XP unlock Problem Solved! *THX*

[Mangler]

I unlocked 3 Athlon XP's today and all work perfect except for one!

Here is the deal:

It is an AGOIA XP 1600+

It seems to be partly unlocked...

If I put the multi at 12.0 I get 11.0
If I put the multi at 11.5 I get 11.5
If I put the multi at 11.0 I get 11.0
If I put the multi at 10.5 I get 9.5
If I put the multi at 10.0 I get 9.0
If I put the multi at 9.5 I get 9.5
If I put the multi at 9.0 I get 9.0
If I put the multi at 8.0 I get 7.0

WTF is the deal guys? I am not that great with what bridges do what, so if anyone does please tell me which L1 bridge I screwed up.

 

[Mangler]

bump

 

[Jeff7181]

I think if you messed up an L1 bridge it would remain locked, or wouldn't boot.

 

[Mangler]

ya, I am not sure what is going on, it does boot up fine, but just not at all multi's

One bridge must be wrong!!! Any other ideas?

 

[mrman3k]

Wierd, maybe someone can comment on this. I am interested to hear the responses.

 

[Mangler]

bump....I am stuck HELP please

 

[yodayoda]

wipe away and reunlock. sometimes if you screw up one bridge, it makes erratic multiplier settings.

 

[Mangler]

I was hoping to see if someone knew what bridge it was...looks like I might be out of luck...I'll wait till morning to see if anyone does, and it not I will redo

 

[Frosty3799]

From what that other guy said, it isnt bridge one because then it wouldnt be working at all.

I wouild suggest just looking at it real close, w/ a magnifying glass. If one looks thinner re-do and re-try.

 

[Mangler]

Credit Given to JohnC over at Xtremesystems!

1X Bit Value L1, second from left was probably "directly connected to ground" during closing mod, which sets a LO, and "direct grounds" CANNOT be reset HI.

How we got there, easy because of your excellent data collection...and knowing AMD's Multiplier Code. First, 1600 chips have .5X, 1X, 2X, 4X Bit Value Multiplier signal circuits HI, 8X BV circuit LO. We can ignore the 8X BV as it's never used in analysis. So then....

12.0 gets 11.0 cause 1X BV should be HI, 11X = 1 less than 12X so 1X BV probably "grounded" at its L1 bridge.
11.5 gets 11.5 cause 1X BV = LO at 11.5X so GD = no problem.
11.0 gets 11.0 same as 11.5X above
10.5 gets 9.5 cause 1X BV should be HI etc...see 12X.
10.0 gets 9.0 same as 10.5X/12X above.
9.5 gets 9.5 cause1X BV = LO at 9.5X, so Gd = no problem.
9.0 gets 9.0 same as 9.5X above.
8.5 gets 7.5 cause 1X BV should be HI etc...see 12X.
8.0 gets 7.0 same as 8.5X above.

Clean up the second L1, esp the pit, be sure no conductive material is connected to either "dot". Then reinsulate, let dry well to prevent "weeping" of conductive material, finally connect the dots.

You should all learn the Multiplier Code, see how easy it is. http://www.beachlink.com/candjac/index.htm link to Code article, also Circuits, Tbred and Palomino articles.

Problem solved!

Thanks to everyone that replied!

 

[loafbred]

Thanks for the link - I just added it to my favorites. I have a similar problem with my unlock job on a 2100+, which left me with only 5, 7,9, and 11 for multipliers. I used whiteout to fill the pits, and after the second wipe-up for a bad trace, I didn't redo the pits, which probably got some conductor into a pit. The pits appeared to be full of whiteout, but conductive paint must have leaked or mixed in one of them. Right now it's not a problem since multiplier 11 is working well with 172 fsb, but I had intended to run 11.5*166.

 

[Mangler]

ya I am running the 1600+ right now at 166x11 for the time being.

I am busy till this weekend then I will redo my mistake, awesome link john had and awesome advice!