Voltage Regulation We will now have a look at a couple things related to voltage and voltage regulation. One common question in forums we see is that of "why does my voltage show 1.90v when I set 1.95v?". A lot of people don't understand that what you set in the BIOS isn't an exact reflection of what the board is going to be supplying. We will be measuring the voltage for a couple components with a calibrated UEI DM393 and comparing those readings to what is set in the BIOS, what the software reports, and discuss any major differences. We will then take a quick look at the vDROOP of the board and ASUS's Loadline Calibration option in the BIOS. First up are the read points we used for the testing.  
 Starting in the upper left you can use pretty much any of the leads closest to the socket for measuring voltage, either from the capacitors or the inductors as circled. The MCH voltage was a bit of a guess as there is no access to the inductors because they are surface mount and the large heat sink covers them. We found the circled ends of the capacitors in the upper right photo to provide what we figured were the accurate vMCH voltage. The third photo is of the vDIMM measure point and again is a surface mount capacitor. These capacitors can be found all up and down the DIMM slot wherever there is a voltage providing pin pretty much. These cap readings are bang on to what the actual DIMM slot pin provides, just much easier to read from on a running system. Here is the chart we came up with for volts set versus load and idle in windows using the various methods of outputs available to us. | BIOS Set | BIOS Report | PC Probe II
Idle | PC Probe II
Load | DMM
Idle | DMM
Load | | vCORE | 1.375v | 1.330v | 1.34v | 1.30v | 1.354v | 1.354v | | VTT | 1.35191v | 1.343v | 1.34v | 1.35v | x | x | | vPLL | 1.56406v | 1.561v | 1.56v | 1.56v | x | x | | vMCH | 1.51097v | 1.488v | 1.49v | 1.48v | 1.508v | 1.508v | | vDIMM | 2.05431v | 2.050v | 2.05v | 2.04v | 2.070v | 2.071v |
For the most part the BIOS set values are pretty accurate. On the whole, the software readings were a bit low, and the digital multi-meter readings a bit higher. vDIMM over volts the very slightest amount but is nothing to be too worried about and with Loadline Calibration enabled we get little to no vDROOP under load as we would expect. If you are concerned about vDROOP, simply enable Loadline Calibration and you should be good to go. Here are a couple charts from OCCT of our max stable overclock with Loadline Calibration enabled, and disabled. vCORE vDROOP Chart from OCCT - Loadline Calibration Disabled
 | vCORE vDROOP Chart from OCCT - Loadline Calibration Enabled
 |
As we can see, without LLC enabled, we get a solid 0.04v droop under load and with LLC enabled we get a relatively steady line across the two hour test. There is some chatter at a couple different spots during the testing period with LLC enabled but nothing to really be concerned with. This option is becoming an almost necessity since so many users are witnessing the phenomenon known as vDROOP in their setups. In reality, vDROOP is built into the Intel specifications for processors and the debate will ensue on whether disabling it is a good, or bad thing.
| 
