Cold Boot? Memory Died? [email protected] = who you need to contact

[Rezag1000]

if the boards are found defective (I hope not), whats going to happen? Can DFI replace all of them? and what about the memory people lost?

 

I hope DFI doesn't go out of business!!

[red930]

What?

 

I think you can rest easy about the future of DFI

if the boards are found defective (I hope not), whats going to happen? Can DFI replace all of them? and what about the memory people lost?

 

I hope DFI doesn't go out of business!!

[Old_Fart]

You mean something like this ? On my picture, it is the wire going from the CMOS 3 pins to the switch with red electrical tape on the backside. Yeah... I've been lazy too already and made thosed things on all my rigs since my AMD overclocking adventure and really crappy cmos reset jumper emplacement. I have 2 or 3 of them somewhere in a box, they are a bit more nice than this one (this one was an old rig ) and only God know how this thing have been useful when NF7-S was a bios killer and cold booting pos :shake:

 

Looks like you're not as lazy as I am. I'm about to leave on a 4 day weekend but when I get back, I'll take a picture or two of the CMOS clear jumper I came made and will post links (IF anyone is interested). Just got a new Nikon 8 megapixel digital SLR and have been wanting to see how the macro function works.

 

I'm still running my OCZ VX PC4000 at only 200MHz (DDR400) but I haven't had a single cold boot issue since I switched the 4v jumper back to it's default position (meaning I'm only running the ram at 3.2 volts). I have yet to try the settings that Andy @ OCZ posted the link to (the Bleeding Edge link) but will when I get back home from my trip.

 

Actually, I'm almost hesitant to try anything different at this point but it still irks me that I can't even get a stable STOCK DDR500 speed out of this expensive VX ram in the DFI motherboard. I'd be willing to wait for a solution to this problem if I thought one was actually going to be provided by DFI, but from what I've been reading lately (posts and emails from RGgone and others), I'm becoming pesimistic that one will be forthcoming.

 

I suspect that there's either a basic flaw in the motherboard or one or more of the components in the voltage regulator circuitry are out of tolerance. Since there are so many others out there who are NOT having problems with identical (to mine) setups, all this is beginning to smack of a QA problem (either with DFI or one of their subcontractors that provide the components they're using).

 

In one of my previous posts, I quoted RGone as advising me to RMA my motherboard and asked "to whom?". I never did get a response to that one but I'd be curious as to exactly HOW I'm suppose to get a fully functional replacement (meaning pre-tested and verified to work with OCZ VX PC4000 ram at 250MHz and up) that DOESN'T exhibit these problems that some of us are seeing.

 

Would ANYONE at DFI care to response to that one?

 

At any rate, I'm outta here for the weekend and probably won't even think about any of this again until next week. Hang in there, gang.

 

Later.

 

DFI NF4 Ultra-D, OCZ PC4000 VX (UNDER-clocked to DDR400 for stability), AMD A64 3500+ Winchester (w/ Danger Den water block @ stock 2.2GHz), Sapphire X850 XT (w/ Danger Den water block), Thermaltake PurePower 680, SoundBlaster Audigy 2 ZS Platinum and other stuff not worth bragging about.

[Jesus_Faction]

i wanna know what exactly these people that are having problems were doing

 

how high did they overvolt?

were they coling the ram?

did they have a quality psu?

hows the rest of the case's airflow?

 

 

there seem to be lots of opportunities for confounding here, and it seems a bit premature at best and shortsighted at worst to blame DFI right away

[kimandsally]

Well I am going to tell you how such shett gets heard. Notice I did not say, 'said', so pay attention.

 

Back in the DFI NF2 days all was well nearly with their early rocket boards until AMD locked the processors and a few folks put mobile processors in some Abit NF7 boards. With all the other shortcomings of the NF7 and I know I had 3 of them myself, the NF7 would accept the Mobile processors and folks just had it in their mind that DFI "must" support the mobiles as well. That was bull shett from the word go but the people kept trying and kept responding that they had cold boot issues. Cold boot issues. Cold boot issues. Gawdamn cold boot issues and ME too with a mobile processor.

 

About this time I got a little closer to the folks in charge at DFI and they asked me what I knew and I confirmed that in fact there was a lot of problem with cold boot issues and the mobile processor. They believed me and asked had I the ability to pull in a couple of "known" cold booting boards and "would" I also submit my OWN board and MY memory? I said OK and arranged for such to happen.

 

At least three problematic boards went to Taiwan along with my memory for test and in about 9 days, I began to get the reports back from the engineering/QC in Taiwan and to put it mildly > I was knocked to the floor.

 

My first friends board was deemed dead. Shett you say! I knew he had just pulled it from his working computer. I said Ok maybe it was damaged in shipment and waited for the reports on the other 2 boards I had had personal experience with, that being my personal board and one other from the second close friend. Report from second friend's board comes to me and it is said that it will not POST. I said what in the h*ll that board used to work. Then about three days later I got the report on my board that it would not POST either and I am shocked and dismayed. My frikken board does not POST. I thought how in h*ll they kill my board and was hoping that they had not also killed my memory.

 

I moped and shuffled my feet around the house for about 20 minutes and then had to go out to run some errands and after about 20 mins drive of the 30 mins it would take to get to my destination>>> it hit me right between the eyes. Just like a brick thrown between my eyes. Engineers in Taiwan were saying one word to describe a situation and most of us that spoke pure english were using another word to reflect the same situation and that put us miles apart in our thinking.

 

I turned my car around and drove back to the house and picked up the phone and called Mr Frank Wong for DFI in CA and told him this. I said Frank sir, cold boot equals no POST. He says to me. Robert what are you saying? I said sir, cold boot problem that all us english speaking folks are talking about is the same as the no POST problem that engineering is seeing with the boards we sent to Taiwan for testing. Mr Wong says to me. "REALLY"? I said yessir, it is a problem of communication of the symptoms from what all us english speakers see as our problem and the words that the engineers use to describe what they see.

 

Mr Wong says will you be beside your phone for the next few hours as I may need to talk with you again? Remember at this time I was not even partially employed by DFI but I said yessir.

 

About 3 hours later the phone rang and Mr. Wong said to me that since I had relayed the words that now brought the two differing terminologies together; that the engineers understood what was happening. HELLO? Hallelujia. From the point of getting all languages onto the same page there was steady progress toward a bios that would in fact let 'unsupported' mobiles often work on the DFI NF2 boards. I learned from that a valuable lesson. I do not take every word as gospel since I myself do not speak 'mandarin' Chinese. Now over a year later, I make every effort to make myself FULLY understood by those that will translate my words into Chinese for the engineers. I write the same thing often from 3 different positions, so much so that many think I talk in circles. But I learned my lesson well when riding down the road, it came like a brick between the eyes that in Taiwan the cold boot = no POST.

 

Now that is how Mr. Wong was heard or understood to say what the questioner wanted to hear; even if it was not the real case in reality. And far past that mis-play on words, I personally know that people doing testing have not found any such blowing up of chipsets. Not even any spiking of voltages has been experienced with recording Oscopes to this day to my knowledge.

 

Now that is the answer to 'how' the words one may say come out all wrong in the hearers ears. And of course the hearer is well pleased to hear what he wanted to hear in the first place. Ah hah, a problem to explain my failures with over-volted parts.

 

As for high Vdimm stick useage and over-volting of memory on the DFI NF4 board; it has already been my suggestion to flash to the N4D510-2.bin and use slots 1 and 3 (yellow) until such time as testing by DFI engineering and the memory companies can truly determine what is the truth of the symptoms of some users. That is it and this time it is my thoughts in plain english and they mean exactly what they seem to say and no more nor any less.

 

RGone...

 

Well Rgone

 

That makes perfect sense to me, over the years my job of tuning and designing racing two stroke engines I had similar problems myself getting over what I wanted manufacturing in Japan, that was quite an easy one as they are quite good with English, my major problems came when having engines manufactured in Russia, in the end the only way to solve out language barrier was for me to go to Russia with an interpreter then a very similar situation to this arose.

 

My heart goes out to the poor guys in Taiwan they are among the hardest working and intelligent people on the planet yet if they don't get the facts as they are they have practically no chance of sorting out the problem, I would offer to do some testing myself if that would help I know the meaning of how important it is to have facts and only facts.

 

I am very confident that all will be well in the end and it would be very helpfull if people who do have a problem can be as accurate as possible if they want to submit information.

[sluggo]

3.6v max voltage for BH-5 (that doesn't mean 24/7, just what is stated by winbond to be max V). http://lenine.biz/elisha/winbond.gif

 

edit: woops link no longer works.. ill see if i have the pic somewhere on my hdd..

 

Here's a link to the Winbond pdf for the part.

 

http://www.winbond.com/s-winbondhtm/partne...t.asp?Pname=918

 

The "Maximum Operating" voltage for this part is not 3.6 Volts. The "Absolute Maximum" voltage is 3.6 Volts, and the max operating is 2.7 Volts.

 

I have some experience with datasheets for electronic components. The convention is that when you say "Absolute Maximum" for a voltage rating on a part, it means exactly that. The person using this information is usually an engineer who understands it to mean "if you expose the part to anything higher than that, it may well break". In this case, "anything higher" includes a number like 3.6000001 Volts. Knowing this, the engineer will design the application so that the part will never, under any circumstances, come close to seeing 3.6 volts.

 

Of course, PC users are not bound by the petty rules that restrict engineers, so they run their memory at an indicated 3.6 volts. A problem comes up, however, when the thing that's doing the indicating (the MB mounted sensor) is not all that accurate, and the power supply that's delivering the 3.6V has at least .2V of ripple. Yes, that means an indicated 3.6V is usually a sine wave with a trough at 3.5V and a peak at 3.7V. And don't forget to throw in some random noise spikes that get ignored by the indicator but can commonly add another 200-300 mv. Now your indicated 3.6V is 4V.

 

The datasheet has a recommended operating maximum of 2.7 Volts. We all know that the part will tolerate more than that and that you can get "something for nothing" by running it hotter, so we do. However, running it at 3.6V (indicated by the mobo sensor) is lunacy. Personally, with a PC power supply and the noise spikes that are all over the typical motherboard, I wouldn't run this part higher than 3.1V (indicated). If I'm really motivated I'll get out my scope and look at what Vdimm is really doing and adjust accordingly.

 

Will the part work at 3.6V and higher? Maybe, probably, who knows. How long will it work? A while. Will it eventually break? Oh yes. Do you care? Well, that's up to you, but if you run your parts at the Abs Max regularly, expect problems that you don't expect.

 

sluggo

[uwackme]

Ive looked with a scope, and fluke 85 and the Vdimm rail is great on both Infinity NF2 and the DFI NF4's. Ive run BH5 for more than a year @ 3.5V on Infinity and had no issues. Did CH5 @ 3.3V for a year prior to that, again np.

 

I have no concerns with VX chips at 3.5V on these boards.

[CJ4]

Just another data point- with the bios set at 3.1V, my MM reads the VDIM at 3.170. This is with the jumper at 4V or regular.

 

CJ

[hagar]

Thx for the update sluggo. I've seem this CH5 datasheet so I guess the BH5 can run a little higher as the process is a few tens of microns larger, but my good old BH5 is actually running 200 HTT @ 1.5-2-2-5 @ 2.6V, discovered that today, and I would like to keep it that healthy.

[OzSnoal]

I will be receiving a new set of VX shortly as my one stick has died completely and no longer allows booting no matter what I do to it.

 

Since no one can say for sure if it is:

a) DFI boards killing the RAM, be it the 5V jumper or 'unstable currents'

B) high voltage memory that just cannot handle the high volts

 

What do you recommend (Angry/Rgone) so that I am not in the same RMA position again?

 

Run at it 2.7V? Since everyone likes this car thing, here's another - that will be like using low profile tires without air in a corvette!

 

BTW, are either of you, Angry or Rgone running VX full time in any of your rigs? I hope you are - you guys could give some valuable insight/feedback to DFI/memory manufacturer.

[loc.o]

Ive looked with a scope, and fluke 85 and the Vdimm rail is great on both Infinity NF2 and the DFI NF4's. Ive run BH5 for more than a year @ 3.5V on Infinity and had no issues. Did CH5 @ 3.3V for a year prior to that, again np.

 

I have no concerns with VX chips at 3.5V on these boards.

 

Mine is running over 6 months now, its on about 5 to 6 hours a day with me VX@3,5v using the 5v rail to supply it. Not a single problem so far. RAM, pwmic and upper right mosfet area cooled by fans.

[Angry_Games]

i have OCZ PC3200 Gold (BH-5) in the San Diego 4000 rig that is listed in the OC Database

 

 

I have OCZ 4000VX in the X2 rig that is listed in the OC Database

 

I have old Corsair CMX512-3200LL rev1.1 (real BH-5) in momma's rig running @ 3.2v with a 90nm FX-55 (one that you guys cannot even buy yet heh)

 

this is why it is so hard for ME to diagnose this problem because I ONLY use high voltage memory in my NF4 boards (TCCD + Angry_Games = big fight cuz it dont like me for some reason)

 

I always use the 4v jumper to give more voltage except in momma's rig (since she has FX55, she dont need overclockin)

 

my OCZ BH-5 = 3.5v

my OCZ VX = 3.4v

 

ive never experienced a cold boot. I've never killed a stick of RAM.

 

this doesnt mean i dont beleive anyone...as i do...I just don't have a clue what the culprit is and refuse to speculate on what it is...that is the engineer's job, to test all of that thoroughly since they build the boards and memory and would know more than me with limited test equipment and even more limited knowledge of the design/implementation of the power circuitry on the boards.

 

 

 

personally...until the issue is resolved one way or another, I would go with low voltage TCCD and a high multiplier cpu (like your 3500+, which is the only cpus i will buy now are 11x multi or higher).

 

memory timings are not real important as the A64 is a superior cpu with an integrated memory controller...you can never feed it enough memory, and cpu mhz is the ultimate goal, not super high HTT, and not super mega tight memory timings...dont get me wrong, good tight memory timings are helpful, but they dont make a difference of more than a couple of percentage points...not worth the effort in my opinion....again, my goal (and yours should be too) is MAX CPU MHZ.

 

I don't mind running 2.5-3-3-7 @ 260x11 to get my super overclock...in the end, I still get 200FPS in UT2004...2-2-5-2 @ 260x11 might get me 210FPS....but that extreme voltage might not be what i want (stress, voltage, HEAT)

 

another good trick I use with BH-5 stuff is 3.2v (never using the 4v line) and dropping the memory divider down to DDR166...I can hit 260x11 this way easily @ 2-2-5-2, 3.2v, as my memory is only running an actual speed of 210Mhz....and only performance penalty i see is memory bandwidth, which as I explained, is a moot point because software (games!) rely on cpu mhz, not memory bandwidth (and again, memory bandwidth can never be fully utilized with an integrated memory controller....there's no bottleneck there like there is on a separate memory controller hub (Northbridge))

 

 

anyway, that is my opinion. hope it is helpful :tooth: