. When I (duh) put my o'scope leads to the 12V supply, it is 12.05 with about .01 harmonic distortion on top of the 12.01 signal, so everything is OK with the current PS, and I expect the same for the one that I replaced.

Looking of the good side, I now have an extra PS that is silent when the one that I am using goes bad. (silent is needed here as there are multiple processors in the same room and I hate the fan noise after working around IBM mainframes for over 35 years.)

Again, many thanks,

Bob

 

[Bill_Bright]

(silent is needed here as there are multiple processors in the same room and I hate the fan noise after working around IBM mainframes for over 35 years.)

I hear you wink: pun intended).

I spent many years myself in equipment rooms with raised floors and the noise truly is deafening. And not just from the ventilation noise, but from everyone shouting to be heard over it!

Obviously, 12.05VDC is well within acceptable tolerances. Since you have a scope and know how to use it, if you are curious, the maximum ripple allowed on the +12V rail is 120mV peak to peak and 50mVpp on the 3.3 and 5V rails. When measuring, outputs should be bypassed at the connector with a 0.1µF ceramic disk capacitor and a 10 µF electrolytic capacitor to simulate system loading. However, in most technical reviews, exceeding 50mVpp on the 12V rail and 30mVpp on the others would be considered excessive for a quality supply. For more info and test setup, see Figure 1 on page 22 of the ATX Power Supply Design Guide.