Once again about L1 briges

saltyfrog

Junior Member
Oct 16, 2000
7
0
0
Hi,
I was trying to oc my Duron 600 by cuting briges, as I have EpoX 8-KTA motherboard
that doesn't support changing multiplyer. I have all of L1 briges already cut and
I disconnected coresponding L3, L4 and L6 briges to be able to get an appropriate
arrangement for 900Mhz (I was using the same engraving tool as shown at Tom Pabst's
website) then I tryed to boot, it started at 600 Mhz, after that I connected appropriate
L3, L4, L6 briges and all of the L1 by drawing the connections with sharpened solder
wire. This time it wouldn't boot at all. So I came up with folowing questions after
all that.
Are the chances high that my CPU wouldn't work at 900Mhz at all even at 1.85V
and with a Chrome Orb cooler or I should check the connections once again??
And if I wouldn't be able to run it at 900, would it be bossible to go back to
600 Mhz just by disconnecting the L1 briges (are they needed to provide voltage
for all other briges?) or I'll have to be messing with L3, L4, L6 once again ??

By the way, I dout that I can connect all the briges back as they were, one of
the L4 looks pretty bad, thats why I'm asking about L1s.


 

paulip88

Senior member
Aug 15, 2000
908
0
0
Some Duron 600s don't make it up to 900. You should try 800, which is what most Duron 800s can get (though many get more). If the L4s look bad, just take comfort that you did this to the Duron 600 and not a Tbird 750.
 

johncar

Senior member
Jul 18, 2000
523
0
0
saltyfrog,
Yes, 900 may be too high, even with higher/1.85 Vcore.

But closed L1 bridges may be connecting interfering/corrupting signals from the mobo to the L3, L4 FSB Multiplier busses...try opening the L1 bridges...see http://members.nbci.com/candjac/index.htm Duron OC article, esp page 2 "Experiences", has item which showed that closed L1 bridges prevented an OC. Bottom line...if you don't have/connect to multiplier switches/jumpers, L1s are best left OPEN. Article explains all about oc'g with photos and circuit diagrams.

L1s "just" connect l3s and L4s to mobo, that's all...article explains all.
Good Luck, John C.
 

saltyfrog

Junior Member
Oct 16, 2000
7
0
0
thank you johncar for a realy good piece of
information. I was looking for something like this for quiet
a long time. Maybe I should have asked about this before I
started. Anyway I managed to go back to 600 without problem,
and now it is working at this speed for a while.
The only question which has been left unsolved is why did
it booted at 600Mhz when I had already cut the briges and
didn't connect anything yet. The L6 setting was for 1Ghz
and can only guess how it is interpreted when L3, L4 briges
are not connected to anything. OK, they are ignored if they
are all closed and and resistors on Vcc supplying briges are
shortcircuit, but what happens when all L3, L4 briges are open?
And I don't think the problem was that there were enough contact
after i have cut the briges so they were not actually disconnected.
 

johncar

Senior member
Jul 18, 2000
523
0
0
Salty Frog,
Go back to http://members.nbci.com/candjac/index.htm Duron OC article, esp "Experiences" page 2. You will see that FID and BP_FID settings should match. FID settings are made via L6 bridges, BP_FID settings are made vis L3 and L4 bridges.

And L3 and L4 bridges are NOT ignored, they actually connect directly to and signal the FSB Multiplier in the cpu. The L6 connect to Northbridge which we believe sets some system timings...so it should match the L3/L4 setting.

Try matching L3/L4 with L6, with L1s open to avoid any interference from mobo. And don't forget Vcore may need to be increased via the L7 bridges.

Hope this helps, John C.
 

saltyfrog

Junior Member
Oct 16, 2000
7
0
0
Sorry,
a bit of misunderstanding. I was talking about this Picture.
Now I see that the settings are not ignored, but the idea is that
it would be easier to deal with resistors wich are much bigger than
the breiges. And if all briges are closed the shorcircuit resistor
acts like a closed brige and a normal one like an open brige, that's
easy.

BTW, I can change voltage by DIP switches on mobo.
 

johncar

Senior member
Jul 18, 2000
523
0
0
saltyfrog, you wrote.......
And if all bridges are closed the shorcircuit resistor acts like a closed bridge and a normal one like an open bridge, that's easy.
End quote....

Careful...not quite....
If all bridges are closed, the shorcircuited resistor sets a logic level "opposite" to the normal resistor...shorting Gd/LO res sets a HI, shorting Vcc/HI res sets a LO. Then to get back to spec setting you must "open the bridge" connected to the shortcircuited res.

Because if you just open the resistor shortcircuit, but leave both bridges closed, the voltage to the signal bus will be exactly 1/2 of Vcc, which might be interpreted as either a HI or a LO..."you don't know". So to be sure to "reset to a spec setting", you must "open the bridge" connected to the shortcircuited resistor.
John C.
 

johncar

Senior member
Jul 18, 2000
523
0
0
saltyfrog, Boy did we blow it....correction coming...
We wrote>>>>
If all bridges are closed, the shorcircuited resistor sets a logic level "opposite" to the normal resistor...shorting Gd/LO res sets a HI, shorting Vcc/HI res sets a LO. End Quote.....

Should have written....
If all bridges are closed, the shorcircuited resistor sets a logic level "opposite" to the unshorted/normal resistor...shorting Gd/LO res sets the spec HI to LO, shorting Vcc/HI res sets the spec LO to HI. Then to get back to spec setting you must "open the bridge" connected to the shortcircuited res, the bridge that was "originally open".

Because if you just open the resistor shortcircuit, but leave both bridges closed, the voltage to the signal bus will be exactly 1/2 of Vcc, which might be interpreted as either a HI or a LO..."you don't know". So to be sure to "reset to a spec setting", you must"open the bridge" connected to the shortcircuited resistor.

Sorry for the error, did it without looking at diagram.
John C.
 

saltyfrog

Junior Member
Oct 16, 2000
7
0
0
OK, that's the same thing I was thinking about, maybe I expressed it incorrectly.
Anyway, I managed to change multiplyer to x6.5, the system is working stable at
702Mhz 1.5V and I think I wouldn't be able to get a higher multiplyer because the
last L4 brige (X4 HI) is destroyed, but I didn't try to make a connection, maybe
there would still be enough contact if I apply graphite or conductive glue there
considering that currents are very low. I'm using solder now and it is not possible
to make a good contact with it in this situation I think.