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

[OzSnoal]

Yeah, I have posted this as a reply to another thread, but think it warrants one of its own.

 

(I have also posted this at the OCZ forum)

 

It's DFIs bad hardware design. Specifically Vdimm, which in my opinion is not performed in an optimally, in the terms of how this voltage is maintained when the power is removed from the mainboard. As a note: there are variable resistors that can be controlled electronically and that do maintain their resistance even after power has been removed from them - this is definately not what DFI is using.

Enough waffle onto the meat.

 

First of all I took a multimeter, black lead into the black wire/GND of the 4 pin molex plug, the red lead of the multimeter onto pin 7 of the Vdimm. See attached picture.

 

My CMOS setup for Vdimm is 3.5V

(1) With my PC on: 3.6V

(2) Shutdown my PC: 0.0V

(3) Measurement at power on: Does not overshoot 3.6V

 

Everything is fine and dandy - exactly as it should operate.

 

Next up - the problems begin

Shutdown the PC and remove the power plug

(4) Reinsert the power plug, the PC is OFF: 2.7V

--- This in itself is strange - Why is there voltage to the RAM? - the machine is off!

--- See (2) above, there the measurement was zero as expected, this 2.7V should not be there.

(5) Turn on the PC. Vdimm = 2.7V and I get the beepity beep!

The voltage remains at 2.7V, and the PC cannot boot!

 

Question 1: Why is DFI not remembering the set voltage when the power is fully removed?

 

Question 2: Why if you shutdown the PC there is 0V on Vdimm, but if you pull the power plug and reinsert it (without turning on the PC) there is 2.7V on Vdimm?

 

DFI is assuming too much on there always being power on the mainboard - not good.

 

 

I just thought of this - I am not sure of DFI's/standard BIOS startup. Perhaps they start at loose timings, and low FSB on startup to detect the RAM, then my theory kinda falls flat. But not completely, because it is the only thing that makes sense to me so far as to exactly why I have a cold boot problem.

 

 

 

 

 

 

 

 

 

 

 

=======================================================

 

 

THIS IS POSTED BY ANGRY_GAMES IN THE VERY LAST POST AND IS BEING POSTED HERE IN THE VERY FIRST POST SO YOU CANNOT MISS IT

 

 

 

 

 

 

 

 

Funny how DFI's employees are "incorrectly reporting" their are issues with the jumper, and we are now being asked to contact the right people who can help us...perhaps a subtle hint?

 

which DFI employees are these?

 

We've contact the right people from the beginning. We have also explained more than a few times that Rgone and I are not engineers, bios writers, nor anything like that that will be able to solve this problem for you.

 

We are tech support, as I have explained. Our job is not to fix these problems that involve engineering/design/bios writing. Our job is to bring it to the attention of those that do those jobs so that they can fix it. You guys need to understand the difference and listen to what Rgone and I have said.

 

You need to email that address he gave you. It is the only way TW is going to come around to resolve your problems. They do not read this forum. They barely speak/read/write in English.

 

We have given you the right place to go. If you choose to call it 'corporate run-around' then that is up to you, but it just tells me you don't understand exactly how a corporation works, especially a multinational one where the top tech support guys only speak english and the engineering/design/bios teams all speak mainly only chinese.

 

 

Id also like to ask why either of you cant inform the right people with the right info for us, especially if it is the last bit of time you are going to spend on this issue..... even an email linking to this thread....after all you guys work for DFI? You know who we need to talk to about this better than any of us possibly could?

 

you need to ONCE AGAIN read exactly what we have written to you over the last week in this thread.

 

WE CANNOT FIX YOUR ISSUE. WE CAN ONLY CONTACT THE CORRECT PERSONS IN DFI TO CORRECT IT FOR YOU. WE HAVE DONE THIS. WE HAVE DONE THIS OVER AND OVER. WHAT PART OF THIS DO YOU REFUSE TO UNDERSTAND? I CANNOT MAKE IT IN ANY BETTER, MORE PLAIN ENGLISH THAN I AND RGONE HAVE AGAIN AND AGAIN AND AGAIN AND YET YOU ARE STILL SEEMING TO BE TOO BULLHEADED TO READ WHAT WE WRITE. READ IT. UNDERSTAND IT. WE HAVE DONE OUR JOB OF CONTACTING THE RIGHT PEOPLE. DO I HAVE TO SAY IT ONE THOUSAND TIMES BEFORE IT LODGES ITSELF IN YOUR SKULL AND YOU FINALLY UNDERSTAND AND LISTEN TO WHAT WE ARE SAYING????

 

If you cant, ALL the people having these headaches will be forced to spend the time messing around with email tech support, when it might not even be the crux of the problem...But your boss probably told you to say that so were S.O.L. I guess.

 

we don't deal in speculation and that is exactly what the last sentence of that is. Pure speculation. Since you don't have a clue what the bossmans and us talk about, it is better to not try and guess because you are going to be wrong even if you believe with all your heart that you are right. You don't have a clue who our boss is and how hard he works to make sure we take care of customers. Speculate all you want. It's up to you.

 

Rgone:

 

Not everyone has magical ram that can do tight timings at 3.2v ....

 

BUT

 

we all have EXPENSIVE motherboards to supply our EXPENSIVE ram with the necessary voltage required to work like your magic bh.

 

ALL OF US.

 

then after so many modules that have given you trouble, why do you still insist on using the same CH based modules and possibly BH modules? Why not move to something else that doesn't require high voltages? Why not move to another motherboard that doesn't give high voltage or does give high voltage but doesn't have this problem with the memory that the DFI NF4 does? Why must everyone be so stubborn as to come here and tell Rgone and I how crappy we are and how we don't support anyone and yet you are stubborn like a mule and refuse to budge about the combination of UTT + DFI NF4? Isn't it time to move on to another board or another type of memory after so many RMA's?

 

No offense but are you trying to rub it in that we all got ripped off here? Are you saying its our own fault for using high voltage memory and maybe we should be smart like you and not use high voltage stuff?? That is the tech support your are giving us on the issue?

 

that is not what Rgone said. Maybe clear your head out a bit before making such a claim and re-read what he wrote. Since we cannot fix the problem you have, we can only suggest a way to avoid it.

 

But since he gave you a direct contact to the people who can fix it, why do you persist on making it an issue to Rgone and I in this thread? Why not just stop doing such things and CONTACT THE PERSONS HE GAVE YOU TO CONTACT SINCE THEY ARE THE ONES THAT CAN FIX YOUR PROBLEM?

 

You think its no big deal I am out about $4200 bucks and lost major face to a prominent client because of this?

 

#1 it isn't our fault either that you decided to purchase this combination of hardware. You decided it. No one forced you to. No one twisted your arm or held your family hostage.

 

#2 why you would rely on an 'enthusiast overclocker' board for important business dealings is way beyond me. That seems to be some shortsight on your part for even thinking about using such a system built around an overclocking/enthusiast board to rely on it to get your important, mission critical business done every day.

 

I only use stock speed Intel P4 and Athlon64 'vanilla' machines to run as servers, business machines, workstations etc. My overclocking/enthusiast parts are left for playing games and fooling around, but I've been doing this for about 8 years and have never relied on a 'play' machine to do my important work. You might be different, but that is the difference...maybe a lesson to be learned for yourself?

 

Seriously, what was the point of saying that or anything else besides telling us to use that email adress from now on because you guys arent gonna help out anymore? The rest of your post wasnt necessary...just more haze clouding the issue.

 

again...stop yapping long enough to read what we have written. WE CANNOT HELP YOU ANYMORE ON THIS ISSUE.

 

maybe if you give me a magical engineering pill or a magical bios-knowledge pill or a magical board/electronics design pill I can instantly learn how to fix all this. Until then, READ WHAT WE WROTE ABOUT WHY WE CANNOT HELP YOU. You are starting to repeat yourself out of frustration instead of actually LISTENING TO WHAT WE hAVE TO SAY. THIS IS PART OF THE REASON WHY WE CANNOT HELP YOU. YOU WON'T LISTEN YET YOU EXPECT US TO TELL YOU HOW TO FIX SOMETHING. LISTEN DAMMIT. IT MIGHT GET YOU SOMEWHERE.

 

OCZ is out 6 sticks of memory and shipping ON ME ALONE and tons of man hours spent trying to figure it all out? Plus Oskar has spent a crapload of time going through the bios, and sending/testing me another board? Never mind all the rest of us suffering with malfunctioning components?

 

and so OCZ and Oskar are working on it, and yet you STILL COME HERE EXPECTING RGONE AND I TO HAVE A MAGICAL ANSWER AFTER WE TOLD YOU WE CANNOT FIX THIS ONLY OSKAR/TAIWAN CAN FIX THIS. AGAIN, STOP TALKING LONG ENOUGH TO LISTEN TO WHAT WE HAVE TOLD YOU.

 

 

I hate to say this, but both of you have cast so much "grey area" and confusion to this issue and clouded all your responses with implied or irrelevant info that I seriously think we might be better off continuing on our own anyways.

 

then do it. since you refuse to listen to anything we say you are much better off moving on and going elsewhere. It will be good for both of us as you aren't happy that something isnt working and we are not unhappy that we have told you to your face many times in plain easy-to-understand english everything we know to tell you to help you out and you still refuse to listen. There's a pattern here. Refuse to listen, come and complain in this thread, get told something, refuse to listen and come back with some response that doesn't make any sense when you look at what we told you to do to help yourself get it fixed and then repeat this 100 times. That is the pattern.

 

LISTEN TO WHAT HE TOLD YOU TO DO. RE-READ A LOT OF THESE REPLIES BY US TO UNDERSTAND WHY WE TELL YOU TO DO IT. STOP ARGUING WITH US ABOUT WHAT YOU THINK WE SHOULD DO FOR YOU AND DO WHAT WE INSTRUCTED YOU TO DO BECAUSE RGONE AND I CANNOT FIX THIS ISSUE. PERIOD. DO YOU NOT UNDERSTAND ENGLISH ENOUGH TO UNDERSTAND THAT RGONE AND I ARE NOT ENGINEERING, BIOS WRITING, OR BOARD DESIGNING GUYS???? DOES THAT NOT MAKE ANY SENSE TO YOU WHY RGONE AND I CANNOT FIX THIS ISSUE? DOES IT NOT MAKE SENSE TO DO AS HE TOLD YOU TO DO AND CONTACT TAIWAN DIRECTLY SO ENGINEERS AND DESIGNERS THERE CAN UNDERSTAND YOUR PROBLEM? DOES IT NOT MAKE SENSE THAT WE HAVE TOLD TAIWAN ABOUT THIS AND OSKAR EVEN CAME HERE AND TOOK A COUPLE OF BOARDS AND IS WORKING ON IT? DO YOU WANT AN INSTANT 15 SECOND FIX THAT MAGICALLY APPEARS OUT OF NOWHERE?

 

I am sure both of you are alright guys and do the best job you can, and its not an easy one to say the least, but something tells me this came straight from the higher ups and it may be a sign of what the "official response" to this issue is gonna be.

 

there is no official response. THIS IS WHAT RGONE AND I TELL YOU TO DO AND YOU SHOULD DO IT.

 

 

 

 

enough already. If you cannot listen to simple instructions then you should demand your money back for the board and buy something else that will give you less problems. Otherwise, stop talking and listen to what we say and understand why we told you to do it.

 

RGONE AND I ARE NOT ENGINEERS

 

RGONE AND I ARE NOT BIOS WRITERS

 

RGONE AND I ARE NOT BOARD DESIGNERS

 

WE CANNOT FIX YOUR PROBLEM.

 

OSKAR HAS bEEN HERE AND OSKAR KNOWS YOUR PROBLEM AND OSKAR HAS TAKEN A FEW BOARDS TO TRY AND SOLVE THE PROBLEM. WHAT ABOUT THIS IS SO HARD TO UNDERSTAND?

 

 

 

You have had DFI engineers working on it - and repeating the problem. Suddenly they are not, and we need to mail them? Come on, thats just poopy. I cannot believe it is being downplayed like this.

 

who said they were repeating the problem? I have not heard that. Send me any email/correspondence that shows this.

 

Who said they are not working on it? I don't have a clue what they are working on in taiwan. We have enough of a job to do over here. Don't you see a forum full of users needing help? Those are people that Rgone and I can help with. YOu guys in this thread we have helped as much as Rgone and I can help you.

 

Again, since you dont seem to understand plain english, I will explain it once again.

 

WE TOLD TAIWAN ABOUT THIS PROBLEM FOR YOU. OSKAR SHOWED UP AND ASKED FOR A COUPLE OF BOARDS. TAIWAN IS THE ENGINEERING/BIOS/DESIGN PLACE, NOT RGONE AND I. WHY DON'T YOU DO AS RGONE SAYS AND CONTACT TAIWAN AND DEMAND THAT SOMETHING BE DONE ABOUT THIS? CONTACT OSKAR AND DEMAND HE FIXES IT? RGONE AND I CANNOT HELP YOU IN THIS BECAUSE WE CANNOT FIX IT. WE HAVE CONTACTED ALL THE RIGHT PEOPLE. WHETHER THEY FIX IT OR NOT IS NOT UP TO RGONE AND I. WHAT IS IT THAT YOU DO NOT UNDERSTAND ABOUT THIS?

 

Disappointed by the obvious brush-off.

 

Damage control has begun....

 

Then contact the people we told you to contact and keep bugging them to fix it like we have continually bugged them since the issue was brought up.

 

If you speculate you are being brushed off without any conclusive proof then that is up to you to believe. I will not try to change your mind. I will tell you to do as Rgone says and contact the people he told you to contact and tell them to fix it for you.

 

If Rgone and I could fix it we would have fixed it and this would not be an issue. We cannot fix it. We are not engineers, we are not bios writers, we are not board designers. WE HAVE CONTACTED THE RIGHT PEOPLE FOR YOU. THE BOARD DESIGNER CAME AND TOOK A COUPLE OF YOUR BOARDS TO TRY AND SOLVE THE ISSUE. WHY IS THIS SO HARD TO UNDERSTAND THAT RGONE AND I HAVE DONE EVERYTHING WE CAN FOR YOU AND IF THE ISSUE IS NOT RESOLVED THEN YOU NEED TO CONTACT THOSE THAT CAN FIX IT FOR YOU AND MAKE SURE THEY UNDERSTAND YOU ARE NOT GOING AWAY UNTIL IT IS FIXED?

 

 

 

 

 

for those of you who think that I do not actually work for DFI, please see this:

 

 

and then call my boss @ 1-510-274-8000 and ask them if Robert (Rgone) and I actually work for DFI USA.

 

 

 

 

 

 

 

 

 

 

 

 

 

since those of you that persist in continuing to throw a fit and say all sorts of things instead of actually listening to what I have to say, I am going to say this once and once only.

 

 

HERE IS THE ADDRESS TO CONTACT THE ENGINEERS/DESIGNERS/BIOS WRITERS IN TAIWAN THAT CAN FIX YOUR ISSUE:

 

[email protected]

 

 

CONTACT THEM. TELL THEM YOU HAVE A COLD BOOT PROBLEM AND EXPLAIN IT TO THEM AND EXPLAIN THAT YOU HAVE SOUGHT HELP IN THE FORUMS AND YOU CAN GET NO HELP BECAUSE THE EMPLOYEES THAT WORK THE FORUMS ARE NOT ENGINEERS, DESIGNERS, NOR BIOS WRITERS AND WE SUGGESTED THAT YOU THE CUSTOMER CONTACT DFITS IN TAIWAN TO GET THE ISSUE RESOLVED.

 

 

there will be no burying of this thread. It is getting locked and stickied and that is the last thing I can do to help you. I do not have time to go around and fool with all of you that might try to be foolish enough to make multiple threads complaining about my response. There is no other reply that I nor RGone can give you to solve your issue. This is the last thing I have to say about it.

 

If you are not happy with it, then contact the persons above and tell them how unhappy you are. But you will not interfere with our ability to do tech support at this forum. Get it through your head. Rgone and I cannot fix this issue for you. I am tired of repeating it to you over and over and over and i wont repeat it anymore.

 

If you are not happy with this response from me or Rgone then either

 

1. contact the address we gave you

2. contact your vendor or regional DFI office and demand your money back and move on to another board that doesn't have the problem that you have with the DFI board.

 

 

 

 

 

 

 

[email protected] = Taiwan, engineers, bios writers, board designers reside here.

[TheMooN]

Any Feedback from the DFI guys would be most appreciated as this Cold Boot saga IS really starting to do my head in .

 

Interestingly for me It still happens when the power cord is left in situ between cold boots :confused:

[jsbond007]

that is some fantastic bit of analysis. Well done.

[OzSnoal]

Maybe Oskar can change the bios - well actually I did this test with 3-10 bios, not exactly the latest. Possibly this had been changed in one of the newer bioses?

 

Anyway, the bios needs to be changed so that Vdimm is to set the CMOS indicated Vdimm voltage before it tries to detect the RAM. I think that would fix my cold boot problem. Perhaps many others - have you noticed how some people come here with cold boot problems after they have changed hardware? The reason being that they removed AC power from the box, leaving them with 2.7V on Vdimm.

[Ubermann]

What readings do you get without any ram in slots ?

[OzSnoal]

Quote from Bigtoe which nicely explains the problem, fits my findings nicely:

http://www.xtremesystems.org/forums/showth...?t=61210&page=4

A major cause of cold boot issue = BIOS

 

Reasons why.

 

When a board first boots it runs default speed to engage the video bios etc, then all overclocks are applied etc and you can enter cmos.Now, if the fsb and timings are applied before the voltage the board will cold boot. Remember we very rarely see the video bios and this all happens in the blink of an eye so your screen will always look black.

 

So to explain what should happen.

 

1 you push the power button.

2 the board boots at defaults to initiate video

3 the voltages should be applied

4 the overclock and other setting are applied

5 the board boots past cmos and you load OS

 

Now if it doesn't go in that basic order the board will freeze on the overclock as the ram has no voltage....this is why you are having a lot of issues with high voltage ram..IE VX or any UTT BH5 ram out there.

 

Cure....bios rework, or have the board set really slack timings at boot which allow the ram to pass the fsb stage and then see the high voltage it needs. I do this in most of the bios mods i release as it really helps with cold boot issues.

 

If you cold boot, measure how long the ram takes to recieve the voltage it needs, remember it should see it very close to the start of the boot cycle, if its a few seconds in there will be issues.

 

I hope this explains the main cause of cold boot issues, and why RMAing the ram is usually not the cure.

Remember the board overclocks the ram, it has to do it correctly for everything to go smoothly.

 

 

1 last point...why does TCCD not have this issue?

 

TCCD or TCC5 does not have this issue as the board is booting at 2.6V or so, many of the modules are able to cope with the fsb at this VDD.

[Sharp]

Hello,

 

Is this the same for all he DFI nF4 boards? (All revisions?)

If I wanted to test the vdimm voltage,

 

Do i just touch pin 7 with +VE multimeter probe

 

And -VE multimeter probe to what?

[OzSnoal]

Where to put the black lead:

First of all I took a multimeter, black lead into the black wire/GND of the 4 pin molex plug, the red lead of the multimeter onto pin 7 of the Vdimm. See attached picture.

 

BTW, put the red probe into the small hole where my arrow is pointing to in the pic, and not into the actual RAM slot.

[red930]

Has anyone else been able to reproduce this problem.

 

My rigs are running fine so I can't reproduce the symptoms.

[OzSnoal]

Basically I am having a 3 LED problem, only when cold booting with AC removal, i.e. RAM is not being detected.

 

You could simulate the exact same thing, by just removing your ram and powering up (after having removed AC). You will notice that Vdimm will not reach the value that you specified in the CMOS. This is the crux of the problem, not applying the correct voltage.

[Sharp]

Hello,

 

I will test the voltages and see what happens.

 

Does anyone experience this problem if they are not overclocking?

Based on everything that has been said it would seem to only effect people that require the high voltages.

[red930]

Removing the memory stick changes everything so I don't see how that could be a valid simulation.

 

If, and this is a big IF, what you state is true, then that would point to the RAM being a possible source of the error.

 

Or am I missing something very obvious? This is getting over my head very quickly. :confused:

 

You could simulate the exact same thing, by just removing your ram and powering up (after having removed AC). You will notice that Vdimm will not reach the value that you specified in the CMOS. This is the crux of the problem, not applying the correct voltage.