Help! Low Vdimm prob w DFI Ultra-D

[Lithotech]

Hi all,

I can't figure out what the hell is going on with a LanParty UT nF4 Ultra-D I have that insists on feeding the ram with only 2.32 - 2.45 volts no matter what I set it to in the BIOS. Nearly every kit of ram I have errors in memtest, usually test 5, at stock speed, with full auto settings or all manually set config. Pump the vdimm up to 2.8 or 3.0, and in the BIOS it says 2.32V. The system is completely unstable no matter what I do, except for one kit of ram I have that I am fairly certain is just good at low volts.

That ram allowed me to the desktop, and the motherboard monitor vdimm readings comply with the bios readings, 2.32 or so. It is a kit of Shikatronics Shika XRam PC4000 3-3-3-8 and should be Hynix D43 or Hynix B ic's, I'm not sure.

I've tried both yellow and orange slots. With and without the +.03V option, with auto volts on everything, and manually set voltages.

I am desperate! I've come here for help knowing many here have much experience with this board. My own personal system is running one, and it works flawlessly. I get 2.74V vdimm, it's set to 2.75 in the bios.

Both boards are currently using ver 406 bios, and I just reflashed the low vdimm board 10 minutes ago and still have the same problem. My personal system runs with an OCZ GameXStream 600w and the other is on a StealthXStream 600w. I've also tried the other with a couple other power supplies I have here. I've tried two CPU's, a Winnie 3200 and a Manchester X2-3800.

I've checked and rechecked the manual, all the jumpers, everything is ok and the vdimm overvoltage jumper is on 3.3v. The only thing I haven't tried is flipping the vdimm jumper to the 5V -- I only have 1 kit of BH-5 and no way am I going to risk it in this board that seems to have no control over the vdimm.

I even tried RMA'ing the board. I took good notes when I first got this board, and should have a ton of pictures too. I sent in a Rev AE1, and the board I got back was a sealed everything brand new (even the box is pristine) Rev AF0!

And it does the same thing! :bonk:

Ugggghhhhh........

Well, not exactly the same... my previous notes show the first board only read 2.1V vdimm, so the new board that came back picked up .02v for me. :cry:


I just can't understand it, how can two of these boards have the same problem like this? They wouldn't have swapped the cmos chip on them would they? And if they did, wouldn't they test it first?? It's the only thing I can think of that could be common between the two. Could a bad cmos chip do this?

Am I missing something in the bios or jumper settings?

I've racked my brains on this for days, spent hours trying to find others with a similar problem, and come up with nothing. Time to ask for help! My last option is to RMA the board again! I hope I don't have to, I hoped the bios flash would help, I've run out of things to try.

Thanks for any help, please ask if there is anything else I've neglected to mention!


.

[uwackme]

Umm there is a VDIMM jumper in the upper right hand corner, you have to set it to use 5V for the Vdimm regulator in order for higher than 3.3V

[Lithotech]

Umm there is a VDIMM jumper in the upper right hand corner, you have to set it to use 5V for the Vdimm regulator in order for higher than 3.3V

Hi, thanks for the reply!

Umm, yes, the jumper is for going over 3.2V (3.3V to 4V iirc).

The problem isn't getting higher than 3.2v, the problem is that no matter what I set it to, I only get 2.3V. That is a little low for almost all types of Ram, and hence it is extremely unstable.

Pump it up to 2.9V or 3.0V, and it still reads only 2.3V in the bios or operating system.

I haven't tried moving the jumper over to the 5V rail, because this is so strange to begin with I just don't trust it to not fry my ram.

Recent tests and trials I've tried as per suggestions from other forums:


Clear the cmos with battery out for 4+ hours
Boot with a single stick in the upper orange slot
Load optimized defaults
Try different ram
Try brand new Antec SP2-450w psu



My main personal system is the same board as this, my next step is to pull the cmos chip from it and try that in the flakey board. If it gives me proper vdimm, then perhaps DFI will just send me a new cmos chip which would be much easier than me RMAing this board again. :yepp:

If it doesn't help the vdimm, then I shall plop the mobo into the cheapest piece of crap case I have here, and pitch it down a few flights of stairs and then grab the nearest fire extinguisher and using it as a phseudo sledge hammer, proceed to pound the living crap out of it until it starts oozing fluid onto the floor from ruptured capacitors. :shocked: :shrug:

Well, maybe not, but that was fun to type and quite stress releasing in itself. :rofl:

Thankyou.

.

[Lastviking]

Hi, thanks for the reply!

Umm, yes, the jumper is for going over 3.2V (3.3V to 4V iirc).

The problem isn't getting higher than 3.2v, the problem is that no matter what I set it to, I only get 2.3V. That is a little low for almost all types of Ram, and hence it is extremely unstable.

Pump it up to 2.9V or 3.0V, and it still reads only 2.3V in the bios or operating system.

I haven't tried moving the jumper over to the 5V rail, because this is so strange to begin with I just don't trust it to not fry my ram.

Recent tests and trials I've tried as per suggestions from other forums:


Clear the cmos with battery out for 4+ hours
Boot with a single stick in the upper orange slot
Load optimized defaults
Try different ram
Try brand new Antec SP2-450w psu



My main personal system is the same board as this, my next step is to pull the cmos chip from it and try that in the flakey board. If it gives me proper vdimm, then perhaps DFI will just send me a new cmos chip which would be much easier than me RMAing this board again. :yepp:

If it doesn't help the vdimm, then I shall plop the mobo into the cheapest piece of crap case I have here, and pitch it down a few flights of stairs and then grab the nearest fire extinguisher and using it as a phseudo sledge hammer, proceed to pound the living crap out of it until it starts oozing fluid onto the floor from ruptured capacitors. :shocked: :shrug:

Well, maybe not, but that was fun to type and quite stress releasing in itself. :rofl:

Thankyou.

.

hmmm looks that you are doing something wrong but maybe i have wrong...

Sett vdimm to 3.2 and save in bios and after that sett it to 3.4v(dont now if your ram will like it but if you use 4v mode(+5v) you dont want to run lower than 3.4vdimm)

if you use 3.3v mode use 2.6-3.0v and check your +3.3v psu line in bios.

Also connect all power cabels to the board.

[PhilDoc]

Seems more like your shorting something out in the mb. The Utra-120 can bump on the first ram slot a little. On mine it doesn't seem to cause a problem, but maybe on yours its pushing it over too far. Try running it with the stock cooler

[Lithotech]

Seems more like your shorting something out in the mb. The Utra-120 can bump on the first ram slot a little. On mine it doesn't seem to cause a problem, but maybe on yours its pushing it over too far. Try running it with the stock cooler

Thanks! Very good thought PhilDoc, I've been struggling to think out of the box on this because I've repetatively eliminated the obvious and easy.

However, I think you are mistaking the system in my sig as the one I am having problems with... it is not. The system in my sig is my personal system, works flawlessly.

The one I am having problems with I bought the mobo recently for using as a testbed on my workbench.

On the heatsink note, the most recent tests with the X2-3800 were done with the stock HSF on it. When I had the other cpu's in I was using a Big Typhoon and I think I even tried a Freezer Pro at one point.

But I was thinking along the lines as you, and seeing as how different PSUs or RAM make no difference, I started trying different everything else:

CDRWs, DVDRWs, usually booting with only an optical and no HD unless I had the Shika Ram in and wanted to hit the desktop, and then just a single Hitachi 80gig sata2 was hooked up. I made sure to try different IDE cables as well.

The board was always setup open air, mostly on an HSPC Tech Station. So I began testing on a plain mobo box thinking just in case there was some funky problem or short happening with the tech station. Any other mobo works fine on the HSPC, as well as on a plain box. So it isn't some sort of nasty newb habit I have either.

The last thing I tried was to shut down and clean my personal system as it has been months since it was last cleaned and was quite dusty. Once I had it dust free, pulled the vid card and disconnected most of the cables, I pulled it's cmos chip to try in the boards that is having the low vdimm problem.

Incidentally, they are both using the same bios, 0406.

I plop my personal cmos chip into the other board, hook it all up, boot and enter the bios to check the ram voltage and...

still 2.32 volts.

:shakes:

Hmmm, so it isn't anything to do with the cmos chip. Fine.

I head into the cmos reloaded page to see what I have saved there, and behold... all the saved settings are for the board I am on! The one with the low vdimm problem! I thought these cmos saved settings would be held on the cmos chip itself, and I would be able to access my saved settings from my personal system since that is the chip I was using??

So now I am back to square one, slightly more confused than before but still learning from the experience.

I am wondering, if the cmos reloaded settings aren't saved on the cmos chip, where ARE they saved? I guess that is why they don't get wiped with a cmos clear with the battery removed, but how would one clear these if needed?

Anyway, I am about to give up on this. It is becoming frusterating, completely baffling because of the fact that two different boards are doing this to me yet my main one works fine -- I cannot fathom how a returned board to DFI gets replaced with a new one that does the same thing yet no matter WHAT I plug into this board the same thing happens to me.

I've got another RMA number, and this one is going back very soon unless some brilliant person here can brainstorm the secret solution. :shrug:

Thanks again for all the help!

.

[kitfit1]

I've read through this thread a few times now, and must admit, it's got me stumped as well. One thing that has slowly crept into my mind (trying to think "out of the box" a bit) have you tried the rig with a different cpu ? It might be worth a shot.

[Lithotech]

Yeah, I did actually, three of them. A winchester 3200 extensively, a venice 3800 briefly, and a dual core x2-3800 manchester most recently.

[PhilDoc]

I can't think of anything else. At this point I'd go for the rma and you never know you may have just got unlucky twice