1. Check out OCAU's review of the SpaceX Starlink satellite internet service!
    Dismiss Notice

Comments on the Palomino L10 Bridges

Discussion in 'AMD x86 CPUs and chipsets' started by BiteMe, Sep 9, 2002.

  1. BiteMe

    BiteMe Member

    Joined:
    Jun 28, 2001
    Messages:
    37
    Location:
    At Home
    I was looking for some comment from people who have opened the "closed" L10 bridge & closed the "opened" L10. I HAD a XP-1600 (may it RIP) & used a 12v battery to blow open the closed bridge - Bad Move in my opinion.

    Any one have any experience & comments & perhaps suggestions ?
     
  2. Vow

    Vow Member

    Joined:
    Aug 23, 2001
    Messages:
    629
    Location:
    Melbourne
    I've blown bridges on a duron using a 9V battery, but am not game enough to do it to my XP.

    Have you tried reversing any of the bridge modding you've done to try and get it working again?
     
  3. OP
    OP
    BiteMe

    BiteMe Member

    Joined:
    Jun 28, 2001
    Messages:
    37
    Location:
    At Home
    I ruined it big time. I tried to put it back but it wouldnt fire up. It still gets hot if I fire it up but its cactus.

    I have a KT7 (v1.0) with the Kt133 chipset. I cannot get the multiplier above 12 & I know the L1's are OK because I can change the multiplier from 5-12x.

    Im looking for the best approach to cut the L10 bridge - As I know by fusing the bridge Open is NOT the best approach.

    I dont know of any ways apart from cutting it with a stanley blade & if it shorts to the conductive ground can that break it permanently.

    There is a way by placing a 180ohm resistor across AJ27 to vcc(thanks to ppmz) & I have tried that a few times. I get the correct voltages to indicate its set to 13x & above but when I link the L1-link5 it refuses to boot - & I do have the multiplier set to 5x (5+8=13x100 = 1300mhz)

    ** Confusion Reigns **
     
  4. Mred32

    Mred32 Old and Decrepit

    Joined:
    Jul 4, 2001
    Messages:
    4,820
    Location:
    Perth, South of the River
    I'm considering doing this myself soon. I've seen a right-up where they used the 5v line from a PSU to blow the bridges.
    I might try that.

    Mred!
     
  5. OP
    OP
    BiteMe

    BiteMe Member

    Joined:
    Jun 28, 2001
    Messages:
    37
    Location:
    At Home
    ??? - OK - ???... I will describe what I did briefly.

    1- Had CPU working with L1 bridges (1-4 only) linked ie. Unlocked

    2- Removed CPU & placed in foil aid in anti-static

    3- Blew the L10 link-1 Bridge (the link that was closed) with 12v supply ie. a real puff of smoke & link was gone

    4- Closed the L10 link-2 Bridge (the link that was open) with wax procedure & conductive silver compound -the same used for the L1 bridges as per "Bravo's" thread on how to unlock XP's.

    5- Removed foil

    6- placed CPU back in & heatsink etc...

    7- turned PC on = DEAD ie. nothing, Kaput, not working, no longer alive, RIP CPU, etc...

    8- removed CPU did the foil thing etc... cleaned up L10 Link-2 & tried to resurrect L10 Link-1 with Solver compound

    9- Tried CPU again Pls see point 7 (above)...

    10- Conclusion = very unlucky person OR perhaps a flaw in procedure

    11- Recommendation = Post a thread here & get comments BEFORE embarking on another victim/cpu !

    :confused:
     
  6. l00b3r

    l00b3r (Banned or Deleted)

    Joined:
    Jun 23, 2002
    Messages:
    8,352
    Location:
    Somewhere Urban..... Bris
    and the purpose of this CPU distruction was, (THE REASON BEHIND IT, NOT KILLING IT)
    :rolleyes:

    -Matt
     
  7. Vow

    Vow Member

    Joined:
    Aug 23, 2001
    Messages:
    629
    Location:
    Melbourne
    I think using a huge battery or PSU to blow the bridges is a bad idea. When a 9 volt battery (which is, compared to other batteries, as weak as piss) has enough power to blow a bridge, there's no need for anything more.

    Maybe the L10 bridges aren't as clear cut as the others, and keeping to the 180 ohm resistor method might be the way to go.
     
  8. OP
    OP
    BiteMe

    BiteMe Member

    Joined:
    Jun 28, 2001
    Messages:
    37
    Location:
    At Home
    L00b3r - The reason behind the mod was an attempt to get the multiplier above 12 (or 12.5 ?) in the bios. If the L10 bridges are swapped this enables an internal addition of 8x to be added to whatever the BIOS is set to - so for example by setting the BIOS to say 6.5x & the L10 Bridge swapped you get a 14.5x

    I am using the CPU in a KT7 with the KT133 chipset which of course does 100mhz FSB only.

    Im just trying to keep up :confused:
     
  9. OP
    OP
    BiteMe

    BiteMe Member

    Joined:
    Jun 28, 2001
    Messages:
    37
    Location:
    At Home
    VoW - Yep I agree the 180ohm from AJ27 to VCC is an excellent idea. However in my situation & therefore probably others that fix doesn't work even though the voltages at AJ27 (ie 1.57v) indicate a 13x multiplier is active. When I join the L1 link-5 the PC wont boot (of course the bios is set to 5 or 6 or 7x) so I remove the L1 Link-5 & its back to square one - ie able to set the multiplier from 5-12x in the BIOS.
     
  10. Vow

    Vow Member

    Joined:
    Aug 23, 2001
    Messages:
    629
    Location:
    Melbourne
    Bugger. :( What motherboard is that on?

    Another reason I don't like messing with bridges on the XPs is because they look different... on a Duron you can see the bridges on top of the ceramic. On XPs, they're kinda under the surface. I know this might not mean much, but it's just another reason I'm not game enough until the process has been proven more.
     
  11. ppmz

    ppmz Member

    Joined:
    Aug 19, 2002
    Messages:
    71
    BiteMe: Setting 5x (13x100=1300) doesn't work for me either. I have no idea why, but the lowest that my CPU will do is 7x (15x100=1500MHz).

    Maybe you've got the same problem. Try 15, I'm pretty sure most 1600+'s will handle 1500 MHz okay.

    -- P>
     
  12. Mred32

    Mred32 Old and Decrepit

    Joined:
    Jul 4, 2001
    Messages:
    4,820
    Location:
    Perth, South of the River
    Any idea how this will go in a 133a mobo running at 133fsb instead of 100.
    I would like to push this cpu a little more with the multiplier.
    This mobo does not support sub-default multiplier so no dropping the mult and upping the fsb for me:(
    If I can get higher than 12.5 it might be worth the effort.

    Mred!
     
  13. OP
    OP
    BiteMe

    BiteMe Member

    Joined:
    Jun 28, 2001
    Messages:
    37
    Location:
    At Home
    Thanks ppmz - Last night after having no luck (& maybe thinking that CPU may be dodgy) I tried it on another XP1600 (agoia wk8) - & the same problem.

    Basically I conclude its the KT7-Raid v1.0 board thats the problem.

    I've tried the AJ27 + 180ohm resistor via a switch to VCC & get all the right voltages on AJ27 - with link5 on L1 connected & L10 as per factory. I removed the AJ27 (by o/c the switch to VCC) & cut the L10 link-1 & connecting link-2 with the Bravo treatment.

    At best when it boots I get 5-12x multiplier & NO sign of it doing 13x or above. I took your advice very early & set the BIOS to 7x but it boots up at 700mhz with AJ27 reading 1.775volts.

    FYI when I swap over the L10 bridge AJ27 goes high at 1.775volts so it does the same thing as to tie it high - so something at least "seems" OK.

    Also by cutting the L10 link-1 it doesnt blow the CPU up- which is something good & if/when it doesnt work just wax it up & conductive varnish the link back.

    What I need to try is a ver 1.3 KT7 with Raid (KT133a). I never trusted ABit since I bought that v1.0 board. I know I might get flamed for saying that but how many people can remember ABit pumping out BIOS updates every week trying to fix problems from there previous BIOS release - I must have done at least 20 BIOS updates from ABit.




    :mad:
     
  14. ppmz

    ppmz Member

    Joined:
    Aug 19, 2002
    Messages:
    71
    Having 1.775V at AJ27 doesn't really tell you it's going into the 8X BUS. If L1[5] is not linked properly (or grounded), then the 1.775V is not getting there.

    BTW I'm using the A9 bios, but I doubt that matters much.

    *** Edit:

    >> I took your advice very early & set the BIOS to 7x but it boots up at 700mhz with AJ27 reading 1.775volts.

    Just thought about this more carefully. Having 1.775V at AJ27 would tell you that L1-bridge-5 is not connected, OR, both L10 are cut. Since there's a 180ohm resistor, you should get:
    V = 1.775 * (1000 / (1000 + 180))
    V = 1.50
    (the 1000 ohms is on the chip).

    You should probably touch up the 5th bridge.

    [​IMG]

    -- P>
     
    Last edited: Sep 12, 2002
  15. OP
    OP
    BiteMe

    BiteMe Member

    Joined:
    Jun 28, 2001
    Messages:
    37
    Location:
    At Home
    I've got A9 Beta BIOS. I will flash the beastie & give it a try. At least Im not breaking CPU's anymore.

    I will get back after a hack.
     
  16. OP
    OP
    BiteMe

    BiteMe Member

    Joined:
    Jun 28, 2001
    Messages:
    37
    Location:
    At Home
    Talk about being teased.

    I actually got the beast to boot to 1500mhz ONCE yesterday- very early. I reset the PC to get back into the BIOS & it went back to 700mhz - Damit.

    I then spent the next few HOURS trying again on both CPU's but no luck. I started back from scratch on several occasions & I noted that both CPU's read about 220ohms resistance across each link in L1 & on L10 I think link-1 was 1000 ohms & link-2 was 220 ohms.

    I dont think the 220ohms helps me much. I tied VCC directly to AJ27 with L1 link-5 closed & it drew over 200ma & still didnt do the trick (I think I was lucky it didnt do damage soemwhere).

    The A9 BIOS update didnt help.

    Both CPU's are unlocked OK. I checked that each could be changed by the BIOS from 5x upto 12abouve (12.5x) - All OK there.

    ppmz - Thanks for the cct diagram - A picture is worth those thousand words.

    I'll be back...:confused:
     

Share This Page

Advertisement: