Problem OC'ing Athlon XP on Asus A7V266-E

[Guest]

All,
If anyone out there can help, please reply. I recently used the method mentioned in Tom's Hardware to overclock my Athlon XP 1900+. I did exactly what the instructions said and then installed the modified processor into my computer. When I fired it up, I was able to change the multiplier in my BIOS, but the processor stayed at 12x no matter what I set it to. I used WCPUID to determine the multiplier after boot up. I checked to make sure that the contacts did not have continuity with other contacts and that the silver paint did not make contact with anything else. I also checked the resistivity between the ground point and the contacts and I get a little less than 1k Ohms. I'm using what I think is good silver lacquer by a company called CircuitWorks. Please help....

BTW, I'm using an Asus A7V266-E motherboard with the 1004b BIOS and all jumpers setup for Palamino. In addition, there is a jumper called "THEMCPU". I have it set to 2-3 or "Reserved" according to the book. I read somewhere that the jumpers have to be set to Athlon/Duron instead of Palamino to get the multiplier to take... but I'm not sure if that is the same problem that I am having.

 

[Guest]

Have you tried resetting your CMOS after changing the multiplier?

 

[Guest]

Hmm... No I haven't tried resetting the CMOS. Somehow, I don't think it will work, but when I get some time, I'll try it and let you know.

 

[labdog]

you haven't correctly unlock your xp cpu, retry.


if you know you don't know, the way could be more easy.

 

[Guest]

labdog,
I have already retried once. I really don't think that I'm going to get better results this time around. Perhaps it is the conductive pen that I am using. Mayble it is generating too much resistance. Anyone think that might be the cause? I'm headed to the electronics store today to pick up another brand of the stuff and see if that helps. Another possibility is that I use heat curing (about 250 degrees F) after I paint the silver lacquer. What do you think about that?

 

[labdog]

try rather another good brand stick instead of maybe to burn your cpu.


if you know you don't know, the way could be more easy.

 

[Guest]

Ok... A third time is NOT a charm. I followed peoples advice and tried to overclock again. This time I was really really carefull to make good connections. I was able to completely reverse the process and start from stratch by using some alchohol and a little exacto-knife. I have the same problem I had before: no matter what I set the multiplier to in BIOS, the processor always runs at 12x. I also no longer think that the problem is the type of silver lacquer I'm using. I measured the resistance between the ground contact and the near side of the L1 contacts and I found that the resistance only went up negligibly when I added the silver lacquer and measured from the other side of the contacts. WTF! I think I'm going to give up now, but if anyone has some ideas let me know. By the way I tried to clear CMOS, but that did not do anything either. HELP! PLEASE!

 

[labdog]

1) are you sure the processor run at 12x. how did you know it ? (no offense)
2) is there a little resistance between the 2 contacts of the L1 bridges ?
4) can you test the locked L1 bridge with for eg an other resistance added to one L1 bridge contact & test resistance between the other L1 bridge contact & this added resistance ?
3) can you test the cpu unlock with a "more sure mechanism" like a little conductor, a little wire sprig put between the 2 contacts & fix a little with some adhesive ribbon or something else...


if you know you don't know, the way could be more easy.

 

[Guest]

Labdog,

Thanks for the post. Let me try to address your questions...

1) are you sure the processor run at 12x. how did you know it ? (no offense)

I am using WCPUID to determine the multiplier and frequency. I also ran SiSoft Sandra and compared it to the results at 12x to confirm WCPUID. All evidence indicates that the system is still running at 12x. By the way, I have found 2 other people on the web with this same problem and all of them have the A7V266-E motherboard.

2) is there a little resistance between the 2 contacts of the L1 bridges ?

Yes, but I think it is negligible. The resistance I measured was just under 1k Ohm before I did the overclocking process. (This was the measurement between the ground point and one side of the L1 contacts.) Unfortunately, I did not record them, but they were something like 989 Ohms and 990 Ohms, etc. When I added the silver lacquer, I ended up getting about the same reading (just under 1k Ohm). I don't remember what I got when I measured the across the L1 bridges themselves. Since I can no longer contact the L1 contact point (now covered in silver lacquer) it's hard to say if that measurement is useful. I think it was around 1-2 mOhms.

4) can you test the locked L1 bridge with for eg an other resistance added to one L1 bridge contact & test resistance between the other L1 bridge contact & this added resistance ?

I'm not sure what you mean here, but I think you want me to record the resistance before the unlock is done and then after so that the added resistance can be (semi) accurately assessed. The answer: Yes. (sigh) I'll probably do that today and get back to you.

3) can you test the cpu unlock with a "more sure mechanism" like a little conductor, a little wire sprig put between the 2 contacts & fix a little with some adhesive ribbon or something else...

I'm already using a Fluke multimeter with attached Gold miniprobes. With a steady hand, it is not difficult to make measurements at all.

 

[Guest]

Update: I just measured across the L1 connections and I get .2 or .3 Ohms on all of them. Then I measured from the ground contact to the opposite side of the L1 contacts and I got 1035 Ohms at the highest point. Most are around 990 Ohms. I'll redo the one that is at 1035 and then try the overclock again.

Another update:
After stripping the silver lacquer away, I found something interesting. When I measure the resistance from the ground point to the L1 resister bank, I get the following in order from left to right: 987, 1032, 990, 986, & 987 Ohms. When I measure the resistance from the ground point to the other L1 connections I get infinity! Is that supposed to be that way? According to the instructions at http://www.3dspotlight.com/articles/athlonxp_overclocking/xpoc-2.shtml I should be getting less than 1 Ohm. I don't think that I got super-glue on top of the contacts, but I went ahead and used a little very fine sand-paper to see if I could get better readings. After a nice layer was sanded off, I got the SAME readings. Maybe my processor just doesn't want to be overclocked. It is interesting that my BIOS will let me change the multiplier after I have performed this procedure, but not before. What is more interesting is that even though the BIOS is allowing me to change the multiplier, the system is still not overclocked (when checked via WCPUID). Crazy! Crazy I tell you! What do you make of this?

Yet Another Update: Ok... I think I found the problem. The superglue that I used is slightly conductive! I dabbed some glue on this paper and then placed two wires really close together (not touching) then took my multimeter to it. The resistivity was at about 100 mOhms! Ooops! Crap. Now I'm not sure what to do.<P ID="edit"><FONT SIZE=-1><EM>Edited by Ultivek on 01/06/02 06:52 PM.</EM></FONT></P>

 

[labdog]

your mustn't have a resistance aroud 1K ohm, it's too high.
normally with no contact between the 2 contacts, the resistance must be high (some K ohms).
if you close correctly the L1 bridge, you must have a maximum resistance around few ohms near the 0 ohm.

_______________________________________________________________________

I don't remember what I got when I measured the across the L1 bridges themselves. Since I can no longer contact the L1 contact point (now covered in silver lacquer) it's hard to say if that measurement is useful. I think it was around 1-2 mOhms.
_______________________________________________________________________

this is important. i think you have to retry with attention this manipulation to be sure of.

there is another think. What is the conductivity of this "silver lacquer" ?
is it sufficiant to make this kind of stuff ?


if you know you don't know, the way could be more easy.

 

[Guest]

All,
It's probably time to bury my Athlon XP 1900+ processor. Tom's process for overclocking the processor has done it in... If you have read through the posts, you probably know what is going on.. In summary, overclocking did not work. I could change the BIOS settings, but this did not change the actual multiplier. Even with the L1 contacts open, the resistance I measured between the ground point and the L1 contacts is about 1k Ohm or just under. Apparantly this is supposed to be negligible. I believe that my processor was not overclockable from the beginning. Anyway, I went ahead an removed the silver lacquer and reinstalled my processor. I found that I could still change the settings in BIOS. Strange I thought so I tested the superglue for conductivity (though marketed as "Nonconductive") and it passed. The crap is conductive. After removing the glue with a chemical remover, I had the same problem. I could still change the BIOS settings.. No big deal, you say? Well, you're right. But then I noticed that the computer will no longer keep CMOS settings when the computer is disconnected from A/C power. I thought I noticed this before. I am sure that removing the glue didn't do it; it was like that before. The system boots to the BIOS menu that deals with CPU speed. This is bad... really bad. I know that I took the risk here, but it would have been nice to use something that we know is not conductive here. I hope Tom includes more of a warning than he currently does for all those other unexpecting overclocking enthusiests out there. Anyone ran into this before? Right now I have in damage control mode. I'm hoping to get everything back to normal.

-Ultivek

<P ID="edit"><FONT SIZE=-1><EM>Edited by Ultivek on 01/08/02 00:04 AM.</EM></FONT></P>

 

[labdog]

from the <A HREF="http://www.OC-Shop.com" target="_new">http://www.OC-Shop.com</A> site

<blockquote><font size=1>En réponse à:</font><hr><p>PK-OCK7 (Golden Socket)
For everyone who has a socket A motherboard without the possibility to change the multiplier (for instance the Asus A7M266 and Tyan Thunder K7)

With the PK-OCK7 you can change the multiplier in a second.

price: € 67.00<p><hr></blockquote><p>the link is <A HREF="http://www.oc-shop.com/website/shopshow/index.php3?shop=oc-shop&productgroup=tweaking_gear&productsubgroup=-&nav=tweaking gear" target="_new">http://www.oc-shop.com/website/shopshow/index.php3?shop=oc-shop&productgroup=tweaking_gear&productsubgroup=-&nav=tweaking gear</A>

it's a little expensive but maybe it's can help you...


if you know you don't know, the way could be more easy.

 

[Guest]

i have exactly the same problem(a7v266-e) except when sometiems when i join the bridges and get a msg on startup that my xp 1600 is running at 9.5 clock mult instead of 10.5

i get have the same thing where after i clean it off i am still able to change to clock, but in doing os it has no effect

 

[OldBear]

I have the ASUS A7V266-E and an unlocked XP1800+. The answer is look on page
20 4), of the manual. Move the jumpers from Palomino 1-2 to Athlon/Duron 2-3.
You will then be able to raise your multiplier.

<font color=blue>Remember.... You get what you pay for. All advice here is free.</font color=blue>

 

[Guest]

All,
So far, I have found dozens of people who are experiencing the CMOS clear "feature" of the Asus A7V266-E board and others by Asus. Some customers tried to overclock it and used conductive glue on accident. Others didn't do anything (their boards came like that). It is my belief that attempting to overclock my computer caused damage to the motherboard; damage that I think can be caused by certain CPUs that are even clock-locked. I was able to borrow a friend's Duron and install it in my system. The problem is still there. I put my processor in his system and he did NOT see the same problem as me. This tells me that the problem is related to my motherboard. The fact that all the people I find with this problem have an A7V266-E or another Asus board supports this theory. To get more proof, I pulled out the old trusty Fluke multimeter and measured the voltage over the CMOS clear jumper. While plugged into A/C, the reading was about 3.3V. I disconnected A/C and continued to read the voltage. It stayed at 3.3V for about 15 seconds and then quickly dropped to about 0.1V. After the capacitance drained, the battery DID NOT TAKE OVER. WTF? If I measure the voltage of the battery I get 3.3V. I even replaced the battery with the same results. Somehow, the flea power line got messed up by the conductive glue in the laserlocks... I have heard those people that say, "So what, the glue in the laser locks will just prevent you from going over 14x" are full of crap. Tell them to try to disconnect flea power and see if their board can remember its CMOS settings. Anyway, I will keep you posted. Thanks for the tip about setting the mobo to Athlon/Duron, I don't think I tried that yet, but to be honest, I think my OC'ing days are past. (Too expensive)

"I have always been here."
--Kosh

 

[OldBear]

So far, I have found dozens of people who are experiencing the CMOS clear "feature" of the Asus A7V266-E board and others by Asus.

Every motherboard has a way to clear the cmos. The only problem is with the ASUS manual on page 24( 12) is
written and shown backwards. The board should not be run with a jumper installed on the "CLRTC". You only
install the jumper with the power off and the battery removed then remove the jumper and install the battery.
Is this the part you are having trouble with?



<font color=blue>Remember.... You get what you pay for. All advice here is free.</font color=blue>

 

[Guest]

LISTEN TO OLDBEAR!! He's right! I have the same problem, thought it was a bad job so redid the cosmetic surgery like millions of times and now all I did was change that jumper like you suggested and I'm up and running already! THank you so much oldbear for stating that, I think the other guy is having the same problem too, hope he reads this or yours.

 

[OldBear]

Thanks for the reply. Now it's your turn to help someone.

<font color=red>Remember...</font color=red><font color=blue>You get what you pay for. All advice here is free.</font color=blue>