Q6600 overclocking, 1:1 fsb ram a must ?

Toggle sidebar Toggle sidebar
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
Hello, so i've started trying to overclock my CPU. Atm the min voltage in which seemed to be stable was 1.35v as i was writing this prime crashed. Temperature 57 degrees in coretemp and 47 in Aida64 and don't think i will be able to run it stable at a lower voltage. Now the fsb is at 333 x9 multiplier with 5:8 fsb:dram ratio and memory is at 1066 5-5-5-15 2.1v (how they are supposed to be running as the manufacturer say). Is it so important for the fsb:dram ratio to be 1:1 ? If so what would be the best way to achieve it: lower multiplier raising the fsb or leaving the fsb at 333 and lowering mem speeds ?
Will lower the ram speeds for the moment to see if that was the problem. Will update...
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
Update: atm running 1.35v on cpu, 2.1v on ram with 1002mhz 5-5-5-18 over 1h on blend test in prime no errors, temps around 57 in full. Seems to be ok as is it, fsb: dram 2:3; should i try lowering timings or it's ok like this ? Not sure if it will work since with 5-5-5-15 at 1066 the memory wasn't stable.
Complete pc specs: Asus P5KR, Q6600 G0 > Antec Kuhler 620, Corsair Dominator CM2X2048-8500CD, GeForce 8800GT (waiting for gtx670ti), Thermaltake W0101RE 550W
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
Will try 5-5-5-15 on 1002mhz, as with 1066 the memory is not stable. Now i'm curious of something: in windows i had memory rating of 7.1 now that i've lowered the mem speed, increased the timings from 5-5-5-15 to 5-5-5-18 i get 7.3. I don't understand...
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
Back at square 0 atm, i find it a bit strange whenever i try a prime95 blend i hear a strange hissing sound coming from something and the moment i stop it the noise stops. If i try a small FFT test (RAM not tested much) no more hissing noise so i've come to the conclusion the sound comes from the ram (also i heard the noise when i was downloading a torrent). It's the same when overclocked or not so could it be coming from the ram, did i actually break something i mean it's a new sound, i have a problem with strange sounds coming from my computer so i'm 100% sure it wasn't doing it before today. Tested out with memtest and no problems, prime no errors ....
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
This sound is driving me nuts and still can't find from where is comming, i've tested with 1 ram at first it was hissing but didn't after. So now with 1 ram it doesn't hiss but with 2 it does...
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
None of the fans are to blame, stopped each fan (including video card fan) and the noise was still there. No disc in drive, ran a hdd test with sisoft sandra no noise. The hissing sound is the same as the one made by a PSU that has a blown capacitor though it doesn't seem to be coming from the psu. The only test in sandra that makes that thing hiss is the Lan speed test.
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
The ram is off the list, it's probably a capacitor from the mobo or the psu. Funny thing i'm running Crysis 2 on ultra no noise, Diablo 3 beta nothing... i refresh tomshardware page i can hear the hissing for a split second. Hissing when loading a video on youtube, stops when done loading. Tested hdd with sandra, tried copying stuff around the hdd no noise.
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
It seems a bit strange it started doing that hissing today after the overclock when i haven't changed anything except the vcore and fsb (underclocking the ram shouldn't count) same thing i did some time ago when i made this pc (temperatures were to high to keep the oc). I mean did i actually do something wrong and broke something or it was just an unfortunate coincidence. Now about the psu screeching under load ... what load, loading a youtube video can't be called load. Crysis 2 on untra is a load on all but it's not hissing when playing.
Now let's say it's the VRM's on the board (do tell what that is) should i worry? Is it a chance the board will die soon or something ?
 
lowjack989

lowjack989

Distinguished
Oct 30, 2009
2,068
0
19,960
VRM's = Voltage Regulating Modules ...When you OCed it you could have strained the VRM's on the board causing tiny fissures ( cracks) in them same goes for the PSU caps could have been stressed to much. Loading a YT video causes about a 50% load to be put on the system at intialization of the YT video...Makes no sense that running Crysis wouldn't duplicate the problem..Thinking......
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
When i say loading a youtube video i don't mean the initial load, just the download part. Hissing starts when buffering and ends when video is completely downloaded.
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
I did say earlier i tested the hdd that's not it. Tomorrow i'll detach all the coolers except the one on the cpu's heatsink (will stop it when needed) and record. Hopefully i'll manage to get the hissing alone as i have tried it now but it's impossible to hear anything but the sound of the coolers. Would do it now but it's 5 am, gonna go to sleep. Btw guys tnx for the answers, hoping for others ideas maybe i'll manage to isolate the problem.
lowjack989 the vmr's are the best bet atm but should they break so easy just by increasing vcore from 1.23 to 1.35 (plus the fsb from 266 to 333) it's not like it was some aggressive OC.
 
jsc

jsc

Champion
Jul 14, 2004
16,893
6
52,960

1:1 gives you the best stability. Running the RAM faster will give you little, if any, real performance increase. Here is one place we talked about that:
http://www.tomshardware.com/forum/251715-29-ratio-myth

Motherboard VRM's are not that fragile. They should work past 1.60 volts will no problems.

There is actually very little DDR2-1066 RAM. Most -1066 RAM is simply DDR2-800 RAM that the manufacturer has tested at the higher speed using increased voltage (over 1.8 volts) and more relaxed timing. That is not the same as memory that was actually designed to -1066 specs.

A Q6600 should be able to run at 3.0 GHz with stock cooling and 3.3 GHz (367 MHz X 9) with even a mediocre aftermarket cooler. 3.4 - 3.6 GHz is generally the upper limit with good stability using a good air cooler..

Mine runs at 3.6 GHz at 1.425 volts in a G'bite P45 motherboard. Load temps are 61 - 66 C, 24 hour Prime95 stable.
 
U

unknownuser

Distinguished
Dec 23, 2007
35
1
18,535
Comp specs: Asus P5KR, Q6600 G0 > Antec Kuhler 620, Corsair Dominator CM2X2048-8500CD, GeForce 8800GT (waiting for gtx670ti), Thermaltake W0101RE 550W

Here's a recap of what i did: Increased the fsb from 266 to 333, increased the vcore to 1.35, tried stability with ram at stock 1066, 5-5-5-15, 2.1v. Prime crashed after it finished 1024k tests and started the second batch of tests 8k or something like that. Lowered ram to 800mhz, voltage on auto, timings on auto, prime ran 1h+ no errors. Increased to 1002mhz voltage on 2.1 and timings on auto (5-5-5-18) prime ran ~2h no errors so left it like this. NB voltage was on 1.45v now is on auto. Someone said it could be the pump which i find impossible, there's no way the pump could start making that noise the moment i start prime and stop instantly when i stop it. Not sure if the hissing is permanent and only starts louder when testing with prime. I think i'm hearing a faint hissing now but it could just be my imagination.

I've disconnected all coolers except the cpu's and psu's. I've recorded a video and you can hear the sound though there is a lot of other background noise. This is the best i can do atm: http://www.youtube.com/watch?v=yQ2uB4hbTfU hissing starts at 00:03 and ends at 00:41. Also i think i managed to narrow down the spot from where it's coming though not sure 100%.


If i press with my finger on the R68 the hissing sound clogged
 
jsc

jsc

Champion
Jul 14, 2004
16,893
6
52,960

No, I am not. I am just saying that they are not that fragile.

The maximum VID (core voltage that the CPU programs the on board power regulator to provide) by Intel specs is 1.50 volts. So the power regulator needs to provide at least that much in an LGA775 motherboard.

Could part of the power regulator have failed? Certainly, but I think that is pretty unlikely. OTOH, I have been wrong before. :)

R68 is a .68 ohm monolithic power resistor. There's really nothing inside to cause any noise so I do not know what is happening.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom