DFI RDX200-DR troubles

[Diverge]

Actually I wasn't clear about the voltage jump - it is not that within Windows the voltage changes - it is that with a reboot the voltage is suddenly higher.

I'm just curious, but what were you using to measure the vcore during bootup; a DMM? If so, where on the board can it be measured?

Maybe i'll look at mine when i get my board back, if i knew where to probe. Also, did you set the cpu startup voltage to anything besides the default auto setting? I never played much with mine, but then again, i usually leave my PC on unless i am tweaking bios settings, ect.

[mike]

*** Read Below ***

I believe we shouldn't jump to conclusions here. Sampsa lost a CPU on a different board from Sapphire, not the DFI one, and other members have also provided proof that their boards are working fine, with minor memory related issues.

While I do believe MikeGuava is being sincere with respect to the problem, this may very well be an isolated incident FOR THIS BOARD. In no way are there CPUs dying everywhere on this board. I imagine DFI goes through extensive testing on many different CPUs before they release it to the public.

It is always up to members to purchase a motherboards they like and many factors go into getting the one they want including compatiblity and such, so it's fine if you would like to see this issue addressed before making a similar investment, but please keep in mind that there is no concrete and consistent evidence that the extent of MikeGuava's problems, which we all sympathize with, represent the majority of motherboards in question here.

Thank You.

Perkam

I totally agree with Perkam. My intentions with this posting is to warn as well to find out as much information as I can. I attempted to be as forthcoming and objective as possible with my posts, and set any emotions aside - I do not want to create any negative spin on DFI or the boards, until we know for sure what is going on.

I would like to find out as much as possible because neither can I afford to write off 3K worth of processors.
Initially, with both boards still working I had made a comment on DFI forums along the lines that I feel like a beta tester that has to pay extra, beause in my point of view the board shipped crapiest bios I have ever sat in front of - IMPOV the boards should not have been released at that time.

Now I feel like a crash test dummy that has to pay for his services. Hopefully that will not be the case - time will tell.

[mike]

I'm just curious, but what were you using to measure the vcore during bootup; a DMM? If so, where on the board can it be measured?

Maybe i'll look at mine when i get my board back, if i knew where to probe. Also, did you set the cpu startup voltage to anything besides the default auto setting? I never played much with mine, but then again, i usually leave my PC on unless i am tweaking bios settings, ect.

Startup voltage was set to 1.4V
I onlt measured the voltages via the ITE Smart Guardian. I had wanted to measure the voltages intitally with a DMM - digital multi meter, but never got to that.

[mike]

The RDX200 is using the same Vcore regulator as the NF4's. The only difference is the fact that the 4th phase has been put into use on the RDX. If there is an issue with the board killing CPU's, I would say that its probably BIOS related, and nothing to do with the regulators implementation on the board.

I've been running a 3700 in my board since Thursday, and I've not had any problems with Vcore spikes or dead CPU's etc. I've had a DMM connected to the Vcore supply and I've seen no noticable spikes either.

Would you mind looking into the combination 1.425V + 102% - startup voltage - between 1.4V and 1.45V

would be very interested in seeing if that can be recreated on another board - espcially with a DMM reading the true voltage.

[persivore]

I've just checked what the Vcore supply does at power up. With the % increase function in BIOS set to auto, the board gives 1.55v to the CPU. This then decreases to the Vcore set in BIOS once the CPU has been detected.

When the % increase is taken off auto (to say 102%), the 1.55v is increased by 102%. After the CPU is detected, this then decreases to 102% of Vcore set in BIOS. This is the only thing that I have found (so far) which could be causing a problem.

If the board fails to post (4 LED's lit), the pre-CPU detection voltage is given to the CPU.

If you have a low Vcore and high % increase set, I would hate to imagine the sort of voltages being given to the CPU (about 2.15v worst case from what I worked out)

[Diverge]

Persivore, is it possible to show where you are reading the vcore readings off the motherboard? A pic, or just a pin number of whatever labeled component? I'm curious and want to check mine out when my replacement board gets here.

[Charles Wirth]

I guess you guys have not seen the AMD death thread where some members killed a half dozen or more FX processors on NF chipset. God forbid we count the memory deaths.

I know the feeling of having a CPU just up and die on you. I had an FX55 die at stock speed while idling during the middle of the night. I am sure it was something I did wrong...

This is one of the main reasons I dislike testing AMD processors.

Sorry to hear about your loss, it comes with the territory. Can't blame it on the chipset or the motherboard when it could a grounding or power supply.

Go AMD!

TBH, I would be more scared to put my FX55 back into a NForce mobo. My Intel based NF boards have been nothing but trouble. Three power supplies and my best DDR2 modules fried. It blew the chips off the PCB. Pictures posted in my C19 thread iirc.

[Charles Wirth]

I have been running the RDX for a few weeks now, near max voltage on the 3500+ E6 and mild overclocked 300x10 the X2 3800+

No problams at all so far with the DFI CF board for me.

[persivore]

Persivore, is it possible to show where you are reading the vcore readings off the motherboard? A pic, or just a pin number of whatever labeled component? I'm curious and want to check mine out when my replacement board gets here.

I've marked some points where you can measure Vcore . Take your pic as to which one you use.

[s7e9h3n]

Hey Mike,
Sorry to hear about all your troubles. Just do me one favor - Don't run that 4600+ in one of those boards . If you like, I'll talk to somebody for you to see what your options could be in terms of an RMA Drop me a PM.....

[Diverge]

I've marked some points where you can measure Vcore . Take your pic as to which one you use.

Thanks, so many to choose from

[xenolith]

This thread is scary... for any motherboard owner.

I guess I've been real lucky. In hundreds of builds over the years, I've never experienced anything like you just have Mike.

Makes me want to go out and get a cheap Sempron for testing everthing now.

[G H Z]

I've just checked what the Vcore supply does at power up. With the % increase function in BIOS set to auto, the board gives 1.55v to the CPU. This then decreases to the Vcore set in BIOS once the CPU has been detected.

When the % increase is taken off auto (to say 102%), the 1.55v is increased by 102%. After the CPU is detected, this then decreases to 102% of Vcore set in BIOS. This is the only thing that I have found (so far) which could be causing a problem.

If the board fails to post (4 LED's lit), the pre-CPU detection voltage is given to the CPU.

If you have a low Vcore and high % increase set, I would hate to imagine the sort of voltages being given to the CPU (about 2.15v worst case from what I worked out)

In that scenario if Mike was only running a 102% bump that should not have taken 3 CPU's with it. Perhaps a particular combination of vcore changes until BIOS saved are a risk on POST. If a BIOS mod could be made to eliminate the % increase being set on the non-post CPU voltage we should be alright.

Either way I don't think I'll be running my FX in this board until this is cleared up.

[RJARRRPCGP]

Possible causes:

Bad caps (again) (If a cap shorts, a component will fry)

Improper heatsink contact (I hope that there's nothing wrong with the heatsink supports)

Did the caps look fine after the symptoms occured?

[mike]

In that scenario if Mike was only running a 102% bump that should not have taken 3 CPU's with it. Perhaps a particular combination of vcore changes until BIOS saved are a risk on POST. If a BIOS mod could be made to eliminate the % increase being set on the non-post CPU voltage we should be alright.

Either way I don't think I'll be running my FX in this board until this is cleared up.

I think this has a lot of potential to be the culprit - from NF4 onwards I had been using the % to raise voltages. Thanks Persivore for looking into it.

Maybe the CPus were hurt in earlier boot ups - and the time when they finally went was just random

[mike]

Possible causes:

Bad caps (again) (If a cap shorts, a component will fry)

Improper heatsink contact (I hope that there's nothing wrong with the heatsink supports)

Did the caps look fine after the symptoms occured?

Nothing that I could see visually in respect to caps. Heatsink contact - no way. I was good on that.

Hey Mike,
Sorry to hear about all your troubles. Just do me one favor - Don't run that 4600+ in one of those boards . If you like, I'll talk to somebody for you to see what your options could be in terms of an RMA Drop me a PM.....

LOL - No way - this baby will stay clean and safe!



Well the boards and CPUs are in DFI's hands now - will see what their findings are. They are very interested in my case and are going to have a very close look into things. I am sure that things will be sorted out soon.

[Dumo]

Good for you Mike Hopefully DFI will sort out any problem

[Jort]

Opteron 146 died WITH the dfi rdx,

i am happy that i have rma on the cpu, shipping next week.

when i get the dfi back THE first thing i am going to do is sell it at this point.

DAMN first MEM killing NF4 and now CPU killing RDX

DFI get down your feet please!!!!!

[Diverge]

Opteron 146 died WITH the dfi rdx,

i am happy that i have rma on the cpu, shipping next week.

when i get the dfi back THE first thing i am going to do is sell it at this point.

DAMN first MEM killing NF4 and now CPU killing RDX

DFI get down your feet please!!!!!

What was the Vdimm and Vcore you were running? Perhaps this thread contains some info as to the possible reasons. http://www.xtremesystems.org/forums/...ad.php?t=79509

[k0nsl]

I'm running the RDX200 now with my 3200+ and it has been online for a couple of hours. No problems at this end, and the CPU is still not killed. The only problems are those annoying as hell lockups during bootup screen and after saving bios values. Really annoying. The bootup image is also fubared while using my X1800XL as confirmed by others already.
Otherwise everything is working perfectly. I am using the latest beta bios.

-k0nsl

[Diverge]

I'm running the RDX200 now with my 3200+ and it has been online for a couple of hours. No problems at this end, and the CPU is still not killed. The only problems are those annoying as hell lockups during bootup screen and after saving bios values. Really annoying. The bootup image is also fubared while using my X1800XL as confirmed by others already.
Otherwise everything is working perfectly. I am using the latest beta bios.

-k0nsl

Yeah, my DFI bootup was also corrupted or 8-bit looking too, with my x1800. If you wanna help troubleshoot the bios lockup's, see oskar's thread here : http://www.dfi-street.com/forum/showthread.php?t=29755

[k0nsl]

Thanks Diverge. Will look at that thread now

-k0nsl

Yeah, my DFI bootup was also corrupted or 8-bit looking too, with my x1800. If you wanna help troubleshoot the bios lockup's, see oskar's thread here : http://www.dfi-street.com/forum/showthread.php?t=29755

[Jort]

i havent got the chance to meet the POST screen so thats out of the question

[Hybrid]

So does this board play well with the Seasonic S12 600w? I know the dfo nf4 boards didnt like this psu. I might pick this board up soon if sapphire doesnt get off thier butts with those mobos.

[calcal]

Opteron 146 died WITH the dfi rdx,

i am happy that i have rma on the cpu, shipping next week.

when i get the dfi back THE first thing i am going to do is sell it at this point.

DAMN first MEM killing NF4 and now CPU killing RDX

DFI get down your feet please!!!!!

another reason to get intel..LOL