Epox 8KHA+ Vcore mod question!!

[Paul.K]

Roadrunner,

I can do 1.808Ghz stable, runs 3dmarks2001, prime 95 and Sandra without a single problem. At this setting my Vcore fluctates between 2.0-2.03volts. Anything greater than >2.03volts and my system goes crazy with random reboots. The interesting thing is I achieved 1.87Ghz with my backup board (Iwill KK266R) and XP1900+. The Iwill KK266R had been previously modded to provide a Vcore of 2.23volts.
What this result tells me is that the AMD Athlon XP using hybrid technologies can indeed withstand a voltage greater than 2.03volts and the constant rebooting seen by having a high Vcore on my modifed 8KHA+ is probably due to a Vfbc on the board. I pray that some day this board will allow me to have a Vcore value of 2.2volts. And will therefore deserve to become a member of the overclocking family.......

I can always hope!!!!!

can't I!!!


P..

[Road_Runner]

Originally posted by Paul.K
Roadrunner,

I can do 1.808Ghz stable, runs 3dmarks2001, prime 95 and Sandra without a single problem. At this setting my Vcore fluctates between 2.0-2.03volts. Anything greater than >2.03volts and my system goes crazy with random reboots. The interesting thing is I achieved 1.87Ghz with my backup board (Iwill KK266R) and XP1900+. The Iwill KK266R had been previously modded to provide a Vcore of 2.23volts.
What this result tells me is that the AMD Athlon XP using hybrid technologies can indeed withstand a voltage greater than 2.03volts and the constant rebooting seen by having a high Vcore on my modifed 8KHA+ is probably due to a Vfbc on the board. I pray that some day this board will allow me to have a Vcore value of 2.2volts. And will therefore deserve to become a member of the overclocking family.......

I can always hope!!!!!

can't I!!!


P..

our problems sound strikingly similiar except my 1800xp starts needing more vcore before your 1900 does. Same deal with me....at my highest stable overclock, the vcore fluctuates from 2.0 to 2.03, and the minute I give it the slightest bit more voltage, I'm lucky if I can get back into WIndows much less run any system intensive app like Prime95, Quake3, etc.

The Epox guys are saying psu or bios flash but I think it's something electrical with the board that no psu or flash is going to fix.

Do you know anyone running higher than say 2.05v on these 8KHA+'s?

[Paul.K]

I'm sorry to say I dont know anyone which is over the 2.05Volt mark. Just another point to note, I am just after increasing my FSB by 1Mhz and now my CPU speed is 1812Mhz. However I am still at 2.03volts Vcore and I am now experiencing random reboots of Win XP when running intensive applications. How can 1Mhz increase in FSB cause my system to go crazy. Dropping my FSB back down by 1Mhz allows me to run stable at 1800Mhz. Go figure!!! Since I haven't adjusted my CPU Vcore why is it rebooting when I increase the FSB by 1. Could it be my memory ???? However I am only at 150FSB by 12X mulitipler and my Cas setting is 2.5 in the bios.

Please need all your input here!!!!!!!

Thanks!!

Paul.K

[dimmreaper]

I can run 2.2V for forever and a day without my system crashing . . . . If I slow the core down, and put a fan on the regulator. From this I have drawn two possible causes. Either:

A) The Morgan core Duron, which is similar in many respects to the Palomino core AthlonXP, doesn't like the extra voltage for whatever reason.

B) The voltage regulator, and it's surrounding circuitry, were never intended to operate at these voltages, that is a fact. As a result I could see the possibility of this causing some troubles with current stability.

Unfortunately we don't really have a way to test the first possibility. The second possibility is easily testable. All you would(in theory) need to do is monitor the Vcore voltage ripple with a sensitive oscilloscope. And what if someone were to find that the voltage ripple is the culprit? I can see two possible ways of "attacking" the problem.

A) Alter the MOSFET switching speed. This is fairly simple to do with these regulators, it would only require a resistor swap. This could be counter productive if you alter the switching too much.

B) Try using a bypass capacitor between pin 2 and the ground. This might help, but It could also make matters worse.

Unfortunately, I do not have a suitable oscilloscope at my disposal, so I'm not prepared to do any of this . . . .

[EPoX Tech]

Due to certain areas of the die that are to be described as "small", the more voltage you apply to a Morgan or Palomino will detract from your overclocking threshold.

That is to say plainly - they are not suitable for high voltage overclocking like Spitfire or Thunderbird.

I have tried them on many boards and can confirm that the 8KHA+ performs admirably with them - as well as 8K7A or any board.

I am presently running 1.95v which seems to be ideal for my particular setup. I found a balance between subzero temperature (loaded) and voltage low enough to be that cold to enable the switching enhancement of the silicon.

i.e. to state - More benefits can be gained by extreme cooling rather than over-volting as with Thunderbird IMHO.

[proze]

paul k. i can run well over 2.1v all day long, like jeff. no sinks or fans on the regulators or anything.

the only time my vcore gave me the problems you've described was when i had a crappy psu. getting a delta with 220w combined 3.3v,5v sorted everything out. but you have a sparkle, so it shouldn't be that.

this is wierd...

[dimmreaper]

I think we should listen to EPoX Tech on this one. Knowing for the most part exactly what his function within his firm is, and the amount of component testing he does, I can state with some certainty that in this forum he is the authority on hardware. I'm sure there are people more knowledgeable than he, but they are few and far between. Most of 'em are probably locked in a cage and presented with the task of engineering the next generation of computer hardware, and hence not able to post in this forum. Obviously it's still alright to question EPoX Tech's statements, as he is still human and as such capable of error, but I would advise doing so only if you have "facts" or "proof" that suggest he is wrong.

[Paul.K]

Dimmreaper,

I completely agree with you, however my point is I am only seeking help by posting in these forums. Some threads in these forums particularly seem to divert away from the context of the questions asked and instead revert to other issues. Now I do like this forum or online community very much but I am seeing a lot of the same debates and issues been fought out here. Please make this community for us (overclockers) be a friendly, knowledge gathering area, where people can join and ask interesting and puzzling questions concerning the art of overclocking.


Please for the sake of all of us!!!!!


Paul.K

[dimmreaper]

My comments weren't directed at you, or anyone else, Paul. And I apologize if anyone took it that way. I was simply retracting my previous statements about the possibility of these being a current stability issue, because I know EPoX Tech would tell us if that were the problem. Also, I was establishing some authority for EPoX Tech, so that we might all grow to appreciate how lucky we are to have him here. People need to be aware that EPoX Tech isn't just some clock punching tech support wiener, or somebody from marketing, but an engineer and an overclocker.

[EPoX Tech]

Paul -

Short of liquid or more ideally twin - TEC cooling, I think that you have already acheived a very respectable overclock with this processor indeed.

I only managed to break 1725MHz with Vapochill and now only 1892 with it with any stability.

I think that Athlon XP is a little strange in the way it behaves. Generally if a T-Bird would boot winodws you had some chance of maintaining stability overclocked whereas Athlon XP will sometimes throw you out just when you thought you were stable.

I think AMD is doing a fine job of bin selecting and limiting these processors so we don't see anything quite of the AXIA-Y or AHYJA ilk that would give the game away from a strategic point of view.

I think I may personally purchase an SK-6 with a focus flow in order to see what results I could acheive.

[Paul.K]

Epox Tech,

Just for your information my XP is been cooled by an SK6 fitted with a delta 80mm fan which pushes 60Cfm of air over the fins of my copper heatsink. I have also performed a mod on my case which draws cooled air through a duct from the side panel of my case over the area of the fan. Now my CPU temperatures using a probe are 29C-30C idle and 32-35C load. What I am really puzzled about is that I can run stable at 1787Mhz with Vcore of 2.03volts. But when I increase the FSB by 1Mhz without altering the Vcore I can load into WinXP , but when I run an intensive application I get random reboots. Hmmmmmm. This really puzzles me!!!! Hmmmmmmm!!!!!

[Road_Runner]

Originally posted by dimmreaper
I think we should listen to EPoX Tech on this one. Knowing for the most part exactly what his function within his firm is, and the amount of component testing he does, I can state with some certainty that in this forum he is the authority on hardware. I'm sure there are people more knowledgeable than he, but they are few and far between. Most of 'em are probably locked in a cage and presented with the task of engineering the next generation of computer hardware, and hence not able to post in this forum. Obviously it's still alright to question EPoX Tech's statements, as he is still human and as such capable of error, but I would advise doing so only if you have "facts" or "proof" that suggest he is wrong.

Well....I hear what you and Epox is saying and appreciate this forum. However, from what I have seen in this thread alone and at amdmb.com, hardocp.com, icronticforums.com, etc, the following things stick out like a sore thumb.

1. I have yet to find anyone running higher then 2.03-2.05 vcore stable with this motherboard. In every case I've seen or read about, going higher vcore-sise results in random rebooting for no apparent reason, or inabilty to boot at all.

2. There are two people in this thread alone that can pull their XP's out of their 8KHA+, and apply more voltage to the cpu in another motherboad configuration and overclock almost another 100 mhz.

3. I know and see many people running XP's at 2.1-2.3v albeit with good cooling. None of these folks are doing it on an 8KHA+..

I'm not trying to be argumentative guys but there are some trends here that seems to be being overlooked. I've just seen too many people run these xp cores at high voltages to agree with the argument that they don't like higher voltage. Most people I talk to say the exact opposite...that to maximize the overclock, these things love juice.

[dimmreaper]

Paul, I've found through my own super cooling experiments that the relationship between temperature and overclockability, is never a linear relationship. By that I mean that sometimes a 10C lower temperature may offer little or no more frequency, but a 20C lower temperature may offer a substantial increase in acceptable operating frequency.

I have found liquid cooling to only offer a small improvement with my Morgan. But at the same time I have found temperatures of 10C and lower to really wake this chip up, with sub-zero temperatures offering even more head-room. And sub -20C temperatures(dry ice) really get it cranking! I had this chip running at 162x10 fo over an hour the other day with temperatures below -60C(dry ice, and a 156W peltier, too cold for regular use as this would eventually damage the wafer).

[EPoX Tech]

Interesting - I think I may try to look at this a little further with consecutive XP dies from production.

[Road_Runner]

Jeff....your sig shows a Morgan core at 2.05v....

what XP are your running above 2.05? What total speed are you at and at what voltage is it running at?

Go to the overclockers forum at http://www.icronticforums.com/forumd...?s=&forumid=21

check out Oppainter's rig and experiences with the 8KHA= (stopped using it), MrIcee's rig, ddtung's, among others. Oppainter's a mod over there (very good overclocker). Check out the 8KHA+ reboot and overclocking threads at www.amdmb.com for others.

I don't remember seeing Proze's rig but I'll look his up.

/edit - where did dimmreapers last post go? he was refuting my 3 points.

[Road_Runner]

Originally posted by EPoX Tech
Interesting - I think I may try to look at this a little further with consecutive XP dies from production.

thanks Mr. Epox guy

[Paul.K]

Just from what I've been reading over on the icronticforums, several people are pumping 2.2volts or more into their XP's and reaching unbelievable overclocks with their XP1800+ processors. What most of them have in common is that they are running their chips on an Abit KG7 DDR motherboards. What does this tell ya??
Have a guess????? Hmmm could their be a fbc on the Epox boards that prevent high Vcores.........

[Road_Runner]

Originally posted by Paul.K
Just from what I've been reading over on the icronticforums, several people are pumping 2.2volts or more into their XP's and reaching unbelievable overclocks with their XP1800+ processors. What most of them have in common is that they are running their chips on an Abit KG7 DDR motherboards. What does this tell ya??
Have a guess????? Hmmm could their be a fbc on the Epox boards that prevent high Vcores.........

that's EXACTLY what I'm seeing. And btw, my 1800xp is the exact same stepping as Oppainter's. The guys overclock kills mine. I think he's running 2.3v on an Abit KG7.

[Paul.K]

Roadrunner,


What PSU are you using???
What cooling are you using???
What are your 5V,3.3V and 12V readings like????


Paul.K

[Road_Runner]

Originally posted by Paul.K
Roadrunner,


What PSU are you using???
What cooling are you using???
What are your 5V,3.3V and 12V readings like????


Paul.K

Enermax 430 watt (never an issue with two ther mobos running 2.1+ vcore.

Watercooled 1/2" ID system 35c idle, 38c load

As I type:

5v @ 4.92
3.3v @ 3.38
12v @ 12.46

and vcore is currently 2.03 (cuz she won't go no higher )