FSB 400 will not work on P5K-E?

A

AmongTheChosenX

Supreme [H]ardness
2FA
Joined
Sep 24, 2007
Messages
7,302
I just do not understand it!

E2180, 400 FSB x 8 multiplier, 1.4375 volts, 1.4 on Load with the Vdroop effect.
Prime95 freezes after about 2 hours on a small FFT test. No BSoD, just a freeze.
I'm running the 0602 BIOS.


but heres what gets me:
at a 320 x 10 it runs great.
but 400 x 8 runs great and more stable, until it runs prime for 2 hours, and then it fails.


Why doesn't it work? What can I do?
 
may just be that you're chip has reached it's max fsb. it may also be your ram that is failing, what is the speed you're running it at?
also, what are the other voltages? you can try bumping up the FSB voltage a bit too
 
I love when people come in, post no information about what they're running (voltages etc) and expect a magic fix-all answer.



Well, here's the magic answer: every chip is different, and just because someone else can run X settings...doesn't mean yours will.






And that's really about the best answer you'll get until you start posting details.
 
like it runs just as good, sometimes better, until its been put under a load for like a few hours, THEN it goes unstable and fails.

anyways,

I tried getting the voltages to the following:

Vcore: 1.4875
RAM: 2.2
Northbridge: 1.4
Southbridge: 1.05
DDR2-800 Speed, 4-4-4-15 timings.
 
thanks for the hekp, but like i said, it isn't stable at 400.

why would I want to run it at 401?
 
A memory divider is when you have reached the maximum speed of your memory, so you cannot run it 1:1 with your FSB. So, you slow it down. That is a very basic explanation of it. Mainly used when you have generic or slower memory.

As for the overclocking, chances are that the chip has reached the limit. Do try higher speeds though, maybe jump 20, just in case there is a FSB hole there.
 
lamboman said:
A memory divider is when you have reached the maximum speed of your memory, so you cannot run it 1:1 with your FSB. So, you slow it down. That is a very basic explanation of it. Mainly used when you have generic or slower memory.

As for the overclocking, chances are that the chip has reached the limit. Do try higher speeds though, maybe jump 20, just in case there is a FSB hole there.
Click to expand...

i was asking him what memory divider he was running, but the explanation should come in handy for someone searching it:p

@ the OP...my OC for my c2d never ran prime95 stable...but anything that ever stressed it (dota, crysis, counter strike, media encoding) always ran smoothly and never failed. if you've exhausted attempts at making it prime stable, just try it out on everyday basis and see where it goes from there
 
omz said:
@ the OP...my OC for my c2d never ran prime95 stable...but anything that ever stressed it (dota, crysis, counter strike, media encoding) always ran smoothly and never failed. if you've exhausted attempts at making it prime stable, just try it out on everyday basis and see where it goes from there
Click to expand...

I have the same experience. My overclocks are never Prime or Superpi proven, but I got loads of machines rendering animations 24/7 at 100% load stable. After a while I just take it as if it does not fail after 24 hours on intended sress use there is no point brooding over it and get on with what you need to do, If you want more security, get to the max overclock and reduce 5% from you overclock (not overall speed). I think most software/games do not demand as much as what Prime will do.
 
OK lol. I have actually found Counter Strike to be a relatively good aid for stability checking. But most intensive stuff won't be as heavy as Prime so there isn't really much point using it for tests, just as the other guys have said.
 
AmongTheChosenX said:
I just do not understand it!

E2180, 400 FSB x 8 multiplier, 1.4375 volts, 1.4 on Load with the Vdroop effect.
Prime95 freezes after about 2 hours on a small FFT test. No BSoD, just a freeze.
I'm running the 0602 BIOS.


but heres what gets me:
at a 320 x 10 it runs great.
but 400 x 8 runs great and more stable, until it runs prime for 2 hours, and then it fails.


Why doesn't it work? What can I do?
Click to expand...

Have you locked the PCI speed at 100MHz? Memory divider set at? Have you tried setting looser memory timings (ei. 5-5-5-18)?
 
Darth Bobo said:
Have you locked the PCI speed at 100MHz? Memory divider set at? Have you tried setting looser memory timings (ei. 5-5-5-18)?
Click to expand...

i was gonna suggest the same, loosen your timings some and bump the fsb 1-5mhz

i kno my e6600 it wont run fsb 380 stable but if i take it over 400+ even as high as 450 same timing and voltage it will run stable.

btw another question what are the temps? do you have the c1e feature enabled?
if so disable it see if it makes a difference.
 
wow thanks so much for the help!


i am going to see if theres a FSB hole

the BIOS does not let me choose a slower speed other than DDR2-800 with a 400mhz FSB. 1:1 is the slowest ratio it allows me to use. :(

I did try looser timings, but not with the 400FSB, as it was only for oerclocking the RAM in general. maybe I should do that as well.

I have the PCI-Express locked at 100. theres no PCI slot speeds to work with in the BIOS though.

PS: Temps are very low. Under prime load it peaks at 45C on rare occassion, and usually runs at 44C.

So I'll check for the FSB Hole and move the FSB to 401 and see if it works, maybe loosen the timings a little more.
I have the C1E enabled, but last time I disabled it I got tons of BSoD's so idk. maybe i'll give it another shot.


PPS:

Sorry i forgot to mention some wierd behavior that was happening with the 400FSB.

Asus Suite would suddenly not show that the system was throttling down when it was idle. with the 10X multiplier it never did that. the 9x didn't do it either. Could the C1E be doing this?
 
I see no actual temp numbers. You're shoving a shitload of voltage down that chip's throat for 3.2GHz. I have a Q6600 running at 3.3 with 1.35V on a P5K-E. My e4300 does 3.2 at stock voltage. WAY too much juice IMHO.
 
mwarps said:
I see no actual temp numbers. You're shoving a shitload of voltage down that chip's throat for 3.2GHz. I have a Q6600 running at 3.3 with 1.35V on a P5K-E. My e4300 does 3.2 at stock voltage. WAY too much juice IMHO.
Click to expand...

well...his chip may just be a crap overclocker...my e6600 can't do too much at it's stock voltage, but it can hit 3.6ghz at 1.46v which is pretty good for a later e6600.
i can't speak for the lower end conroe chips, but you're voltage seems to be okay for that chip considering the temps are fine and you're not 1.5v+
what speed ram are you running? if the stock speed is DDR-800 (PC2-6400) then that should eliminate the ram being the problem. if you still cannont run at FSB 400 with all the tweaks mentioned, you can always try putting in different ram to see if your sticks are the problem
other than that, just keep trying and, hopefully, it will come...overclocking isn't a short process:p
 
are you watching the temps right before it fails? you could just be over heating.
 
the voltages are so high because I am working with the vdroop feature. I crank the juice higher so that it drops in windows.

thhe BIOS may be set to 1.4875 but windows only picks up 1.44 at Idle and 1.4 at load. so its safe.


and because the voltages after the vdroop are quite low, the temperatures are:
idle: 25-27
load: 43-44, peak 45.

also, I was able to get 3 ghz with stock voltage, 1.3125.
I did bump it up to 1.325 to accomodate Vdroop.


note that I can work without Vdroop coming into the issue, but Vdroop causes my temps to shoot down like 8-10 degrees Celcius, so i see it as more benificuial to work with it rather than against it.
 
does your board have a setting that accomodates for vdroop? i remember hearing about something like that on some boards...
it may well be the vdroop that is causing you're instability. if you want to, you can bump that voltage up more to see what happens. if it does turn out to be the vdroop, you can look into a mod or a new board
 
well if anything working with the Vdroop has made my OC even more stable then before, so i don't think thats causeing a problem.

you have to remember that without working with Vdroop, windows wouldn't even boot with a 400 x 8 setting in the BIOS, even with like 1.5250 volts. there was no way i was going higher.

and I just bought this board I don't wanna bnbuy anotehr.
 
AmongTheChosenX said:
well if anything working with the Vdroop has made my OC even more stable then before, so i don't think thats causeing a problem.

you have to remember that without working with Vdroop, windows wouldn't even boot with a 400 x 8 setting in the BIOS, even with like 1.5250 volts. there was no way i was going higher.

and I just bought this board I don't wanna bnbuy anotehr.
Click to expand...

wel...if it won't boot up with that voltage...then you've probably reached your max fsb:rolleyes:
 
yeah i can hit 3.33 ghz @ 1.5125 volts guaranteed. I've done it before with the voltage damper enabled (no vdroop).


but my temps are lower with the vdroop in effect so I'm going to try and set the voltage to about 1.5375 and see If I can get 333 x 10. that'll probably be my max overclock at the lowest temps possible.


I think i just cannot get a 400FSB. oh well at least I can still get that speed with the highest multiplier.
 
Ok i investigated my issue with this 400 FSB.


I tried a 395 x 8 combo at 3160 mhz, and it ran prime for 20 minutes fine. stable enough windows booted and performed quicker too.

tried 401 x 8 after that and windows booted fine, didn't seem too stable though. Prime froze the whole system in under 2 mins. Tried upping north bridge voltage to 1.4, same result. (Ram was at 2.15)

Now I'm at 395 x 8 and its running prime just fine (as I am typing). windows performs great, boots up quicker too. running prime fine right now with the Voltage damper enabled, Vcore set at 1.4375, AI Suite reports 1.4160 and Everest reports 1.42, it stays whether at idle or full load, passes OCCT for 30 mins.

somehow my temperatures have dropped from 44-47 load to around 40 with a full load of prime. extremely WIERD!. I can't understand it, other than maybe that the northbridge is picking up some of the processors work, so the processor doesn't ahve to work as hard, but thats just my speculation. the temp sensors might be bugging so i'll check it next time i reboot.

The ram is also being set at a 1:1 ratio now, DDR2-798.



So heres what I've come up with, please bear with me I know its alot.

its not a Vcore issue because I can do 1.5125 Vcore and hit 3.33 ghz @ 333 x 10.
its also not a processor limitation ON OVERALL SPEED because again I am able to hit 3.33 ghz with 1.5125 Vcore. so adding more Vcore wouldnt solve anything

The northbridge over volt seems to have zero effect at this point. it doesn't add stability like I thought it did. when I thought it was the system ended up freezing anyway so the North bridge is not the problem.

so I've brought it down to the following:
Its either:

The FSB is maxed at 399, and cannot do 400. It could be either the processor can't do it, or the board doesn't want to do it with this processor. maybe i got a bad chip.
as far as I can tell, i see that E2160 overclocks never get past a 400 FSB, so maybe its possible that anything over a 400FSB on these chips is not possible.

The other possiblity is that my RAM is limiting my overclock. These Platinum revisions don't OC well in my system. maybe they dont pair well with the E2180's
I'll have to try my crucial ballistix when I get them and see if they work to improve the overclock. I have a feeling they will, and it could be that the RAM cannot handle the 400 FSB (which is odd cause theyre rated for 800)

Note that I only have 2 Gigs of RAM right now as I sold my other 2 gigs.


I;m still so :confused::eek: as to why my processor is only at 38-39C under a prime load! ?!?!?!?

EDIT: I am going to try a 401 x 7 and see if it runs stable. If it doesnt then i guess tehres a FSB wall that cannot be broken. If it does then i don't know what to do. Im guessing its the RAM at that point, but thats jsut my guess.


EDIT: 395 FSB was stable, not 399
 
somethins up with your temps, i just built a machine for a family member

e6300 @ 3.04 1.42vcore cooled with the same gemini II idles 43c loads 54c
2gb drr800 gskill @ 435 fsb
gigabyte p35 ds3l

now it could go higher but the temps wont allow it.
possibly your chip or your memory has reached its limit, from the sounds of it, it may be the limit of your memory
 
yeah i think the temp sensors bugged out or the room was colder than it usually is (all 5 computers were off but mine, and it was like 2 AM, and 30 degrees outside so the room was chilly).


at 3.2 ghz i load at like 45 C with a prime load. an OCCT load puts it at like 53 i think.


as for the FSB limit, I need to try to get a 400 FSB with my new ballistix RAM, as my Platinum REvision 2 RAM might have been causeing the limit. although I'm reading that its common for E2180's to have FSB walls between 390 and 415
 
well possibilty your chips fsb wall is 395.

those temps sound bout right cause my system watercooled e6600 @ 3.4 1.35v core idles
at roughly 30-33c with window closed , if window is open sometimes will drop 3-4 deg,
on load does 40-42.
 
What bios rev are you running on your p5k-e? I found that my board was a crap OC that was very limited in terms of ram speeds, like ddr2 800, 1000, 1066 were all I had to pick from. With the bios update 0906 I was able to pick just about anything I could think of from 200-800 base clock which would give room for ddr2-400 up to 1600.

I would also suggest getting a 40mm fan and strapping it on your north bridge, mine was running at 80c without a fan on it and I was having issues with prime locking up within 3 hours. Once I put the fan on the NB the temps dropped down to the 50c range and I was prime stable 24hr at 400x9 fsb.
 
I'm running BIOS revision 0602. I can't indepentally overclock the RAM from the FSB. I wish I could do that cause I'd like to run my sticks at on OC without bothering the processor so i could see what theya re capable of.


alright ran the FSB at 400 and set the multiplier at 8x. Played COD4 online for about an hour and a half. no freezes, but there was mad stuttering, and it didn't load the game very quickly. so i'm guessing it is a FSB wall at 395 for stability.
I didn't try upping the north bridge voltage but last I checked it didn't do anything so i gave up on that.

an d the processor seems to be stuck at 3.2 ghz. Anything higher requiers insane amounts of voltage
for example:
I can run it at 3.2 ghz with 1.4375 volts and its like 45 load max with prime
to get it to 3.3 it needs 1.5125 volts for stability and it runs almost 10 degrees hotter, 53-54C load with prime. an OCCT load sends it to like 63C. thats way too hot, and too much voltage.

so ehres my max OC's for everything so far:

E2180: 3.2 at 1.4375 Vcore
FSB of E2180: 395 (400 can run, but is not stable)
RAM: haven't checked yet
8800GT: 700/100 on stock cooling.
 
AmongTheChosenX said:
duh. that why i said SPECULATION.

read and think before you speak.
Click to expand...

This is priceless.



Anyway, you've been told the answer multiple times in the thread, yet you've ignored it.

1) The chip runs at 3.2 just fine.
2) The FSB is fine up to 395.
3) The FSB is not fine above 399.


THEREFORE, you are FSB limited. People tell you this and you continue to whine and conjecture about all kinds of random and incorrect things. You asked the question, you just don't like the answer. Guess what? It's not going to change either.
 
first of all the post about the northbridge i said before was my bad. yes it was stupid. i'm still fairly new to OC'ing.

secondly, i never said i didn't like 3.2 i'm quite happy with 3.2

thrid yes I know the FSB is stuck at 395. fine by me too. the freakin chip only cost me 90 bucks so i think it was well worth it.
heck even 3dmark thinks I have an E6850
 
You must log in or register to reply here.
Back
Top