• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Vid Pin Wire Wrapping The Northwood

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

batboy

Senior Moment
Joined
Jan 12, 2001
Location
Kansas, USA
VID PIN WIRE WRAPPING THE NORTHWOOD P-4 FOR MORE VOLTAGE

by batboy (senior member) private message are welcome if you have questions or comments



Warning and disclaimer

Do this modification at your own risk! If this Northwood VID pin wire wrap procedure is not done correctly, your CPU is going to be an expensive piece of fried junk, which potentially could short out the motherboard too. Be aware, this will undoubtedly void your warranty. Don't blame me if you damage components or injury yourself. However, this procedure has been successfully performed and tested by myself and several others. NOTE: it is now known the overvolting Northwoods (like above 1.75v) for long periods of time can lead to the CPU dying.


What is VID pin wire wrap and why would we do this modification?

Well, this is a modification procedure of wrapping a single strand of wire around certain CPU pins in order to fool the motherboard BIOS into thinking the default CPU core voltage is higher than the original factory default. VID pinning or wire wrapping is done in order to overcome voltage limitations often associated with many motherboards. The idea is to increase the number of available vcore settings. A successfully performed wire trick is easy to undo in case you later decide to sell the CPU, all you need to do is remove the wire and it's back to normal.

Specifically, Intel internally connects different combinations of VID pins in their processors which sends the motherboard a default voltage signal. Most overclockers are aware that we often need to increase the CPU core voltage to help stabilize an overclocked system. The higher you push the limits of a computer, the more voltage you will probably require. However, the BIOS of most motherboards generally will only allow a certain amount of voltage increase above default which keeps "mental midgets" from unwittingly frying their CPU from too much voltage and not enough cooling.


Heat is the enemy

Overclocking your computer creates considerably more heat than normal. Increasing voltage also results in much hotter CPU and system temperatures. Too much heat can destroy computer components. Cooling is critical to prevent damage and to reduce the risk of electromigration, which is a fancy term to discribe the damage done to a CPU due to overvolting. A cool running computer will be more stable and your components will have a longer lifespan, which ultimately saves money and down time.


References and credits

I have some previous experience in doing this wire wrap trick to increase the default voltage with a Celeron 566 and a P-III 700. However, because I never VID pinned a Northwood P-4 before, I followed the procedures (with some modifications) outlined in an article written by Andy Owens (see below listed link) and a Japanese article (see below listed link). The photos and diagram used in this document were borrowed from these two articles.

http://www.hexus.net/review.php?review=278

http://www2.justnet.ne.jp/~kiti/Paso/00pen4-3/northwood01.htm


Actual VID pin wire wrapping procedure

The first thing you should do is to use a low "user define" BIOS voltage setting before beginning. Do not leave the BIOS setting at default or the computer will automatically boot up to the "new" vcore default the first time after you wire wrap the VID pins.

The second thing you should do is to unplug the computer power cord and ground yourself to protect against static electricity. Always touch the metal computer case to discharge any static prior to working on components inside a computer.

Now, you must decide how much of a modified default voltage you desire, keeping in mind that currently the real default in the Northwood P-4 is 1.5v. Next, you need to figure out which VID pins to connect to obtain your vcore goal. The easiest to do is to connect VID4 to VID3 pins for a modified default of 1.7v. Since there is only two pins to connect if you choose this option, it might be easier to bend a tiny piece of wire into a "U" shape and drop it down into the appropriate socket pin holes. I initially tried this method, but my Abit TH7-II motherboard only increased the max setting from 1.625v to 1.85v, so I decided I wanted more. NOTE: Currently, wiring up just VID #3 and #4 seems to be the best plan, since too many Northwoods have died from overvolting recently.

Another option is to connect VID4, VID3, and VID2 pins together for a modified vcore default of 1.8v. This would probably be a good compromise and should provide more than enough voltage for most overclocking projects. For extreme overclocking freaks (like me) that want maximum voltage settings, the last option is to connect VID4, VID3, VID2, and VID1 pins together for a modified vcore default of 1.85v. This allowed me access to a maximum BIOS setting of 2.2v on my Abit TH7-II motherboard, which is more than enough voltage to get me into major trouble if I am not careful.

IMPORTANT NOTE: this paragraph was added later once the new C1 stepping was released. The new C1 stepping Northwoods have a default vcore of 1.525v. With the original B0 1.5v default Northies, VID0 was connected internally and VID1 was open (not connected), but with the C1 1.525v default Northies, VID0 is open and VID1 is internally connected. VID4 is internally connected on both B0 and C1 steppings. So basically, there are only two easy VID pinning options for the C1 stepping CPUs, which are as follows:

Connecting VID3 and VID4 to give you 1.725v
Connecting VID2, VID3 and VID4 to give you 1.825v



pin478r_.gif


For my project, I decided to throw caution to the wind and wire wrap VID pins #1, #2, #3, and #4 to give me the most vcore I could get. It was not entirely an easy task. I worked carefully and took my time knowing if I made a mistake that it would be very costly. My tools used were a magnifying glass, small tweezers, and a medium sized sewing needle. Having lots of patience and a steady hand would be an added bonus. Be advised that socket 478 CPU pins are tiny and extremely close together.

image003.jpg


I took a short length of fine wire strand that I stripped out of standard automotive grade speaker wire. You could also use a single strand out of a length of computer fan power wire or something similar. Next, I made a loop in the middle of this strand of wire by first wrapping it around a sewing needle. Then, I lassoed VID pin #1 with the loop, which was harder than it sounds, because you need three hands to hold the magnifying glass, the tweezers, and the sewing needle (which I used to guide the wire loop down over the pin).

After the loop was over VID1 pin, I carefully pulled both ends of the wire to tighten up the noose. Next, I criss-crossed the wire in a braid pattern around each of the other VID pins for maximum contact with the pins and also to add more friction to help keep it from sliding off the CPU VID pins. Once all of the desired VID pins were connected, I used the tweezers to twist the wires tightly together at the end in a pig-tail. Be careful not to bend the delicate CPU pins. The tail of the wire was trimmed with scissors and curled around the edge of the chip out of the way. This also helps prevent the wire from falling off the pins when the CPU is inserted back into the socket. Make sure the CPU is pushed solidly down into the socket all the way.

Conclusion

After everything was back together, I booted the system and went into the BIOS. Yippee, it worked! The default voltage registered 1.85v since I had wrapped all four VID pins. Better yet, I now had considerably more additional vcore settings than I previously had before. However, I must caution you good folks not to get carried away with running too much voltage. That's an invitation for disaster if you don't have exceptional cooling. If you're using air cooling, I'd think twice before going above 1.8v (real measured vcore) for any extended period of time. That's a 20% increase in voltage above the original 1.5v default.

Good luck and happy overclocking. This new found vcore voltage options should allow you to unlock the full potential of your Northwood. Prior to wire wrapping the VID pins on my Northwood 2.0a gig CPU, the highest I was able to overclock my system was 2.5 gig, even with lots of cooling mods. Now with higher voltage, I have benchmarked my computer up to a maximum of 2.66 gig. I hope your system will soar to new heights like mine did. Just be careful and watch those temps.

The top photo is how the author wrapped VID4 and VID3 pins in the original article for 1.7v modified default.

The bottom photo shows in red how I criss-crossed the wire in between VID pins for 1.85v modified default.
 
Last edited:
This post deserves sticky so we won't see so much posts about how to vidpin the Northy.

Bump for respect.

Excellent work batboy, your instructions blows away the referals you gave :D..

Yodums
 
Thanks Yodems, I spent a lot of work writing and editing this article. I've alreading sent a private message to the mods about making it a sticky. Mainly this article is just describing the method I used and is merely a guide to help others. Feel free to message me concerning any suggestions to improve the artcile or other comments. So far, I've had lots of people telling me they followed my method and it worked great. Only one person has told me it has not worked for them. I hope this helps those extreme overclockers who want more Northwood voltage and are brave enough to try wire wrapping. Please have good cooling first before attempting this mod. Good luck.
 
That was the best write up about vidpin wrapping I have seen yet. Excellent work! I have one question for you, however. In your post, you state that after you have wrapped the pins "the tail of the wire was trimmed with scissors and curled around the edge of the chip out of the way." Did you just wrap it around to the other side of the processor? Or did you tape it? By the edge of the chip, do you mean that side of the chip or the top of the chip? I have been doing a considerable amount of reading on this, and every article I have read kind of skims over this part of the process. Maybe I am just seeing too much into this, as it might be a trivial step.

I, and am sure everyone else, appreciates the amount of time you spent on an oustanding write up. Thanks!
 
That is a good question that I obviously didn't go into much detail to explain. I did curl the pig-tail end of the wire all the way around to the top of the CPU. However, you need to be careful here not to allow the wire to lay up on top of the heat spreader (the metal cap that the heatsink sits on). I trimmed the wire so that it just barely hooked over the top before it reached the metal IHS. If you accidently sat the heatsink on top of the wire, then the cooler would not sit flat and square onto the heat spreader and this would probably cook your processor and generally would be a bad thing. The heat spreader is a little smaller than the rest of the chip. Once you look at it closely, I think you'll see what I'm talking about. There is no reason you have to curl the wire ends up all the way to the top except to keep the wire from falling off when you inset the CPU into the socket. If you sat the case up right so the motherboard was on it's side, that would work too. One guy even carefully used a bit of glue to hold the wire down. I didn't do that because I didn't want the glue to cover any of the pins which could create more problems.
 
I just did my P4 the other day to get 1.85v. I did it exactly how batboy has it outlined in red for the second pic. I tried a few different types of wire but finally I got a strand of cat5e to work. Any other wire but copper didn't seem to make good enough contact to work. I just wish I could go beyond 1.85 with vid pinning.
 
Hey, Ben, What P4 do you have and how well does it overclock? I just picked up a 1.6A the other day, and even with the pin wrapping, it's unstable at 1900.
 
I have a P4 2000 Northwood with 256 MB mushkin DDR on an Asus P4S333-M. I was able to get 2400-2500 at 1.5v, 2700-2800 with 1.7v and with 1.85v 2800-2900. What kind of cooling and what kind of ram do you have for the P4? Were you sure to check the voltages to see if the vid pinning worked? It took me a few trys but I think that was because I wasn't using copper wire.
 
Last edited:
Yep. Checked everything. It won't even post on a cold boot at 125fsb, so heat isn't the problem. I've given up on it. My 1.2G T-bird outperforms it, even at stock speed. Well, not in benchmarks, but I notice a difference. It could be the motherboard, I have a shuttle AV40. I'm not going to worry about it. I'm just going to sell it. I have a watercooling setup coming that will work with my AMD stuff, and the T-bred will be out soon. Anyways, I've worked on a 2.2GHz northwood, and I can't tell the difference between it and a 1600XP.
 
What kind of resistance do you get when inserting the cpu in the socket after the mod. Normally the cpu gets 'sucked' in there with a sort of magnetic feeling with no resistance at all ..... so I was just wondering what one should expect to feel when instering the cpu in

Does it go all the way in with the wire mod
 
i just want to thank batboy,for his help and dedication to the overclocking community.with his excellent "guide",i'm now able to
take my baby up to and beyond what i had expected to get out
of my lil' 1.6a.again thnx and kuddos to the man.before 2.4 was all she wrote now she's singing at 2.527Ghz on all stock cooling
with some AS3 luvin.:)
 
Kaiser_Sose, with the wire wrap done, the CPU should slip right in. With mine after I did the wire wrap and inserted the CPU, it looked like the corner with the wired pins was slightly higher than the rest of the CPU, so I pressed it securely down a little more and it slipped in the rest of way in with just a little nudge. If you use a thin enough wire, there should be no problem seating the CPU fully down into the socket.

Redskull, you bring up a good point. It's critical NOT to bend any of those tiny delicate pins. Dropping a "U" shaped wire into the socket holes actually worked quite well for me when I was only connecting two pins (VID3 and VID4) for 1.7v modified default. But, make sure you connect the correct socket holes. If you do this method, there is more resistance when inserting the CPU into the socket. When you push down on the CPU it sort of "snaps" into place. Again, use a very thin wire that is not too long. However, I don't see an easy way to do this method if you plan on connecting more than 2 pins. Good luck all.
 
Last edited:
:) thankyou so much for the info it worked fantastic never thought i could reach this speed.
one question though why cant you just use a pencil to connect the pins like you do to unlock an amd chip?
 
Goinglobal said:
:) thankyou so much for the info it worked fantastic never thought i could reach this speed.
one question though why cant you just use a pencil to connect the pins like you do to unlock an amd chip?

Welcome to Overclockers.com

A pencil would be nearly impossible to get between the pins of a northwood. Even if you could connect the pins with a pencil it probably wouldn't work since the resistance of the led is to high. If you use a conductive pen the resistance is near zero. Wire wraping is the simplest method of getting extra voltage.
 
And think how difficult it would be to remove pencil marks or conductive ink. A nice thin wire is the best solution.

If the appropriate pins showed on the bottom of the motherboard, it might be possible to sodder on some switches and mod it so you could change the voltage by flipping a switch. Probably more effort than it's worth, tho.
 
Ok maybe some of you guys can help me figure out why my wire wrap wont seem to work. I did it once with some really thin copper wire from my headphones, and it wouldnt boot, so I tried it all over again and this time it will boot, but it just wont seem work. I can run windows everything fine, but the highest vcore is still 1.7 volta any help would be greatly appreciated. I also did it just like in batboys second pic with the red outlining. Thanks
 
Back