Over 1,9v on Abit NF7-S...

[ray]

hi xtreme forums,
i'm a new user!

i'd like to expose you my problem.
I have a 1800+ tbred b on an abit nf7-s; my best overclock is 2641Mhz with 1,9v (1,85v effective), but if i increase the voltage, there is not boot in any case, also if i downclock my processor.

why??

My configuration:
Abit nf7-s rev 2.0
athlon 1800+
256Mb PC3500 geil platinum ultra
enermax 465w
geforce 4 ti4600

and a liquid kit for cpu e gpu.

many thanks.

[iceman2g]

You might wanna increase your 3.3V rail if that doesn't work disabling APCI in the bios (it'll lower performance though).

WELCOME TO XTREME

[ray]

thanks a lot iceman2g!

my 3.3v now is effective 3,31v (with winbond); it's low?
i'll try to disable apci, thanks!

[TheDude]

Welcome to Xtreme ray :D
:toast:

[Hell-Fire]

Is there an overvolt protection on the NF7? Also, whats the stepping on that 1800+ of yours...thats a nice o/c for a 1800+ man.

Not to mention that if you are pumping lots of volts thru that cpu of yours, you need decent cooling, it may be shutting down due to heat. You have to have good circulation thru the case and on the cpu.

What fsb are you running and multi? I agree with the 3.3v rail idea, up that puppy to around 3.65 and see if that stables you out.

[faruquehabib]

welcome to the sickness!!! and yes, hell of an oc on an 1800+

[Keeper]

I have the same problem with mine using a 1700+, won't post over 2volts. I don't know if it's the MB or the CPU, I get it into Windows at 2800MHz but it needs more voltage to stay stable for benching, as soon as I up it to 2.1 no post, and random crashes at 2.03.

[PiLsY]

Same problem with mine over 1.825v. 1.85v will post 75%, 1.875 25%, 1.9v never.

Better than my old board though - that one was unstable using anything over 1.575v - nightmare!

PiLsY.

[TheDude]

unstable using anything over 1.575v

Pilsy...is that a typo? Are you serious? WoW! Even 1.85v is bad. I just killed one of my Nf7-S boards (hand slipped/shorted it) and the replacement has silver caps while my old ones all had black Rubycon caps. Has anyone noticed this on the latest boards? I will try to push it today on air with a sweet 1700 and a 510 PCP&C Psu. If it does well, I will try it in a prommie. My old board took my last 1700 up to 2900 at 2v....may they both R.I.P. :D

[ray]

i have a waterblock for cooling the cpu.
do you want to know the entire initials??
for 2640mhz i put this configuration...155X17, but it is the same if i change the molti and the moltiplicator!

[Marci]

I've found that my OVP on my board is directly related to the Overheat Protection temp which can be set at bottom of SoftMenuIII.

If I have it set on the lowest option... no post at or above 1.95v
Next option... no post at 2v or above. Next option... no post at 2.05v or above etc etc...

Don't know exactly how or why this setting should affect things... but it definitely affects my boards' CPU Voltage limit...

[Evil_Spork]

thats not your mobo, thats your cpu. mine does the same. i have a ROCK steady 3.4v

i post at 1.825 75%, 1.85 25%, 1.875 about once. 1.9 ot sure if i ever got it. i run a dlt3c 1700, so we have the "same chip"

same board my 2100 dut3c wil run at 2.2vcore every time.

[Keeper]

I will check that out Marci, because I tend to place my temp protection at as low as it will go.

[Garrett]

I have no problem with 2.06V either, so it's the cpu if you ask me...
My cpu is in my sig...

[Keeper]

Mine is the CPU, over 2 volts and no post regardless of thermal protection. Too bad too, because it will do 2800 MHz at 1.98, with a little more juice no telling what it would do.

[PiLsY]

Nope - no type m8. 1.575v was the boards stable limit. Anything over and I had to drop cpu speed until 1.9v - then it matched clock with 1.575v.

Simply swapped boarda directly - nothing else changed and can now scale perfectly up to 1.85v and have gained another 200mhz or so. Still on the black caps here.

PiLsY.

[TheDude]

Thanks, interested to see if these silver caps make any difference. I didn't have any time to try it today. ;)

[ray]

i know that the problem is thc cpu, because i also have a barton but this problem don't shows!

but if i try to block the voltage on the socket, is it the same??

[PiLsY]

Marci is along the right lines - by setting every warning/shutdown temperature to maximum I can now use up to 1.93v with 100% post. 1.95v and above causes a no post though.

Ive also noted that the caps and 3 coils nearest the cpu are running at 60 - 80c under load at high mhz/vcore. Im wondering now about uprating the caps and coils. My voltages are rock steady, but as im using the mcw462 (really old blue top version) which basically touches the caps its heating the waterblock up to the tune of +8c cpu temp at high cpu wattage.

Also a note on temps - theres no difference for me in clocking ability at up to 69.5c cpu temp. As soon as it hits 70c the rig reboots.

Still, got my old 1700+ @ 2511mhz on straight water so cant really complain.

Dont fancy starting another thread on NF7 bioses, so i'll ask here. My max performance is at 218mhz, but I can do 223mhz. The board is currently running bios 17 - is there any other bios thats slightly more agressive at the cost of a few mhz fsb? Or a bios thatll loosen things up a little but gain me a few mhz? I need to use either 11 or 11.5 multis, so im wanting either 218mhz or 228mhz. Can currently manage 213mhz @ 11x.

Cheers!

PiLsY.

[LBJGH]

The power regulation on the NF7-S must be a little wonkie... The board I had would consistantly undervolt the Vcore badly. I actually lost o/c'n potential because I didn't want to exceed 1.85v with my air cooling but the Abit board would crap out at 1.90v and above. I went back to my old Epox 8k5a2+ and everything was 100% stable again... same psu/memory/cpu/cooling.