Unlocking Tbreds

[Mercutio]

I just got four tbred XP2100s from newegg.com. Serial numbers are contiguous (except for one) and they're all - and if I'm reading the little black sticker right "actually" 2800s.

Black sticker
ADXA2100DUT3C 9819807280064
AIUHB0308WPGW


The odd man out is marked as a 2600.

1.) SWEET!
2.) What the hell do I do with them? Does the pencil trick still work, or am I gonna have to run them out way out of frequency spec (170MHz or whatever) to get 2800-ish speed?

 

[blakerwry]

Can't you just bump up the multiplier in the BIOS (assuming your mobo supports it)? I thought t-breds are unlocked.

 

[Mercutio]

I haven't tried to overclock anything for a long time. I honestly don't know. One place says you can pencil trick L1s on tbreds. Another says you don't need to. A third says you need a special motherboard and a fourth says they're un-unlockable.

This is all on the first page of google results, mind you.

I figure Groltz or CougTek or Clocker probably knows more about the state of overclocking than I do.

 

[jtr1962]

While we're on the subject, does anybody know if it is possible to unlock the multipliers on either the Pentium II or the Tualatin Celeron? I've never seen any references to this, although there are plenty of sites showing how to unlock AMD processors. My M/B has 112 MHz and 133 MHz FSB settings with nothing in between. Both processors(PII-450 and 1.4GHz Tualatin) work fine at 112 FSB but not 133 FSB. It would be nice if I could play around with the multiplier as well. Maybe 14.5 or 15 would work with the Tualatin.

P.S. Sorry about the thread hijack(at least it's still on the subject of multipliers :mrgrn: )

 

[blakerwry]

As far as I know all T-bred rev. B are multiplier unlocked by default...

All Intel chips since maybe the pMMX have been multiplier locked... I have read some stuff about doing hard mods to the CPU's (such as cliping pins, soldering on resistors, or shorting pins) that will do various things to aid overclocking (like increasing voltage or increasing the default bus speed)... but nothing on multipliers...

 

[time]

AFAIK, you still need a motherboard that can "unlock" the CPU. nForce 2 boards as well as other boards from Epox (and Abit?) can do this. I optimistically tried a Gigabyte KT400, but it was no go.

 

[blakerwry]

i've read that the kt266, kt266a, kt333, and kt400 as well as nForce2 all support changing multipliers.. but it is of course upto the motherboard manufacturer to implement this feature in the board.

if you can't change the multiplier, you can just go right to changing the FSB...

going from 133 to 166FSB is about a 25% speed increase, if the motherboard was made to support 166mhz FSB then it should have the proper dividers to run PCI and AGP within spec.

luckily enough the 2100 at default multiplier and FSB is a 1.73gHz part.. upping the FSB 25% gives you 2.16gHz (approx AthlonXP 2600+ speeds)

So I see no reason why you cant just up the FSB to 166mhz if your mobo suports it.... if it does not run well at 166mHz FSB then you can try changing multipliers... if that doesn't work you'll have to settle on a more mild overclock of say ~145mHz FSB

 

[CougTek]

While we're on the subject, does anybody know if it is possible to unlock the multipliers on either the Pentium II or the Tualatin Celeron? I've never seen any references to this, although there are plenty of sites showing how to unlock AMD processors. My M/B has 112 MHz and 133 MHz FSB settings with nothing in between. Both processors(PII-450 and 1.4GHz Tualatin) work fine at 112 FSB but not 133 FSB. It would be nice if I could play around with the multiplier as well. Maybe 14.5 or 15 would work with the Tualatin.

AFAIK, both the PII and the Tualatin are un-unlockable. What's limiting you in the case of the Tualatin isn't the FSB, it's the CPU maximal frequency. Tualatin are good up to 1.5 to 1.6GHz. A 1.4GHz Tualatin with a 14x multiplier therefore won't be able to reach 133MHz FSB since it would make the CPU operate beyond 1.6GHz. If I were to buy a Tualatin (quite foolish these days considering their price compared to low-end Athlon XP), I would choose a 1.13GHz or 1.26MHz, not a 1.4GHz, because of the wider overclocking possiblities.

The easiest way to unlock a locked T'bred is this one (there are derivates, but they all end up doing the same thing).

 

[CougTek]

luckily enough the 2100 at default multiplier and FSB is a 1.73gHz part.. upping the FSB 25% gives you 2.16gHz (approx AthlonXP 2600+ speeds).

Close but no cigar. A 2100+, typically working on a 133MHz FSB frequency, will operate at 2.17GHz if FSB o/ced to 166MHz. 2.17GHz is the default frequency of the T'bred 2700+. The 2600+ works at 2.133GHz IIRC.

 

[blakerwry]

luckily enough the 2100 at default multiplier and FSB is a 1.73gHz part.. upping the FSB 25% gives you 2.16gHz (approx AthlonXP 2600+ speeds).

Close but no cigar. A 2100+, typically working on a 133MHz FSB frequency, will operate at 2.17GHz if FSB o/ced to 166MHz. 2.17GHz is the default frequency of the T'bred 2700+. The 2600+ works at 2.133GHz IIRC.

you're correct, the overclocked 2100+ would operate at default with a multiplier of 13.. resulting in 2158mHz when ran on 166mHz FSB.. somewhere inbetween 2600+ and 2700+ (but below the orginal speed of 2800+)

 

[CougTek]

you're correct, the overclocked 2100+ would operate at default with a multiplier of 13.. resulting in 2158mHz when ran on 166mHz FSB

No.

A 2100+ operates at 1.733.....GHz with a 133MHz FSB. At 166MHz FSB with its 13x multiplier, it will be clocked at 2.16666....GHz. 2158MHz is when you use simplify numbers. The real frequency of an o/ced 2100+ on a 166MHz FSB isn't 2158MHz, it's 2.167MHz (the 6 is theorically periodical, I put a 7 at the end to simplify).

So a 2100+ with a FSB of 166MHz won't be close to the 2700+'s frequency, it will be right on the 2700+'s frequency.

 

[blakerwry]

you're correct, the overclocked 2100+ would operate at default with a multiplier of 13.. resulting in 2158mHz when ran on 166mHz FSB

No.

A 2100+ operates at 1.733.....GHz with a 133MHz FSB. At 166MHz FSB with its 13x multiplier, it will be clocked at 2.16666....GHz. 2158MHz is when you use simplify numbers. The real frequency of an o/ced 2100+ on a 166MHz FSB isn't 2158MHz, it's 2.167MHz (the 6 is theorically periodical, I put a 7 at the end to simplify).

So a 2100+ with a FSB of 166MHz won't be close to the 2700+'s frequency, it will be right on the 2700+'s frequency.

I don't get you coug... 13 x 166,000,000 = 2,158,000,000 = 2158mHz = 2.158ghz which is closer to the 2700+ than the 2600+, but still inbetween thw two.

Are you saying that the FSB is not 166,000,000 (+/- a few percent)? is it really 166,666,666? I think that really just depends on the clock generator used... and individual differences between electronics.

 

[Mercutio]

Anyone here familiar with the literary works of O. Henry?

Not a single motherboard I have in my apartment is capable of attaining a 166MHz FSB. My GA-7AX(P)s are a PCB rev1, without the 166MHz selector. I've had the newer ones. I've sold them. The most current shipment I got (newegg) had two in it, but both of them are the old "up to 165MHz FSB" version that I already had.

Maybe there's a BIOS update...

Argggh.

 

[CougTek]

I don't get you coug... 13 x 166,000,000 = 2,158,000,000 = 2158mHz = 2.158ghz which is closer to the 2700+ than the 2600+, but still inbetween the two

Ok, I give up. Can someone with more patience than I have make him understand that both the 2100+ and 2700+ share the same 13x multiplier and therefore will both operate at the exact same frequency when set at a given FSB frequency?

 

[Mercutio]

13 x 133.33 = 1734 (same as XP2100)
13 x 166.66 = 2167 (same as XP2700)

I'm not seeing any room for debate.

I've got one of these chips up at 13x166, after getting a BIOS update for my #$%#$% motherboard. Hope it stays stable.

 

[blakerwry]

I don't get you coug... 13 x 166,000,000 = 2,158,000,000 = 2158mHz = 2.158ghz which is closer to the 2700+ than the 2600+, but still inbetween the two

Ok, I give up. Can someone with more patience than I have make him understand that both the 2100+ and 2700+ share the same 13x multiplier and therefore will both operate at the exact same frequency when set at a given FSB frequency?

doh... that's all you had to say... i hadn't done the math to figure out the multiplier on the 2700+

 

[CougTek]

doh... that's all you had to say

FINALLY!

haaaa....relief

 

[time]

Mercutio,

Do you mean that the first one you tried hit 2167MHz?

 

[Stereodude]

It all depends on the Mobo. The MSI KT400 board I have you can just change the multiplier even though the chip is not unlocked. You can do whatever you want without touching the chip. It's pretty nice. I got a T-bred B 1700+

Stereodude

 

[Stereodude]

Mercutio,

Do you mean that the first one you tried hit 2167MHz?

I pushed my 1700+ past 2gHz without any problems. I'm only running it at it's default speed at the moment.

Stereodude

 

[Mercutio]

It seems to. I bumped the voltage up two notches (to, uh.. 1.80V maybe?) on the GA7VRXP I installed it on. It's run a couple of DVD transcoding jobs, completed them, and is now sitting quietly waiting for me to finish my next batch of rips.
The only weird side effect I can find with the OC is that the SB128 onboard sound is... well, it sounds like someone has been playing with the environmental audio control, but of course I haven't. Everything sounds fine if I drop to 133 FSB.

 

[Mercutio]

It seems to. I bumped the voltage up two notches (to, uh.. 1.80V maybe?) on the GA7VRXP I installed it on. It's run a couple of DVD transcoding jobs, completed them, and is now sitting quietly waiting for me to finish my next batch of rips.
The only weird side effect I can find with the OC is that the SB128 onboard sound is... well, it sounds like someone has been playing with the environmental audio control, but of course I haven't. Everything sounds fine if I drop to 133 FSB.

 

[Mercutio]

Forum unresponsiveness/connectivity issues on my end led me to "submit" the preceding post five fimes. I deleted a couple so y'all don't questioning my ability to point and click.

I'm not really trying to rub it in.

Can someone who isn't a moderator tell me if this shows up as having been edited?

 

[blakerwry]

i dont see anywhere were it says things have been edited(other than your comments) :wink:

 

[CougTek]

The only weird side effect I can find with the OC is that the SB128 onboard sound is... well, it sounds like someone has been playing with the environmental audio control, but of course I haven't. Everything sounds fine if I drop to 133 FSB.

That's because your GigaByte doesn't have AGP/PCI bus lock when the FSB is bumped. Several nForce2 boards have this feature, which can be very useful when you have sensible peripherals hooked on the PCI bus.

 

[Clocker]

Use this Site to Identify your T-Bird

Use this Site to Identify your T-Bird

http://mp3.zonebg.com/cpu/cpu.php

Here is the one I just bought from Newegg:

 

[Clocker]

BETTER PIC:

 

[Newtun]

I can't see them dslreports pix. Do you have to be a member?

 

[mubs]

Methinks you're in good company, plump one. I can't see no pictures either.

 

[Handruin]

They don't allow hot linking. Copy and past the url into your browser to view the image.

http://www.dslreports.com/r0/download/318337~a375e6eda2efa24c8c3e564c420f0648/2100DSLR.png

 

[Handruin]

nevermind... I see it shows up... I swear it wasn't there the first time.

 

[mubs]

Doug: I still don't see 'em. I can't even see the URL when I mouse over the "pic"; there's just a placewholder there for me. Thanks for pasting the URL in your response, though.

 

[mubs]

Aha! I cut and paste the URl in my browser, saw the graphic on DSL reports, hit the back button on my browser, and can now see the same pic in the "Better Pic" post made by clocker. I'll bet if I clear my browser cache and come back to this thread the pic will be gone.

 

[Handruin]

Ah, good eye mubs. That must be why I now see his image in the forum. I thought it didn't display the first time. I right clicked on the red X and looked at the properties to get the URL.

 

[mubs]

Thanks for that right-click tip; I didn't know that. Learn something new everyday.

I cleared my browser cache and came back here; sure enough, the pic is gone.

 

[Howell]

Ah, good eye mubs. That must be why I now see his image in the forum. I thought it didn't display the first time. I right clicked on the red X and looked at the properties to get the URL.

heh. I quoted his post and then grabbed it out of the quote.

 

[Mercutio]

If you were using Mozilla, Netscape or even Mosaic, you could've right-clicked to "copy link location". Just one more way IE is keeping you down.

 

[time]

You left out Opera: "Copy image address" or "Copy link address".

 

[time]

I bought some Athlons Friday, including two XP1700+ at US$60 each. Here's my overclocking results with one in an Epox 8RDA+ (nForce 2), topped with a Falcon Rock II cooler.

Standard frequency = 1466MHz (PR1700)
At default voltage (1.63V reported) = 2061MHz (~PR2500)
At 1.70V (1.74V) = 2161MHz (PR2700)
At 1.75V (1.79V) = 2206MHz (~PR2750)

The system posted at 2300MHz, but remained unstable above 2200MHz. It was stable enough at the above marks to complete at least 3DMark 2001.

That's a 50% overclock. I'm haven't been this impressed since Celeron 300A (68%), and suddenly I have plans to upgrade all my PCs ...

 

[CougTek]

What temperatures report your monitoring software? Did you just bump the FSB or did you change the multiplier to reach those frequencies?