High CPU Voltage = Reboot

[Twister1234]

hm,mine runs fine @ 1,925 vcore-even if mbm5 shows heavy fluctuations...
perhaps i should cool my mosfets-like i used on my abit,where i was runing 2,15 vcore(for benching only)

 

[enduro]

Yeah, I'm not sure if the same applies to DFI, but I've seen infrared pics of nforce2 boards under use, and the NB and the mosfets are the two hottest spots on the board, next to the cpu of course. Cooling the mosfets can only be beneficial. Just make sure that you don't touch the mosfets to see if they are hot. I'm pretty sure that they are around 60-70°C. As for the rails. Like what was said above, MBM is extremely inaccurate at detecting voltage fluctuations or in just detecting the correct voltage If you have a digital multimeter, just test your rails from a molex connector to check for flucuations. You can use the orange cable on the PIV power connector to check the 3.3V rail I believe.

 

[obsolete]

Like I said, it's a problem with the motherboard. I could push to 2.0v and beyond with my NF7-S, but with this DFI, like so many others, going past 1.85v causes vcore fluctuation. I have a Fortron 530w, and my rails are spot on when measured with a DMM. MBM5 is completely innacurate for checking rails, and I would recommend against using it. My DMM says 12.05 or so, and MBM5 reads 11.61v. It's not really the fault of the program, but the fault of the onboard sensors.

I have heard of this before. But never have I encountered this. All 5 or so DFI boards I had ran 1.9 no problem. I never payed any attention to MBM, I always used a digital multimeter. I checked the readout on the board itself under load & the highest the vcore would go is 1.9. Anything higher didn't make the slightest bit of difference. Not sure.....Not a problem now cause I have an Abit so the vcore owns. 2.1vcore currently

As for the rails. Like what was said above, MBM is extremely inaccurate at detecting voltage fluctuations or in just detecting the correct voltage If you have a digital multimeter, just test your rails from a molex connector to check for flucuations. You can use the orange cable on the PIV power connector to check the 3.3V rail I believe.

You can check the molex connectors to see if it's your PSU that is the problem. But you might also want to find the read points on the actual motherboard. That way you know what the motherboard is really getting oppose to what the PSU is trying to push through. Black & orange is correct for the 3.3v line.

 

[julianba]

Once I ditched my antec true 430 with 20A on the +12v line and replaced it with a sparkle 550 with 36A on the +12v line, my reboots @1.85+ vcore went away. I've only used up to 1.95vcore so I don't know the stability beyond that. Before I changed the power supply any vcore over 1.85 was very unstable.

The voltages under load were identical for the two power supplies so I'm assuming either the bigger +12v line or the active pfc of the sparkle is responsible for the added stability.

I also noticed slightly more vcore is required on my dfi than previous motherboards to keep the cpu stable at a given speed with the same cooling.

 

[johan851]

I have heard of this before. But never have I encountered this. All 5 or so DFI boards I had ran 1.9 no problem. I never payed any attention to MBM, I always used a digital multimeter. I checked the readout on the board itself under load & the highest the vcore would go is 1.9. Anything higher didn't make the slightest bit of difference. Not sure.....Not a problem now cause I have an Abit so the vcore owns. 2.1vcore currently

That's interesting. I might give it a shot, because I loved having a high vCore with the Abit when I had it. Did you use mosfet cooling on your DFIs? Currently I just have an 80mm fan over the mosfet area, which probably does a better job than heatsinks with no airflow...cools everything down quite nicely. If heatsinks on top of that would help (depending on your experience) I'd be up for it.

 

[obsolete]

I never used Mosfet cooling on any of my DFI's. Your 80MM fan is probably plenty. Unless you have done volt mods, I don't think it's something to worry about. But if you do, stick a heatsink on your SB as well.

There is a guide over at Xtremesystems that has some useful information posted by Lastviking. Not sure if you have seen this.

 

[mitro]

3 different DFI Boards, 2 different power supplies (Fortron 400 & Antec 550), and 2 XP-Ms and all result in the same thing for me... reboots at anything over 1.9v. I have done the volt mod and sinked the mosfets with identical results with or without.

 

[Samurai Jack]

Is there maybe a droop mod that can help this, like adding Caps to stablized the voltage at these high settings.


SJ

 

[obsolete]

While the DFI does have the option in bios to go over 1.9. It does not without mods. If you grab your multimeter & read the points on the board for the vcore, you will see that it maxes out at 1.9v.

There are vcore mods, vdimm, vdd, you name it. Check out that link I provided. Thats how you can fix it & go over 1.9v.

 

[Twister1234]

hum mine seems to stop @ 1,925...higher voltage doesnt bring stability...nothing

 

[lokkers]

With my Infinity/XP-M2600 any more than 1.9v brings a hard lock in windows under load. I am going to try the wire trick to force the default voltage higher, so I'll post if that works, although I doubt it very much

 

[CandymanCan]

What are youre temps for the cpu, if youre raising the voltage that high and youre cpu temps are like sky rocketing thats the reason for the lock up's. Remember raising the voltage doesnt always have the desired affect

 

[lokkers]

Sorry yeah forgot temps.
no higher than 45c under load, watercooled with a LRWW, ehiem 1250 and a 120.2 rad with 2 7v 120mm fans. Thats socket temp, diode never see it above 34c. Even taken with a pinch of salt, I cant really see it being temps.

 

[CandymanCan]

You ever think that youre cpu just isnt stable @ 2.7ghz? Or is it prime95 stabke @ 2.7

 

[lokkers]

No the cpu is good at 2.7ghz, its the voltage thats the problem. Even at 2.2ghz @ 2v, the same thing happens. It seems to be problem with mobile chips and high voltages on this motherboard.