NF7 max fsb overclock decreasing??

[Stang_Man]

ok, i got a NF7 this monday, and been just trying different things to get my max fsb up.

well, with 14b, i was able to get 225x10.5 stable on tuesday.

come wednesday, 225x10.5 isn't stable, but can manage 3dmark2001.

come thursday, 225x10.5 won't even POST at all. 220x10.5 is all i can manage in 3dmark2001.

i've tried different dimm slots, L12 mod, different bios' (14b and d10)..

and i dont know what to do

AMD Tbred B 2100+ (1.95vcore, can do up to 2550mhz on chiller and 2.05vcore)
NF7 rev2 (stock cooling, 1.7vdd through bios, no vmods)
256MB TWINMOS PC3200 BH-5 (2.9V)
ATi Radeon 9800Pro
Antec TruControl 550 (with 12.18V, 5.06V, 3.31V)


what is causing the decrease in my max fsb overclock???

thanks

[Jrocket]

That sucks! Anyways you could try reseating your nb hsf or putting a little better cooler. Thats all i could really think of. If anyone else knows... it might be a bios bug or something.

[quicksilver98]

check yer PSU readouts in bios....

[KnightElite]

I've had this happen to me before. Hardware wearing out I think, though in my case it was a CPU that would no longer go the speed it used to. Are you overvolting it much?

[Evil_Spork]

could be damage to the cpu form heat and or voltage.

[RAMMIE]

If it's a newer board try d18 and only use 11.5 multi.You might be surprised!

[afireinside]

Probley the CPU taking the blows from 2.05V... How long has it been running at that?

[Stang_Man]

Originally posted by quicksilver98
check yer PSU readouts in bios....

measured with a multimeter this time..

12.41V
5.21V
3.39V


Originally posted by KnightElite
I've had this happen to me before. Hardware wearing out I think, though in my case it was a CPU that would no longer go the speed it used to. Are you overvolting it much?

most i've ran through it was 2.05V, and that wasn't for very long at all, maybe 5 minutes? it's been at 1.95V and 2.34GHz a lot though, which i think isn't much at all, and cpu temps always stay below 50C with this chip..


Originally posted by RAMMIE
If it's a newer board try d18 and only use 11.5 multi.You might be surprised!

it came with the d18, and the max i could do was 220x10.5


update:
215x11 = 2365mhz works fine...
but 225x10.5 = 2362.5mhz doesn't

gonna try the L12 mod, and see what it does..

i'm currently on the d10 bios.


2nd update:
227x10 = 2270mhz locks up in 3dmark2001
225x10.5 = 2362.5mhz won't POST
225x10 = 2250mhz doesn't lock up in 3dmark2001, gets 19076
220x10.5 = 2310mhz doesn't lock up in 3dmark2001, gets 19283

this is without the L12 mod.


225x10.5 was able to get me to 19467, but now it just doesn't like that 10.5 multiplier...

[saaya]

i got the same, flashing from bios to bios on my nf7-s 2.0 and the max stable fsb was getting lower and lower. i finally flashed back to the original bio the board came with and i think it got me back to the high fsb speeds again (not really high though, around 215/220 iirc)

[IamAnoobieCheez]

Originally posted by Stang_Man
ok, i got a NF7 this monday, and been just trying different things to get my max fsb up.

well, with 14b, i was able to get 225x10.5 stable on tuesday.

come wednesday, 225x10.5 isn't stable, but can manage 3dmark2001.

come thursday, 225x10.5 won't even POST at all. 220x10.5 is all i can manage in 3dmark2001.

i've tried different dimm slots, L12 mod, different bios' (14b and d10)..

and i dont know what to do

AMD Tbred B 2100+ (1.95vcore, can do up to 2550mhz on chiller and 2.05vcore)
NF7 rev2 (stock cooling, 1.7vdd through bios, no vmods)
256MB TWINMOS PC3200 BH-5 (2.9V)
ATi Radeon 9800Pro
Antec TruControl 550 (with 12.18V, 5.06V, 3.31V)


what is causing the decrease in my max fsb overclock???

thanks
Hey hey yo yo.. I guess I'm not the only guy with bad luck(extreme) ah?

There are few users with abit nf7s v2.0 that top out at blistering 190s FSB with exceptional setup.

I am one of them.


*Some* nf7/s v2.0 boards behave differently by flashing to different bios. *some* nf7/s v2.0 boards make ABSOLUTELY ZERO difference with bios flash. Some will make ABSOLUTE ZERO difference no matter what the hell you do to it... I'm talking L12 mod, Vdd mod, multipliers, bios flash, supercooling on the NB and SB.. I had some of these boards that make absolutely NO difference by doing ALL OF THE ABOVE. That's... world record I must say.. how pathethic the boards are..

When you try to push a dud board, you ain't going no where.. you will be stuck, period. I'm not saying you got the dud board but just saying what's out there.


One of my nf7s v2.0 boards performed worse with L12 mod, and no difference in flashing bioses.

All v2.0 boards I've tried supercooling on the NB, made ZERO gain in FSBs.

Overvolting the Vdd made it even worse. Kicks out to desktop even faster.. at some point, restarts, Nope... it's not the PSU..

First thing though, if you want to brag FSBs, I suggest you brag the MOST STABLE FSB in hours of gaming(most intensive 3Dgames, not just the 3dmark). You need to be able to pass the gaming non-stop, and here comes even more important part.. HOW LONG WILL IT LAST. That's the question.. it is just for one night thing?.. or two night thing?.. or randomly? like as if the board has its own mind and wanting to run certain fsbs on day1 and wanna do poorly on day 6?. You need to give it time frame of 1 month for checking stability, not 1 day or few days. I know how you feel.. you are a bit pissed... I feel the same with mine... When checking for stability on these hardware, you really gotta give plenty of time observing. The systsems tend to run good the first day, or two days at the max.. and then the third day, WHAMMMMMMM.... SURPRISE.. total failure. You can never trust these hardware... gotta give it a month at least.


My current Abit NF7 v2.0 system tops out at blistering 180 FSB and is completely stable in every damn thing. It is incredible.... (incredible my butt....) I had to lower the FSB all the way down to 180 to make everything totally stable. I really miss the old nf7 boards.. they performed quite well.. 220s~ 230 fsbs.


I have noticed I see more and more dud Abit NF7 v2.0 boards flowing around.. More boards I buy, the lower FSB oc'ing result I get. Eventually it will go down to 130 fsb and then down to 100.. or less who knows?...

If you have trouble running 3dmark at 220FSB, you would be LUCKY to get yours run totally stable at 200 FSB in games.. don't forget.. this is not a one night thing to observe.. It might fail on ya soon or later. All the sudden BAMM.


For now, I'm going back to the Via KT600 which does not give me trouble at 220 FSB in games. Too bad.. the Abit NF7 boards are made poorly lately.. They must be using some left over dud parts or refurbishing the parts that were returned by some oc'ers... the very early part of v2.0 releases are made better.. They just aren't the same as how it's made today. and of course.. the v1.0, 1.1, and 1.2s are hell lot better too. I had great experience with all of those. I was able to run mega stable at 220FSB in ALL GAMES with those old versions. Vdd mods and Superb cooling on the NB worked flawlessly with these boards.. This doesn't work with the v2.0 boards i'm sorry to say.. too bad.

[G|-|oST]

IamAnoobieCheez, I think you have a good point about a huge difference in performance one might get from two different 2.0 boards. Although I don't think _all_ old ones are better than _all_ new ones.

For example, my 'old' NF7-S 2.0 (plastic CPU lever) needed a Vdd mod and lots of messing around to get 220 FSB. And eventually it got very confused (thats the best I can describe what it was doing) and died.

The new one (metal CPU lever) does 230 FSB @ only 1.5 Vdd, but hates BIOS 10.

I just think that Abit should've assigned some more detailed revision numbers so that people could tell the diff between several board revisions. Like that whole deal with silver/black caps, etc.

Its just like buying a cat in a bag - you don't know what you got until you get home and open it.

[RandomDude]

Can't you RMA a rev.2 board that can't do 200FSB???

They're supposed to be guaranteed to do 200...

[Stang_Man]

is there a way around the OVP protection?

[Stang_Man]

bump

[CSOFT]

Originally posted by Stang_Man
is there a way around the OVP protection?

On the NF7? What OVP?

[Hell-Fire]

Originally posted by Stang_Man
is there a way around the OVP protection?

If you are talking about the OVP for the vcore, yes.

I think the controller is the HIP6301...pins 7 and 10 are the ones we are looking at. Same as many other Abit boards.

Pin 7 is for Vcore and pin 10 is the OVP pin. Put a 50k on both and adjust the pin 10 vr slighty as you push the Vcore up on pin 7.

Once you hit a wall on Vcore that causes shutdown, adjust pin 10 again. etc etc etc.

Obviously be careful with this one as you are adjusting 2 vrs for one voltage so to speak.

I think MrIcee put a guide up for several boards, and the controller on the NF7 is the same as those boards.