Discovery on locked barton. How to Exploit ???

[YoupY]

Hey guys,

I've just got an locked barton and made a discovery on it. Something is different than the unlocked ones.

I'm not sure if the thingies on top of the proc are resistors or elco's but there is a difference in them. I always thought they were elco's but since I can measure them in Ohm I'm not sure anymore.

On a unlocked barton I measure 9.3 Ohms through those resistors. This value is the same for all the resistors except the one circeled in red. That one has no resistance at all.

Then I took a locked barton and measured this again. And I saw an maybe important differance. The value of all the resistors is 23.2 Ohm, except for the one in red. This one also has nog resistance, but the one in yellow is different. On the unlocked proc this one has a value of 9.3 Ohm. On the locked proc I can't measure a value anymore.



Please dont mind the pcb in the pic. This is still one from an old proc and still has a resistor / capacitator on the left side of the core, but imagine it isn't there

Now the big question is: is there any way to exploit this!!!

[Vlad Draculea]

i am too dense tonight but it seems i am just going to remain dense forever

[k1114]

My first reaction would be that we should pencil in the bridge until it gets down to 9.3 or lower

[Vlad Draculea]

i have slept a bit so now i can continue being dense as hydralgirium.

[YoupY]

Then the question is, which one should I pencil in? All of them?? I think not, because in comparison to the unlocked barton only one thing is realy different, (except for the other values) and that is the resistor? / capacitator? circled in yellow.

This morning I took a quick look at the traces on the proc en it seemed like they go from the one circeled in yellow to the L1 bridges. However to be sure I'll have to take a closer and longer lo look, since I was rushing to work this morning.

But if it is true, then the L1 bridges are not closed, since there isn't a resistance mearurable on the resistor circeled in yellow. A way to unlock the proc again could be as simple as pencilling that resistor to a value of 23 ohms (like the other ones on the proc) except if the values of the other resistors drop, due the pencilling

[Vlad Draculea]

dont ask me questions i told you all that i am too dense. AND A SPAMMER

[celemine1Gig]

You better try that on a locked Duron Applebred instead of a Barton, because if you screw up, your loss will be only one third in comparison to the barton.

[YoupY]

I just took a closer look at the traces that are on the proc near the resistor, but I'm not sure if they really go to the L1 bridges. Since I don't have a magnifiyng glass I'm not sure if the traces go there. It would be nice if someone could post a realy high res pic of a barton.

Before I will try to pencil the thingie I would be nice if I know what that thingie is meant to do. For now I still believe that the resistor circeled in yellow is a clue to unlocking the barton. Probally pencilling it will work, but before I'll give it a try, I will study the 104 page datasheet of the Barton, so that I know what I'm doing, so I wont kill more hardware this year. I have allready killed a lot this year

If anyone wants to go Xtreme and give it a try, I wish you luck, because pencilling may do the trick, since it's one of few differences that exist between a locked en unlocked proc, but don't hold me responsible if you blow up a proc.

EDIT::
Never mind the idea. I measured through the proc again, but I just saw completly different values than yesterday. Even on both the places I circeled there I a resistance now

So the thingies must be capacitators after all. I just held the multimeter on one of those for five minutes and I saw dropping the restistance very slowly. However I find it strange that I couldn't measure a value yesterday and I can now. What was happening , sorry guys

But there is still one difference. By measuring the proc I can conclude there are other capacitators on it, because I'm still reading different values on a locked and unlocked proc. Looking at the proc confirms this, since some of them have other colors.

[Vlad Draculea]

i have been measuring in an old duron 750 and i can tell you that i dont feel like telling you anything, and i dont like my paint made duron pics so you shouldnt.

[YoupY]

Today I still measure 20.7 ohm while I did 23.2 yesterday. When I had my multimeter on it for 5 minutes I saw drop the value from 20.7 to 20.5 ohm.

The resistance is the same on all the thingies.

I can be I measured some different values yesterday, because the proc came straight out of a running system, so maybe those inductors were still charged.

[xDUCK]

I would say that it is no coincidence that there are 12 sets of bridges and 12 of these resistors... I will try to get my hands on a locked chip so I can experiment myself. VERY GOOD DISCOVERY! Could be the breakthrough we need! :thumbsup:

[tennvols_69]

just to let you know .2 ohms isnt all that big of a deal general rule of thumd for electronics are 10% if the resistance is within 10% of original then its still good. im willing to bet these are caps and yuor reading a resistance somewhere else in the circuit. only true way to tell is to remove them from the circuit. however they might of changed the value of the cap causing the rc time consant to either lenthen or shorten to lock the processor. i might need to but me a locked chip to play with.

[tennvols_69]

well i used an dead tbred to verify that those lil brown thinggys are indeed capacitors. i measured it it measured 26 ohms then i removed it from the circuit and measured it it read open then started to charge then it discharged so it is a cap. it you measure it you will see resistance decrease then increase back to open. however it might bee something on that circuit that is open causing that so if we could figure out what it is then set it back to defult then we could probally unlock these things.

[Vlad Draculea]

i talked to the store, and they told me they are very busy and havent sent it, so i will have to wait at least some days to get the giant human killing robot.

[saaya]

very interesting! good work guys!

[tennvols_69]

im gonna look at amd's tech sheets and see if i can find this circuit and see whats all invloved in it. if it doesnt go to the die of the chip and is only a packageing lock you could build something to run away from the chip and set the correct values kinda like the GF device did with the slot A. ill have to researce this more but i think we found the circuit that is responsible for the lock just have to figure whats exactly differant so we can change it back to how its supposed to be.

[saaya]

i dont think you will find too much about it in the data sheets, i dont think they wrote the details about their multiplier lock there. would be way top easy for somebody who wants to remark and resell fake amd cpus to find it there...

[tennvols_69]

any one know where i might be able to get a dead locked thorton or barton so i can slowly sand down the layers to find where they might be locked at? ill be willing to put in the time if someone has a dead chip.

[saaya]

buy a dead ship on ebay... or ask newegg if they can send you a dead cpu

[Vlad Draculea]

the robot is made from metal, and kill humans very well, i am quite satisfied with it overall.

[YoupY]

Is there someone here that knows what the thingies circeled in yellow are?



On the locked barton the 2 circeled, are orientated differently. (the marking 102 is upside down) On them I also measure different resistance values, but since I don't know what they are I don't know if this could be meaning anything.

X-Ray pics should be really cool. That could make some things a lot more transparent

Edit:

I just checked with a m8 of me, and he has looked at 5 other procs, but it seems that they are placed random

[Vlad Draculea]

they are flux capacitors packs, wait for a thunderstorm and go to run with it to the higest place possible.

[Popcicle]

This is a very good Thread and deals with what
you guys are attempting to do.

http://forum.oc-forums.com/vb/showth...hreadid=245742

Check page 4..... even lifted the die-core off a
locked Barton trying to figure this locking mess out.

[tennvols_69]

VLAD i would love to have some xrays of the chip. i think it might provide us with alot of useful knowledge like where runs go. please post the xrays when you get them or send me a copy of it.

[xDUCK]

You know what we need? An AMD-employed deep throat. :smileysex