DFI Vcore modding weirdness, cock up uncovers something?

I've copied this post from one I made at DFI/Amdmb, I'm hoping that some of the Vmodding gurus here may be able to determine what I actually managed to do here:

I recently got myself an XP2500-M and it is totally kick-ass! I had it upto 2.7Ghz at 2 volts (well, 2volts DFI fluctuation style) and I still had low temps (<18 degrees) so I was convinced that if I gave it 2.1 volts it would hit 2.8Ghz.

Well this mean't doing the vmod. I decided to do the vmod only (no vsense) because I only plan on going upto 2.1 volts. Well, to cut a long story short, I made a pigs ear of the soldering 1st time round and ended up spending hours bleeding solder away from the 3 IC legs that I'd bridged by using too big a soldering tip (I'm actually quite a good solderer, but my prescision soldering iron is in a box due to me being in the process of moving).

Any way, I powered it back up after cleaning up the mess with the vmod fitted. Nothing :eek: :(

So I removed the Vmod and then noticed that Pin 7 had become unseated from the mother board, my bad. So I re fitted it and powered it up without the vmod. It worked :)

But, regardless of what I set the bios to, I could only get UPTO 1.85 volts. So I did a little research, I looked at a PDF of the chip (Sadly I do not have the link to hand, can someone provide it please) and found that the IC that DFI have used for the Infinity IS RATED FOR 1.85v operation. So I guessed that DFI have actually vmodded theier own board to get the 2 volts available as standard, because they are using sub spec components. Somehow, with my worst soldering episode ever, I had managed to diable the factory vmod and leave the chip at it's default capabilities.

My only option now was to attempt to fit the vmod again, which I this time managed successfully. I can now go almost as high as I desire. I saw 2.2volts before I got scared :eek: and wound it down to 2.1 volts which hovers now between 2.08 and 2.11.

The other thing I now noticed was that I have experienced none of the stabilty issues that guys have come up with for Vmodded DFIs, could it be that by inadvertantly disabling DFIs "mod" to make 2volts possible, I have introduced stability into the system?

I hope that some of the modding gurus can share some light on what I have somehow done. Any ideas?

and was it worth it? Oh yeah!

3Dmark at 2.8Ghz

this is very interesting indeed :eek:

I have been starting to wonder about the vcore fluctuations, kinda weird that they've modded the IC they've used :)

Quote:

---

Originally posted by bigZ
this is very interesting indeed :eek:

I have been starting to wonder about the vcore fluctuations, kinda weird that they've modded the IC they've used :)

---

Well, I'm guessing they have, because the IC is rated for 1.85V according to the PDF I looked at. I'm not on my rig right now and won't be for a week or so. I would be greatful if someone with a DFI Infinity/LP could check to see what the code is on the IC and find a link for it's PDF to link here, Thanks.

just to make sure, you guys arent finding these fluctuations with MBM or BIOS are you? onboard sensors suck. ive had three Infinitys and none had any problems with voltages when read on-board.

yeh, funny about their mod. im not surprised.

I'm using the Winbond hardware monitor. I've noticed that my voltage fluctuations seem a lot smaller since doing the Vmod.

Quote:

---

Originally posted by dobbz
just to make sure, you guys arent finding these fluctuations with MBM or BIOS are you? onboard sensors suck. ive had three Infinitys and none had any problems with voltages when read on-board.

yeh, funny about their mod. im not surprised.

---

I've been looking for a measuring point for vCore on the board, can you point one out to me? :)

well, try the mosfets under the socket. i forget if its the ones with the legs pointing toward the socket or the other three, but i get a 0.1v over BIOS VCore when reading them. migt be somewhere else to check but thats the only place i know of.

but Malves and macci both told me to check there and that the overage is probably accurate.

edit: i wouldnt go poking around those particular mosfets unless your hands are very steady at the time. i killed my last board shorting one of em. :D: whole machine shutoff and i smelled cooking, but components turned out fine.

theres some PDF's for the chip in this thread: http://forums.amdmb.com/showthread.p...ighlight=vcore

This has 3 pages on the AMDMB forums, where is ervery one on here :(

Errrr I don't get it at all could someone summarize this for me. Cactus jack just unsoldered Leg 7 and soldered it again to the board ? That's it ?

Check your PMs.

I accidently lifted pin 7 clear of the board when cleaning up the mess I made. The only resoldering I did was to put pin 7 back in place.

http://forums.amdmb.com/showthread.p...hreadid=312394

Linking the threads up to share knowledge.

Quote:

---

Originally posted by ricjax99
This has 3 pages on the AMDMB forums, where is ervery one on here :(

---

because noone looks in this part of the forum to talk... you shoulda posted this in the amd section for more replies even though this is about a vmod.. its not a tutorial or anything

Its the same IC as found on the Epox 8KTA3+, but i'd like to know what you did, too..

I'm running @ 1.9v bios (unsure what that is real, dunno if its a good idea to measure @ the choke and the mosfets dont give me numbers anwhere even NEAR that) and haven't had stability issues yet, but i'd really like to get 2.05v stable as i'm sure that would let my CPU hit 2500mhz...

Quote:

---

Originally posted by DisposableHero
because noone looks in this part of the forum to talk... you shoulda posted this in the amd section for more replies even though this is about a vmod.. its not a tutorial or anything

---

Erm, I did post it in the AMD section and so did somebody else.

Does anyone know if there is another site I could ask at? There doesn't seem a lot of interest in this finding here.

Hiya,

I can take a look at this for you.

Can you please summarise EXACTLY what you have done to the board.

You said you removed solder from 3 legs.. are these all now stuck back where they should be?

It might be that you damaged something when you powered the board with pin 7 lifted.

if you guys just want to manually set up the 6301 then its pretty easy to do. This may be a better option than trying to mod the existing DFI setup.

Quote:

---

Originally posted by Holst
Hiya,

I can take a look at this for you.

Can you please summarise EXACTLY what you have done to the board.

You said you removed solder from 3 legs.. are these all now stuck back where they should be?

It might be that you damaged something when you powered the board with pin 7 lifted.

if you guys just want to manually set up the 6301 then its pretty easy to do. This may be a better option than trying to mod the existing DFI setup.

---

Woohoo! Cheers Holst.

Right, The solder I removed from the 3 IC legs was sloder that got spread by me in a moment of clumsiness, the only leg that lifted was Pin 7 and you make an interesting point about powering up when the leg was lifted. I'll have to have another look at the PDF. I think what people would like to do is see if it is possible to safely disable DFIs own 1.5Vmod when fitting their own. I'd love to be more consise, but I'm in the middle of something right now. I'll get back with more details if you want them,

Cheers, Gaz.

Any news here?

I'd really like to know if lifting the pin 7 disables DFI's mod or not, or if it's been a short circuit between the 3(2) legs connected with solder...

Nothing I'm afraid, lifting pin 7 only enables "refusal to boot mode", so it's best to leave it attatched, lol.

Over at this thread at Amdmb it is being speculated that using a pin or wire mod to raise the default core voltage seems to cure the high volts stability problem. It is being suggested that the default voltage level dictated when OVP kicks in.

Quote:

---

Originally posted by Cactus Jack
Nothing I'm afraid, lifting pin 7 only enables "refusal to boot mode", so it's best to leave it attatched, lol.

Over at this thread a Amdmb it is being speculated that using a fin or wire mod to raise the default core voltage seems to cure the high volts stability problem. It is being suggested that the default voltage level dictated when OVP kicks in.

---

Thanks Gaz :D

I did the 1.7...V defaulting mod and gonna try later if there're still problems.

1.7v defaulting mod?

Quote:

---

Originally posted by STEvil
1.7v defaulting mod?

---

For mobile... wire mod so it defaults to 1.725V instead of 1.45V, should stabilize things a bit and reduce boot up probs...

Quote:

---

Originally posted by SAE
For mobile... wire mod so it defaults to 1.725V instead of 1.45V, should stabilize things a bit and reduce boot up probs...

---

Please let me know if it improves stability at high voltages!:slobber:

oh, ok... duh ;)

Quote:

---

Originally posted by Cactus Jack
Please let me know if it improves stability at high voltages!:slobber:

---

No. Seems there's no more stability by doing this mod... :(

Vcore and high fsb don't wanna go together... at least not superpi and prime stable...

Can you please make a step-by-step how-to for me? I need all details of your VCore modding adventure ;)

Please PM me, Gaz :D

Maybe, if everything else fails I 'll try your way to get it going.

Well, my LPuB had a completely repeatable coldboot problem, classic.

I did wire mod to the underside of the MB...did L12 same time... and the cold boot problem is TOTALY GONE. Works like a charm now.

It belongs in the sticky for DFI boards. NF7 doesnt have this because they store the Vcore in a way that stays during coldboot, so when you set it to 1.85V-etc, it BOOTS with that always. DFI seems to boot at CPU's native default, and at 1.45V with 200Mhz fsb and 12.5 multi..... it AINT gonna BOOT :slobber:

Quote:

---

Originally posted by uwackme
DFI seems to boot at CPU's native default, and at 1.45V with 200Mhz fsb and 12.5 multi..... it AINT gonna BOOT :slobber:

---

Hmmm....Don't know about that one as I'm running @ 2620mhz (no mods). Could that be bios related?

This cold boot issue was fixed in more decent bioses, IIRC ;)

new-b - maybe you started with more recent DFI bioses or simply did not used with the old bioses mobile CPU? :p: ;)