HOW-TO: Tbred Unlocking
 

This is a comprehensive guide to unlocking the 2200+ Thoroughbred processor, brought to you by XtremeSystems.org.

This procedure was developed out of a necessity encountered by OPPAINTER and I after trying to run our 2200+ T-Breds on the Abit KX7-333R. The problem exacerbated itself as a chronic NO post. I had updated my bios prior to installing the 2200+, to Abit's latest 9k bios...but to no avail. I would power on 15 times and get 1 post...it would show as a 1500+. I'd go into the bios, set it for optimum settings and reboot...and nothing. It would again take 15 power ups to get 1 post..never did I make it into Windows after 3-4 hours of trying everything.

If any of you have read the 2200+ review at HardOCP, Kyle had this same problem with the Abit AT7. He tried an updated bios supporting the Thoroughbred CPU..but to no avail. Apparently after contact with Abit's tech and or engineers..they offered to send him a version 1.2 of the AT7, which supercedes versions 1 and 1.1.

I am proud to announce that thru our good friend and fellow staff member Frank here at Xtreme, XtremeSystems.org has the fix for the no boot dilemma utilizing the 2200+T-bred, along with being able to unlock it and enabling all multipliers from 5x to 12.5x.

Here is what Frank AKA DaGoochmeister, a moderator from our very own XtremeSystems.org had to offer on the dilemma:

Hearing that these 2200+ T-Breds were not booting on the KX7-333R(Kyle at HardOCP had this exact problem with the AT7), I started searching the web for detailed pictures of a 2200+ and any other Thoroughbred processors to compare the bridge differences. What I found is that the L10 and L4 bridges that were on the Palomino, were missing on the Thoroughbred. In the Palomino, the L10 controlled 12.5X and below or 13.0X and above multiplier manipulation after the L1 bridges were reconnected. The only other difference was the L11 bridge lasering,and I found the difference for the specifications was that the 1700+ had a default Vcore of 1.5V and the the 2200+ had a default Vcore of 1.65V.One would assume that this bridge would not be important for multiplier control.

Concerning the L3 bridge, the only difference was that the 1700+ had no laser cuts and the 2200+ had a cut on the left, middle and the right which would be represented as - | - | - ,and the 1700+ as | | | | |.

Therefore, I concluded that the bridge controlling the multiplier was the L3 bridge whereas, in the case of the Palomino, it was a combination of the L3 bridge, L4 bridge for default multiplier and L10 for high or low. As many, many know, the bridges are bits of 0 or 1 where there is continuity or no continuity. As of now, I am not sure how the high or low is exactly determined for the TBred other than I have a hunch that a certain default multiplier determines access to low or high.

I also remembered that with the KR7A, originally the 2100+ had problems on the board probably with no definition within the bios of what a 13X multiplier is and/or what to do in this situation. Therefore, I presumed this was a repeat on the KX7333 with the 13.5X multiplier and relayed this information in a post last night in the admin/mod section after hearing that OPPAINTERS, and then today, MrIcee's, 2200+ were not booting on the KX7333.
Seeing that there is no high or low bit as in the Palomino L10 bridge, one would assume with proper multiplier manipulation within the bios and support , users will have access to all multipliers from 5X to 13.5 if the motherboard manufactures wish to do so at a later date.

Concerning the possible later bios support, full manipulation of all available multipliers should be available due to no L10 hardware high or low bit selector as in the 2100+ Palomino.

Here is a pic of the T-bred with the target L-3 bridges marked:


I have found thru my own testing and OPPAINTER will concur...that testing of the 5 L1 bridges with a multimeter conclusively shows that those bridges are indeed connected and don't need the traditional pencil or conductive ink/rear defogger kit treatment. They all show continuity when tested.

The next pic, and I apologize for the quality, shows the 3 target L3 bridges connected on my 2200+, I used the familiar superglue to fill the pits and conductive pen silver to connect the bridges


Once installed... the puppy booted up the first time as an 1100Mhz cpu...I then went into the bios and set all the normal parameters...rebooted and went right into Windows flawlessly. I checked all multipliers and indeed have everything from 5X to 12.5x...which for the multitudes of AMD O/Cers out there are what is necessary for the Xtreme FSB overclocking we do.

I'd like to give a big thanks to Frank for his diligent research on this issue...and Xtreme for publishing this guide. Good luck all and happy overclocking !!

Randi