Bad unlock or what?

[SBeaver]

I just got my xp 1800+ today and I unlocked it with superglue and rear window defogger repair kit and everything seemed to go well.
I booted once and got into bios, i tried to change multipler and rebooted afterwards to se if the unlock was good, it probably wasn't because now I can't boot at all even if I reset bios.
What has happened? Is the chip dead or do I need to redo the unlock?
It's a agoia athlon xp 1800+ (it's green btw, if that helps):(

[Yodums]

It most likely is. You may shortened something now when booting up it is shortening something. The unlock is tricky, remove the heatsink clean the defogger pit off and try again. Many fail the first time.

[SBeaver]

But can you tell me for sure that the chip itself is intact?

[maxima88]

Try reseating the memory stick if you removd it to get to your cpu.
If that doesn't work, then you didn't unlock it properly.

Check all the bridge connections and make sure none are touching.

[fletch]

if you are using the XP-333 (i see in your signature...) then im willing to bet that the chip is dead. my xp-333 wouldnt boot at all when my chip was dead

[Yodums]

Originally posted by SBeaver
But can you tell me for sure that the chip itself is intact?

Nope. The defogger may be shortening two bridges or something. So unless you can clean off your chip and boot again, we can't conclude anything, sorry :/

[Polar2]

I had the same problem with my xp1600. I now use an ohm meter to check the connections before installing in mb. With the meter, you only need to fix the wrong connection.

[SBeaver]

I redid the paintjob and it booted once again, now with working multis.
But after a while of changing them around I found myself in the same situation as before, it wouldn't boot.
But the unlock did actually work for a while so maybe I didn't let it dry correctly?
On the box it says 5 hrs before use but thats when using it for rear windows defoggers so I didn't give a damn, maybe I should have?

[SBeaver]

I redid it a third time, this time I took the advise that i read on another thread that is supposed to fix unlocking for these green xp's, I sanded down the bridge a bit.
It worked just as before, then I let it rest a while and watched a movie (machine off) and now it wont boot again.
What is going on?
And also, when it does work, some multiplers are unavailabe!

[SBeaver]

I think I know the problem now.
One or more of the bridges are not connected propperly, therefore I can only use some multiplers.
When I select one that isn't unlocked, I get no boot and maybe because the CPU is reported as bad, the board does not reset CMOS properly until a "good" cpu is inserted.
This seems to be it but I still can't understand.
It's the same multiplers that cause trouble all these three times but I'm sure the bridges are connected!
Is there something new on these chips that noone has found before?

This might be isolated to the Iwill XP333 because it actually powers on when you reset CMOS, something my kt7a-raid did not do.
I bet that if I took you the CMOS battery it would reset and boot fine, wanna bet?:rolleyes:

[SBeaver]

I tried to test the bridges to see if there were any cross connections, I'm not sure I did it right but I put the multimeter on 200 ohms and checked between the bridges.
They were all connected, but what should the reading be?
I thought it was supposed to be zero but this they were all connected and still the chip boots.
I'm waiting for the CMOS to clear the slow way right now, I'll get back to this as soon as I have booted up again.
This is just too wierd

[SBeaver]

hmm hmm.... who's the man? who's the man?
The machine booted after the battery removal and is now running.
I still don't know whats wrong but I'm gonna close up the rig for the night and let it fold a bit.

edit: sorry about the spam, but here is an update:
I'm now running it very smoothly and I'm writing this from that machine.
It has for some reason set default multipler to 11x instead of 11.5 But this does not scary me.
It just gives me more room for fsb :)

edit 2: the multipler is still messy but at least I know how to fix that now.
Bad move by Iwill's engineers to make the CMOS jumper require a working system just to clear.

[SBeaver]

After a last touch up of the paintjob with, not paint, but a knife, the chip seems to work a bit better.
It now boots att all multipler (not tested but it seems that way) but I cant get multiplers 9.5x, 10x and 10.5x.
If I select them I get 9x.
I'll have to get better memory or OC this stick a little better Since I'll have to go for fsb only

[John C.]

Originally posted by SBeaver
It now boots att all multipler (not tested but it seems that way) but I cant get multiplers 9.5x, 10x and 10.5x.
If I select them I get 9x.

You are still probably having problems with the 2 leftmost L1 bridges. Here's why....

9X needs 4X + 2X Bit Values set HI.
9.5X needs 4X + 2X +.5X Bit Values set HI.
10X needs 4X + 2X +1X Bit Values set HI.
10.5X needs 4X + 2X + 1X + .5X Bit Values set HI.

So you can see that if the .5X and 1X Bit Value signals were disabled from being set HI, that would explain your data/expereiences. And the most probable cause would be "grounded" .5X and 1x signal busses...which most probably happened during the L1 closing operation. (See how easy it is when you give actual data and know the Multiplier code).

Here are 2 possibilities....
You did not insulate the pits good enough and the conductive fluid seeped in and connected the 2 bridges to the reported sub-surfac ground plane. This would reset them permanently LO. Solution would require cleaning all the way down into the ground plane. OR....

You grounded on the surface to some lettering or icons which are also reported to be at ground level, and maybe cross-connected the 2...that would ground both signal busses.

Remember...,5X L1 = leftmost, 1X L1 = next to the .5X L1. http://www.beachlink.com/candjac/index.htm Link to Palomino article for pics and diagrams explaining above.
John C.

[SBeaver]

Originally posted by John C.


You are still probably having problems with the 2 leftmost L1 bridges. Here's why....

9X needs 4X + 2X Bit Values set HI.
9.5X needs 4X + 2X +.5X Bit Values set HI.
10X needs 4X + 2X +1X Bit Values set HI.
10.5X needs 4X + 2X + 1X + .5X Bit Values set HI.

So you can see that if the .5X and 1X Bit Value signals were disabled from being set HI, that would explain your data/expereiences. And the most probable cause would be "grounded" .5X and 1x signal busses...which most probably happened during the L1 closing operation. (See how easy it is when you give actual data and know the Multiplier code).

Here are 2 possibilities....
You did not insulate the pits good enough and the conductive fluid seeped in and connected the 2 bridges to the reported sub-surfac ground plane. This would reset them permanently LO. Solution would require cleaning all the way down into the ground plane. OR....

You grounded on the surface to some lettering or icons which are also reported to be at ground level, and maybe cross-connected the 2...that would ground both signal busses.

Remember...,5X L1 = leftmost, 1X L1 = next to the .5X L1. http://www.beachlink.com/candjac/index.htm Link to Palomino article for pics and diagrams explaining above.
John C.

Thank you very much John, that's just the info I needed.
I just need to know one more thing:
I have found that the multiplers 8.5, 8 and 7.5 are all reset to 7.
This certainly seems like the same thing and I "think" the same is for 11.5.
Allthough I will never use those higher multis anyway, the lower ones like 8.5 and 8 might be requied if I some day got better memory for some reason and wanted to go for 233fsb.
Which bridge is that? I mean which has to be wrong for that to happen?

[John C.]

Originally posted by SBeaver
I just need to know one more thing:
I have found that the multiplers 8.5, 8 and 7.5 are all reset to 7.
This certainly seems like the same thing and I "think" the same is for 11.5.
Allthough I will never use those higher multis anyway, the lower ones like 8.5 and 8 might be requied if I some day got better memory for some reason and wanted to go for 233fsb.
Which bridge is that? I mean which has to be wrong for that to happen?

SBeaver,
Yes 8.5X, 8X, 7.5X all booting to 7X are 1.5X, 1X and .5X OFF from 7X right??? So it is the same problem, the .5X and 1X L1s, Just 2X lower that's all. Surprised you weren't sure of that by yourself...we're hoping to get people to start learning the Multiplier code and doing the fault decoding, then teaching others.

(Guessing that 11.5X boots at 11X, .5X OFF???)

Btw, default 11.5X has only the .5X Bit Value set HI, all others = LO. So the 1X "fault/problem" could also be a "still open" 1X L1, not just grounded. A still open L1 would prevent the default LO from being reset to HI by mobo/bios when needed. But repair recommendation stays the same...clean up and redo only .5X and 1X L1s. Hope you'll read the articles, learn the Multiplier code, it's not rocket science just some simple arithmetic with some Remapping thrown in.;-)
John C.