Hey dinos mate,
do you know the VDIMM READ Point?
I have some strange values in Everest and I want to verify that
pretty sure i was reading it off one of the points next to that 24-pin power connector
just poke one of the two there and you will get a reading i'm sure![]()
this is what i have been using for vDIMM. it scales with BIOS selected and with the actual voltage measured at a VDD pin of the DIMM slot. i generally just measure straight from the DIMM slot but with the modules in the black slots, it is impossible to get at. BTW, VDD pins in the DIMM slot report exactly what the BIOS monitor reports and is just slightly lower than selected when using the board mounting grounds as ground.
i am in the middle of a short HCI run but once it is done, i will put together a little chart.
i am also curious about Everest reporting vDIMM. what board is that noobzed? i don't get anything but vCORE in Everest pretty much.
i will also be swapping out the stock cooling on my board this weekend so at that time i will do some of the vmod testing Dinos has posted and get some confirmation for us reference board kids![]()
yup, more users will probably find similar results but i doubt all boards will be exactly the same. just measuring a VDD output pin in the DIMM slot confirms other read points...just be careful when doing so as it is easy to short DIMM pins.
another key item is the ground location. different ground spots can skew readings up to 0.03v and more in my experience. i always ground at the motherboard mounting locations when taking DMM readings.
next time your taking a measurement, do the HD mount hole and then do the MB mount hole. i just took the vDIMM for my dual32M 1020MHz...
HD hole = 1.983v
MB hole = 1.964v
the difference obviously isn't huge but it is there. i found that the MB ground appears to be 'most accurate' based on readings from GPUs. you see, we are aware of what volts are suppose to be for GPUs and i always found that GPU readings were bang on with what they were suppose to be when grounded at the MB mount holes. everything from 8800GTS 512MBs to my 3870X2.
clearly not a technical explanation but i can't explain it, i solder, not design the electronics![]()
my DMM is not good enough to read as many digits so it doesn't bother me a heck of a lot hehehehe
Has anyone had luck with the vdroop mod? I've talked to a couple people that tried without success. I'm tempted to try myself...but would at least like to talk to someone that's had it work.
Thanks dinos, you rock!
ok 100K is better in that case just in case
but 50K will have finer adjustments``
alright, took the afternoon for some quality time with the EVGA 790i and can confirm the vdroop mod on the first page Dinos posted is correct and working with a 50kΩ VR. i believe a smaller VR could be used because the 50kΩ VR needs to be tuned waaaay down before vCORE starts to raise under load to what it is at idle. just don't blame me if you have problems using a 20KΩ VR, i am just speculating.
being typical me, i forgot to write down what the resistance is across the resistor without the mod but i thought it was 2382Ω. with the 50kΩ VR soldered across the resistor and the VR tuned to 50kΩ, the resistance is 2272Ω. i found the vCORE to start raising under load at around 1750Ω and to receive close to a flat line for vCORE, resistance of the circuit needed to be down to 600Ω. after that vCORE starts going up at idle and load.
i don't know what the value of the VR is at that point because i don't feel like de-soldering a lead and re-soldering it but at some point i might. either way, here are the quick 10 minute before and after charts from OCCT...
before:
after:
and of course, a couple photos of the mod:
of course, this doesn't eliminate vdrop where what you set in the BIOS equates to about 0.035v less for actual voltage so don't complain that you are still getting vdroop after the mod...droop is simply difference between idle and load...nothing to do with what is set in the BIOS
i will mention that this mod is quite tricky unless you have very thin wires. i always use floppy cable strands as it is single stranded and easy to work with. unfortunately the insulation is rather thick and this posed a problem with this mod because the width of the resistor is so small that the two wires were quite a bit wider. notice how the insulation on the two wires are almost sitting on top of each other? i don't like to come from the top because they you can't insulate over it, so i always come from the side. if i had to do it again, i would simply separate the two wires further back and come from a different angle on one side of the resistor.
i forgot to bring home my tools from work on friday so i did this the old fashioned way, no magnifiers, proper tweezers, or picks. i would have also grabbed some wire wrap wire had i known it was this tight for the mod...it would have been a lot easier with the correct tools and wire.
in addition to the mod, i also took a whole set of photos of the board as well as changed out the stock cooling. while i was at it, i found a few more voltage read points and will post those soon enough. just have to get some work done before i goof around any more today![]()
No READVCORE READOUT(haven't checked yet >> someone pls check and let us know)
i have a capacitor where Dinos pic is showing just pads.
with that said, i don't get a reading there either. on the opposite sides of the caps i am getting my 12v reading from both, but nothing from the side closest to the socket.
shouldn't we just be measuring the R47 leads on the underside closest to the socket for vCORE?
http://www.youtube.com/watch?v=wt4JTGHXjlk
P03- Revision2![]()
Bookmarks