Sign in with
Sign up | Sign in
Your question

Q6600 Mid VID, Supprisingly reached 3.8 at Spec 1.5v, 8 Hour + Stable!

Last response: in Overclocking
Share
May 17, 2008 1:30:31 AM

I take back my earlier statement that Mid VID chips couldn't reach 3.8 and stay prime stable for 8 hours +.

Due to my 1.2000 VID being temporarily off line, I used my second best chip that was a Q6600, a 1.2500 VID chip.

After tweaking the night away playing with GTL Vref settings and junk, I am rather surprised to see that it is possible to stay within intel specs and reach 3.8 and be over 8 hours Prime 95 small fft stable.

Who would have thought that by studying up on the processors signaling specs, you could actualy modify what it accepts as a 1 or a 0, and by watching how the ringback effect happens, you can modify it accordingly, and eek out some decent speeds at lower than expected Vcore voltages!

(If 3.8 sounds easy on a q6600, go give it a shot and stay prime stable for 8 hours! (And run some good RAM timings while yer at it!) You'll find that it's not that easy, or everyone would be doing it! Look how many people have a Q6600.)

Here is a link, since I posted the results in the OC competition thread already, and I dont wanna spam all over with pic links!

This chip is on its way to E Bay here soon, so if anyone is interested...

http://www.tomshardware.com/forum/page-246979_29_40.htm...

--Jedi Master Lupiron!
May 17, 2008 1:36:16 AM

Oh, the actual high core temps were 73c, so that is a 24/7 stable OC, or not to bad of one.

I have no idea why, but the temps at like 68c, decide to go 68, 75, 68, 75, then settle at 69c. Dunno why it does that. The colder pair of cores did it as well, at like 50c, it went 50-63-50-63, then turned to 51 for a while, but it went over that one, so it doesnt matter.

Anyone know why they do that? Computronix? Where are you when I need you? Hehe, later!

--Lupi
May 17, 2008 1:41:57 AM

are there any retailers that give the cpu VID before purchasing. My q6600 VID is 1.35v. 1.5v will get 3.6ghz, but not very stable. Takes over 1.6v for a very shaky 3.8ghz. Runs nicely @3.4ghz though
Related resources
May 17, 2008 1:47:28 AM

Lucky bastard. I hate you. I hope your cpu dies. :p 

My q6600 needs 1.6v for 3.8ghz, and 1.65 for 4ghz. It's not fair.
May 17, 2008 1:55:31 AM

My Q6600 sees to fail Prime at 3.6 @ 1.375ish but it 100% stable in games, never crashed even after 5 hour gaming binges.

I guess it doesn't matter cause games don't come close to full loading the cores
May 17, 2008 2:07:02 AM

Hehe, yeah. I wish my 1.2000 VID would comply. The IHS was lopsided, so I had to do some serious lapping. Its almost flat now!

And nope! No retailers give out the VIDs, nor does the box list them. Because it doesnt have to, they all will run 2.4ghz all day, no matter the VID. I will admit that I took a flash drive into an office depot or two, lol, and got a display model! 2 chips were 1.2250, which would be awesome! And the last one was 1.2000! Which would be awesome! If it wasnt lopsided, lol!

But if you know someone who sells OEM chips, maybe he will let you "test" one or two before you buy! And fair warning! The VID of the past 6 moths worth of chips have been 1.3000 or higher!

Get OEM if yer gonna buy one!

Or get to smooth talking me out of this one, hehe!

--Lupi
May 17, 2008 2:11:35 AM

Hey Clown! Its 1.3500?? Wow, thats nuts. On the new Core temp? That means that intel decided to toss a pair of duo throwbacks into a quad die! They had to add that VID recently for the Throwback "e" series dual cores. All the "e"s I scanned were 1.3250 or higher, lol!

And Horendus, whats the stock VID in the new core temp of your chip?

Also, you are correct in assuming that you can be stable, and not be prime stable. I have done the same thing in the past! But you can also be prime stable and BSOD during a game, because you are putting a different kind of drain on your system.

--Lupi
May 17, 2008 2:32:56 AM

sorry...1.325v VID on my chip
May 17, 2008 2:41:21 AM

Thats better, as 1.3250 is supposed to be the max. But they really did add 1.3500 to core duos.

You'll need around 1.45volts Loaded for 3.6. Mind me asking what MoBo you are working with over there?

I did manage to tweak mine down to like 1.42 loaded, but that took some work too! I am a Lazy bastard!

--Lupi
May 17, 2008 2:57:57 AM

P35 DS3L. Takes 1.5v for a solid 3.6ghz. That's as low as I managed to tweek it. Fails occt, but will run prime95 for at least a few hours. Stable for gaming though...no bsods. I run it at 3.4 though. It takes 1.425v for a very stable 3.4ghz. Guess I got the runt of the litter when it came to q6600s. Fast enough for me though. I had to add a fan to the NB on the ds3l to keep it cool @3.6.
May 17, 2008 3:25:53 AM

Thats a good speed. All you have to do is look at the price of your chip, and then the 3.0 and 3.2 Ghz ones, and you see why. ;) 

You'd have to look and see what yer VDrop and VDroop are that that board. Its easy. Set the VCore to 1.3250, the factory VID, then boot into windows and look in CPUz to see the idle VCore. That will reflect your direct VDrop. Then when you run prime 95 on all 4 cores with small ffts, wait a min for it to run, then note the lowered VCore in CPUz.

And there you have then!

--Lupi
May 17, 2008 3:33:33 AM

@1.5v droops to ~1.425v under load
@1.425v, droops to ~1.36v under load
May 17, 2008 3:42:45 AM

Damn nice overclock Lupiron. Makes me wish i hadn't bought a 680i board :cry:  The damned systin temps reaches 70c under prime95. And that's only at 3.45ghz ><
May 17, 2008 3:52:23 AM

Clown, thats quite the drop/droop combo! That hurts an OC, because it forces your chip to idle really high, just so it can load up to that lower Value. I cheated. My Maximus formula has a 0.250 Vdrop and a +0.008 Vdroop with LLC enabled!

Thanks Lucuis! And yeah, those 680I boards are very picky! I had one, and sold it, lol!

--Lupi
May 17, 2008 3:57:09 AM

Lol yeah, i'd gladly sell mine. I'd love to have an ASUS P5k-E or one of the newer X38 boards...mmmmm

ooo, /drool, that LLC sounds sexy.
May 17, 2008 4:18:58 AM

To stay within intel specs, I had to use my Formula. The P5k-e wifi is also an excellent overclocker, but it still has double the vdroop/vdrop combo of the formula. The P5k-e hits a stable 475 FSB, while the maximus is stable at 450ish.

The formula is double the p5k-e wifi price! So take yer pick!

--Lupi
May 17, 2008 4:22:03 AM

I'm not sure of the difference between vdroop and vdrop to be honest. When the Vcore is set to 1.425v, it idles at 1.38v and droops to 1.34v. (just revised those numbers with a quick OCCT run).

Besides a pencil mod (which I don't care to do) is there anything that can be done to lessen it?
May 17, 2008 4:36:09 AM

Well, some boards have an option that lessens the effect. On Asus boards, its called Load Line Calibration.

VDrop is a specification calculated by intel that makes sure your VCore voltage can never spike over the VID of your chip. Thats your idle windows voltage difference from BIOS VCore setting.

VDroop is when your processor is in full use, and say... you terminate the processor related to the usage. The power supply can not instantly release that current it is sending to the Processor, so VDroop is in place to prevent a spike in power! It lowers the actual VCore usage a bit, to counter the effects of the power spikes.

Pencil Mods can be erased with your finger! Why not try it. Just be sure you pay attention to the few small areas you are to basically draw a line to with a pencil.

--Lupi
May 17, 2008 4:48:40 AM

I wouldn't trust a pencil mod. I tried it on mine and it did nothing, i later learned that if the mod had worked it would have severely stressed my system.

Your essentially creating a partial short...with graphite. Needless to say, your once clean voltage is now very dirty.

If you wanted to do it "properly" you'd have to removed the desired resister and replace it with a very fine potentiometer. Like a Trim Pot. And you'd better be an excellent solderer. Even if you did all that successfully there's still no telling what adverse affects the mod would have on your system over time. And i'm not talking a long time either, it could very well be a short time before things go wrong.

However if your willing to risk it, why not. Hell my curiosity made me do it..lol.
May 17, 2008 5:08:26 AM

well...too late. I busted the side panel open and went in swingin with a No.2. Damn resistor is microscopic. They tried to teach me to soldier this small crap in a military tech school...didn't do so well. Looks like the immediate effect is about a .015v improvement in the vdrop/vdroop. I'm gonna run the chip@3.6ghz for a while and see how it does. Idling @ 42c with my super ultra sweet arctic freezer pro 7, and loaded up to a resonable 66c. Warranty is busted on this chip anyway since i lapped it. Might as well burn it into the ground and hope my next one has a lower VID.
May 17, 2008 5:29:39 AM

LOL, nice. And thats true, look at the price of one. I am gonna sell this one for about retail, so before you buy one!

Hehe, and I guess you can re trace the pencil mark once or twice and it may add a bit more. I dont know. Just be sure to read up on it first! I only glanced over them, because I thought ahead on my boards, hehe!

Let me know how it goes!

--Lupi
May 17, 2008 5:44:00 AM

I blew the dust off the old multimeter before I went crayola on it, and made sure I didn't completely short it out or anything. I dropped the resistance by ~50%. Fine tuning it more is gonna require lifting the mobo from the case, so I'll probibly be too lazy to do that. And I'm certainly not brave enough to solder in a potentiometer.
May 17, 2008 6:08:13 AM

Well i'm glad your mod worked, still don't understand why mine did nothing.

And don't worry, there's very few people that can confidently solder on projects as small as a computer motherboard, so don't feel too bad.

Hell i know i can do it, but i still wouldn't on anything expensive. It's just too damned risky.
May 17, 2008 6:32:26 AM

Me either, but I am glad it worked out. I like seeing what tricks come in handy! Its quite fun, really! 50% less is quite a little gain in VCore stability.

--Lupi
May 17, 2008 7:53:13 PM

Of course when I place an entry, Shadow takes the week end off!

Waaaahhhh!!

--Lupi
a b K Overclocking
May 17, 2008 8:41:59 PM

Updated on my thread.
May 18, 2008 9:05:34 PM

:) 
May 19, 2008 12:56:28 AM

Yes, Going back to the point about VID's. My e6750 is currently running at a VID of 1.3500. Now many might think that is a very high VID, although I must concur that new VID's must be BS. I believe it is Intel trying to play it safe, and just branding all of the new CPU's with a high VID. Because my proc can run at 1.175v stable at stock clock with a VID of 1.35? Now I overclock it to 3.2 and it runs stable at stock VID of 1.3500 running prime95 for 8h.
May 19, 2008 1:30:49 AM

Makes sense. It would be like Intel to cover it up like that.
May 19, 2008 1:50:39 AM

Yup, Already thought of that, so it will take some extra testing! And since I know what 3.6 should be, thats where I would start the tests, because I have so many VID chips I have shots of them at their speeds!

:) 

I just try and get validation before I blab about stuff! But I would at least do that if I was intel! Label them all the same because it doesnt effect their stock performance.

--Lupi
May 19, 2008 6:26:34 AM

Anyone happen to notice that my 95/105 watt processor is outputting 140.xx watts of power? Thats over a 33% increase in output. No wonder its hard to overclock! Look at that draw on VCore currant!

--Lupi
!