How do I unlock a thoroughbred b?

Getting a 2600 to work in an a7a266...

I want to try to unlock my thoroughbred b 2600+. I want access only to the higher multipliers. I believe that my bios on my a7a266 cannot read high multipliers, but I think maybe it can set them manually if I unlock the chip. I don't care about access to the low multipliers.

Anyone know how to do this?

Ok I used the wire trick to unlock the cpu. Now I can run it at any multiplier from 3.0x to 12.5x. Unfortunately, the chip is supposed to run at 16x133, so I need to find a way to access higher multipliers. Does anyone have any ideas how I might be able to do that? The motherboard has to support it, as there are several chips (for example the 1400 200 fsb version) that it does support that use multis higher than 12.5.

If your chip is unlocked, then look in this thread. It might be "for Mobile bartons", but it will work on any unlocked Tbred/Barton.

My chip is semi-locked. It's not a post week 39 factory superlocked chip, but I needed to use the wire trick (different one from the one you linked) to unlock it.

At that point, I was able to use multipliers from 5.0 to 12.5 (these are the only ones that are available through the dip switches). If I used jumperless mode, I was unable to change the multi, and I don't know why. Supposedly, I should have multiplier control through the bios, not just the switches, but I do not.

Would you suggest doing both wire mods simultaneously, or just this one?

If you can change the multipliers using dipswitches or BIOS settings, then your chip is FULLY unlocked, there is not such thing as semi-locked. its either unlocked or locked.
In your case its the motherboard that is limiting you, and the link I posted will get you to x13-x24 multiplier range from x5-x12.5

I understand what locked means. By semi locked, I meant locked but unlockable, as opposed to locked and not unlockable (as the new ones are). I did a wire mod to unlock it. Without any wire mod at all, the chip is locked. That's why I am asking if I need to use both wire mods at once (one to unlock, one to get me higher multis) or just the second one.

Also, how will I set the multipliers? Does it remap all the dip switch settings to higher multipliers? For example, will an 8.0 setting on the dip switch become a 16.0 setting?

all tbred bs are either fully unlocked or super-locked, there's nothing in between. what wiretrick (in what socket holes) were you using to try and "unlock' your CPU?

I was under the understanding that none of the thoroughbred b's were superlocked but some (or maybe all) needed to be unlocked. There are quite a few links on the web about unlocking locked thoroughbreds, like the wire trick I'm referring to. In the link below, it says it's for t-bred b's.

Thoroughbred b's are always unlocked with some chipsets and in some motherboards, but not in all chipsets and all motherboards. Some of us with older/crappier motherboards need to do a wire trick to unlock them. My a7n8x can unlock them all without additional modding (I think, haven't tried, but I've read that it's true. I use a barton in that mb.)

The chip is a 2600+ thoroughbred B (266 fsb version).
AXDA2600DKV3C
JIUHB
0337MPM

I am trying to just get it to run in spec in an Asus A7A266 rev 1.03 motherboard.

When I put the chip in initially, and used jumper free mode, it would boot into safe mode, but then no matter what I set the FSB to (the multiplier was locked in bios), the pc would not boot. Each time I changed it I had to clear the CMOS. The multiplier coming up in this case was incorrect (I think it came up 10.5, but not sure, it should be 16).

I tried changing the multiplier and FSB manually with the dip switches, but when I did this, the chip would never boot.

I did the wire trick I'll link below. After doing this, I was able to use any combination of fsb and multiplier (from 5.0 to 12.5) and the machine booted and ran just fine (albeit under spec, as it should be at 16.0x).
http://www.abxzone.com/forums/showthread.php?s=&threadid=31987&highlight=wire+trick

In jumperfree mode, the multiplier was still locked, but locked at different numbers, seemingly randomly. It was locked at various times at 8.0, 10.5, 5.5, 5.0, 12.5, and a few others. I was unable to figure out where the number was coming from in jumperless mode. In jumper mode, I had control over the multi.

The multiplier is actually correctly reported in the bios, as I benchmarked it at many different multis and got appropriate results on Sandra.

I still believe the chip was locked but was unlocked by the wire trick. I've seen a lot on the web about unlocking tbred b's.

This "unlocking" myth is an unfortunate result of review sites who didn't know their stuff playing around with the XP2100 and discovering that it wouldn't POST with any multiplier other than default. The problem being that the default was the lowest in the "high" range and the palomino core hit the wall around XP2100 rating, so setting it (using only 4 bits) to anything
other than the minimum wouldn't work. Then, they realised that if they pulled the 5th bit (though they didn't actually call it this) low, they could (unsurprisingly) set whatever multiplier they liked in the low range. Hence they concluded, incorrectly, that this was a magical "unlocking" bit and to "unlock" the chip you pulled it low. This is also the cause of the "NF2 unlocks CPU" thing.

You're in exactly the same position as the reviewers. If you try to set a multiplier of, say, 9.5, then your chip (because it read all 5 bits, but your board only sets 4 and the 5th bit is high) interpretes that as 18.0x and has a few problems trying to boot. By pulling the 5th bit low, you are making the chip change to the low region and are getting what you expect.


The problem with your BIOS is a little different. In jumpered mode, the board is directly fiddling with the BP_FID pins and the BIOS is using the "default" multiplier the chip sends out (which is changed using the BP_FID pins). The 5th multiplier bit is not changed by the 4 dipswitches, so you have access only to the high range of multipliers (though can use anything in that range).

In jumperfree mode, the BIOS uses a different (relatively complicated) technique, and gets upset when the multiplier doesn't stick. When the BIOS tells the CPU to use a multiplier of, say, 9.5x, the chip reports back that's it's booted on a multiplier of 18.0x (because the 5th multiplier bit, which the board doesn't support, is set hi), and the BIOS freaks out.

When you pull the 5th bit low, then the BIOS gets what it expects and is all happy. In jumpered mode, the BIOS doesn't ask for a specific multiplier, so it doesn't care what the chip reports back.

See
http://fab51.fc2web.com/pc/barton/athlon-e23.html
for lots of great info on multipliers and the like.

Now that I reread your posts. If you want to get your chip running at its stock speed. go into the BIOS, set the FSB to 133 and the multiplier to x8. Then reboot into windows, run GCPUID (My sig, red link) and tell me what you get? it should show 2133MHz.

If I do not do the wire mod, my system will not boot if I set the fsb to 133 and the multi to 8 with the dip switches. Without that wire mod, my system simply will not boot ever except into safe mode (5x66). Without the dip switches, in safe mode, it will not let me adjust the multiplier. If I do the wire mod and set it to 8, it comes out at a true 8x, not 16x as I would like.

To be more clear:

With wire mod:
Using dip switches, I get multipliers from 5.0 to 12.5
Using jumper free mode (which is supposed to be able to set the multi) I get it locked at some random, rather lowish, multiplier.

Without the wire mod:
Using dip switches, it will not boot (including at 8x).
Using jumper free mode, it boots into safe mode (5x66) to let me change the fsb in bios (multi is still locked to something lowish). Whatever I set it to, it will not boot after this and I have to clear the cmos or use the dip switches.

I have been unable to get past the bios to boot the OS without the wire mod.

I'm not too familiar with your board, but I had both A7V133-C and A7V266-E, they both acted pretty much exactly the same with a high multi Tbred. at 133MHz FSB I could only get a POST at x17 multiplier, at 100MHz FSB I could only get a POST at or above x15 multiplier.
Here's a table that shows how each low multi is interpreted on high multi range:
******************************
x5 = x13 (No POST on my boards)
x5.5 = x13.5 (No POST on my boards)
x6 = x14 (No POST on my boards)
x6.5 = x21
x7 = x15
x7.5 = x22
x8 = x16
x8.5 = x16.5
x9 = x17
x9.5 = x18
x10 = x23
x10.5 = x24
x11 = Reserved
x11.5 = x19
x12 = Reserved
x12.5 = x20
******************************

So basically you would suggest doing no wire or bridge mods at all, setting the board to use dip switches for fsb/multi, and just trying every multiplier until I get a POST?

Yep

Ok how's this for weird... I got it to boot at 133x16.5... a thoroughbred b should be able to handle this kind of modest oc'ing without problem. However, it shows the frequency in bios as 16.5x133=1800.

When I try to start windows to benchmark it in sandra, I get an error that ntoskrnl is not present and am told to reinstall it. Any clue what's going on?

My experiences seem similar to yours.

I can make the machine POST at multipliers of 16.5 and 15.0 only.

At 16.5, I get the ntoskrnl error if I try to run at a 133 bus.

The next option down in the bus list is 126. I can run 16.5 at that bus speed.

That leaves me two choices:
16.5x126=2079
15.0x133=1995

Of course in the second one, the rest of the system is slowed too.

What do you think is better, and does anyone have any idea how to make it run perhaps at 16.5x133? I can't imagine a tbred b not being able to handle it, but I don't know how to get around that ntoskrnl error.

Locking up at 16.5x126... I guess I'm stuck with 15.0x133 unless anyone has any ideas on the ntoskrnl.

I tried to overvolt it a bit, but even when I set the jumper to be able to change voltage in bios, it comes up locked. I would have to actually mod the bridges on the chip to overvolt it I think.

This motherboard is such a piece of crap.