2600 :eek: now that's what i wanted to see! :toast:
Printable View
2600 :eek: now that's what i wanted to see! :toast:
I'm gonna put my hands on a 1.4 MIXIB 0336 next week I think :). Btw, top resault now with air is 211x9.5@1.6v Which is not that great but I think it'll go higher once I fix my WC rig...
Here is mine, just got it today, still haven't played much but does 3dMark at this speed, i seem to be getting prime errors but think its due to the ram having serious probs.
http://www.rudzer.com/pics/duron1400.jpg
obs: don't ask about the stepping since i was more worried in try it that look at it..will write it on the next restart when i move this great cpu to other pc on home-lan :)
§§§§ now this is REAL VALUE! :P
I got one too, I filled the tiny L2 pit with silverglue and this is the result:
http://home.planet.nl/~dikke332/Duron2511.jpg
It's cooled by a Maze4 and a BlackIce Extreme with 2 fans @ fullspeed... stressed about 50C but hey... it's stable :D
will mod mine next week, hopefully it will work
btw Garrett what voltage you use and stepping you got?
Well, well, I think I got a dud! It's a MIXIB 0338 XPMW and it gets extremely hot. My Barton 2500+ has about 45°C under full load @2,2GHz, cooled by a Swiftech MCX462+ with a Delta EHE @5V.
When I tried to run the Duron at 2,2GHz it got hotter than 60°C and was everything else than stable! It has more than 40°C even when stressed at it's default 1,4GHz.
I rechecked if the heatsink was seated properly, as I couldn't believe that it got soo hot, but the Swiftech made perfect contact.:(
I think I'll sell this one and get myself a new one to play around with.
http://www.rudzer.com/pics/duron.stepping.jpg
this is the stepping of mine, but it seems there are others much better , still very happy with it :D
Well I got another one, 1.6GHz "old style"..... but don't have a conductive ink and my tryouts with the pencil are mover 10 now.... anybody that made a successfull L2 cache mod with the pencil??
Are the multipliers locked on these chips? I was thinking of getting one as the last 1700+ I got was a super dud. Only 1.9 GHz max. Stupid A cores. And I would much prefer a wire trick for the extra cache as apposed to cutting and painting in bridges.
MIXIB...the best stepping there is for Duron....Quote:
Originally posted by Rudzer
http://www.rudzer.com/pics/duron.stepping.jpg
this is the stepping of mine, but it seems there are others much better , still very happy with it :D
got myself a MIXIB...2650Mhz pifast on plain water, 2620Mhz prime on 2.05V
You can't say that generally all MIXIB are good, because that is just pllain wrong. I got a MIXIB 0338 XPMW here myself and it's horrible!!! It can't even do 2 GHz stable! :stick:Quote:
Originally posted by NoStra
MIXIB...the best stepping there is for Duron....
got myself a MIXIB...2650Mhz pifast on plain water, 2620Mhz prime on 2.05V
That's definately NOT good. There may be some MIXIB that do very well, but definately not all and that's the point. Overclocking depends on your luck and not only on the CPU stepping that you get.;)
So has anyone successfuly done the L2 mod on the new packages. I just got mine in today and would like to try before I drop in the board
I have ;)Quote:
Originally posted by SPQQKY
So has anyone successfuly done the L2 mod on the new packages.
I did ;)...just fill the right L2 lasercut with some silverglue....Quote:
Originally posted by SPQQKY
So has anyone successfuly done the L2 mod on the new packages. I just got mine in today and would like to try before I drop in the board
any1 know where to get this conductive paint
These chips are doing better than the 1700/1800 dlt3c's. I'm not satisfied with my barton. It takes 1.9v to get to 2.2ghz. Maybe I'll try one of these durons.
My Duron at 2465 (top benchable speed) vs my barton at 2.2 the barton seems stronger in 3dmark, but i'll have to do some benches with the vid card at deafults to make sure.
I'll be ordering a 1.6ghz sometime this week hopefully i'll have some luck with it.
Get some of the rear-window defogger repairing stuff. Works pretty good :)Quote:
Originally posted by Daishi
any1 know where to get this conductive paint
Quote:
any1 know where to get this conductive paint
Got mine here..
http://www.web-tronics.com/2200-mtp.html
And all of you looking for those Volt mod Pots
They are 1/4 of the radio shack price.
http://store.yahoo.com/webtronics/pa...tiometers.html
There's a minimum $15 order, so you'll have to buys something like some pots or whatever
hehe Bag of 100 LED's $1.50
http://store.yahoo.com/webtronics/bagolspec100.html
Main page im sure you'll find some goodies
http://store.yahoo.com/webtronics/index.html
Shipppings actual UPS rates to your address, So you might wanna get a few things to cut that down if your only thinking about spending $15
Directron has them too.
http://www.directron.com/conductivepen.html
but you'll end up paying more for the pen from them with shipping, unless your ordering other stuff from them.
Rear window defogger repair is the way to go. It's very conductive and fast drying. You can even thin it out with nail polish remover so that it leaves only a small amount in the pit.
is there no way to mod the pins to enable the extra cache?
Yeah, I would think someone would have discovered it by now. I wish I had the knowhow. One more question on the L2 cache unlock, on the new wafers, do you just fill the pit or do you fill the pit and paint it to the bridge. Are the new chips unlocked?
all new chips are locked afaik, even the durons.
i closed the gap between the cut L2 bridge, then i scratched a bit of the surface to reach the 2 small spots of each side of that cut bridge and connected them with silver paint... still only shows 64kb L2 cache in all proggies and the bios :(
i put on my wizard hat and robe
http://fab51.fc2web.com/pc/palomino/athlon-e5.html
so is there a way to enable the cache with a pinmod?
That does not seem to be the case. I have a Duron 1.6 that I just dropped in an Epox ep-8k9ai and I set the cpu peramiters to 166x10 and she booted just fine at those settings. This is the new green wafer (I think it was MIUGA 0337). Haven't done the cache mod or any serious OCing yet.Quote:
all new chips are locked afaik, even the durons.
i hope that my multipliers aren't locked when i get my duron 1.4 on friday.
Hey saaya I'll look into a pin mod tonight I should have some time to play with that.Quote:
Originally posted by saaya
http://fab51.fc2web.com/pc/palomino/athlon-e5.html
so is there a way to enable the cache with a pinmod?
My chip is one of the newer ones from newegg and is definately locked. 1.6 MIXIB 0342WPMW
cool docah :D
i talked to ita, the owner of fab51 and he says he doesnt think its possible... :/
i dont know crap about the pullup resistors they are talking of.
maybe its possible to solder a tiny wire from the leg of the little black chip that leads to the cut L2 bridge to VCC_vcore and put a pullup resistor in between? on his site he did exactly the same thing with a palomino, he used a Pullup to Vcc_CORE with 300 ohms and it made the entire 256kb cache work on a mobile duron wich had only 64kb cache before!
afaik there is no diference between the palomino and tbred L2 and L9 bridges, so it should work.
if it works then we can try to bypass the internal multiplier control circuit on the new locked chips and unlock them! :D
http://fab51.fc2web.com/pc/barton/athlon-e20.html
I don't think that would be possible becuse it would do the same as the bridges wich is connected to the core but has no longer any function. But it might be worth i try, i would like to see how it goes.Quote:
Originally posted by saaya
if it works then we can try to bypass the internal multiplier control circuit on the new locked chips and unlock them! :D
http://fab51.fc2web.com/pc/barton/athlon-e20.html
The cache one i think would work. The pullup is just a connection to vcc_core with an 300ohm resistor, but pulldown is internal.
and i conjure the :banana::banana::banana::banana: of the beyondness
Vlad Draculea
As i said they are connected (It's L3 btw) to the core (pins too ofcourse) but they don't have any function any more. Hoots theory on that was that they gave the the circuit inputs a lethal shot of current or some microcode in the core that is programmed to ignore the multiplier controll inputs.
Yeah, i couldn't find the pins. This locking kinda sucks.... :-/
i think they probably blew some lines of the copntrol circuits inside the pcb by bumping a high voltage through them... :(
but if we can rebuilt the circuit external, wich doesnt look too hard since we only need to connect some pins of the black chips on top of the cpus to vcc_vcore with a small resistor in line it should work just fine...
and if it works for the cache there is no reason why it shouldnt work for the multis.
vlad, do you have a link to that thread on ocers.com? what exactly did they try?
here it is.
httpnorl
and you explode in a fine mist of blood
well, as i said before, i dont think my chips are real tbred a chips, they reach 2.3ghz and probably more with better air cooling!
and i really think its the package... if your frind could take a high res pic of the chip that would be great!
yes the chips are the same i mentioned in the other thread. the 1600 is unlocked! the 1400s are locked. they are both in the new packege, so amd probably started locking them a bit later.... maybe the locking mechanism didnt work from the start, or didnt work on all chips, so they needed a second revision of the new pcb design?
the unlocked 1600 is older than the 1400s, the 1600 was from week 37 and the 1400s from week 46. i dont know if the little lettwer and number on the edge of the cpu are the revision numbers of the pcb of the cpu, but it could be possible.
on the 1600 it says a1 and the 1400s show e5 ...
EDIT: my 1700+ week 08 shows B4...
i tried everything i could think of and nothing worked :/ if you cant get a high res xray pic let me know and i will cut my way through every layer of this crappy duron and check out the circuits ;)
i connected vc_vcore to the leg of the black chip that leads to the cut L2 bridge, i cut the trace between the black chip and the cut bridge and connected vcc_vcore to both sides, side of the black chip no difference, side of the cut bridge, no boot...
then i scrathed off the upper layer above the L9 side of the L2 bridges and connected the spot of the secons L2 bridge wich was not cut to the point of the cut L2 bridge on the L9 side... no diference...
i also used a 300ohm resistor, but that didnt change anything... i think the bridges are not working at all... amd just leaves them on top and cuts them to fool us ;)
oh, i tried the 166fsb mod as well, didnt work either...
because i am a level 1000000 magician, and you are just a level 5 druid.
Seems like this might be permanent... but i had that feeling before.
nah i dont think so, theres nothing they can add we cant remove and nothing they can remove we cant replace ;) it will only get more difficult to unlock chips, thats all :)
vlad, i dont think a usual x-ray pic will show anything, it wont let you see the traces since they are all on diferent layers, and since they are above each other you wont be able to see whats under a trace of the first layer...
are you sure somebody already sanded his chip? can you send me a link? i have a locked and an unlocked duron with the same new package! so there should be a diference between those packages if its not inside the core!
so the first layer of the package is what? and the second layer etc? how do i dissolve the layers without sanding them? just let me know and i will throw my duron into any acid etc you tell me :D
im really really busy atm with getting all the cards ready to ship them and with a bunch of exams in school, so i have little to no time to do any research, so just tell me whatto do and ill do it :)
is there no way to force the mainboard to work with a special multiplier? like forcing all the multiplier pins on the socket to 1 or 0 so they have to accept the multi? or will the cpu simply refuse to work then?
no pics for ya, sorry level 5 masters of the light made me do it.
do you have a high res pic or can give me a link to it?
I've just gotten this little baby for an overclocking comp:
Someone at OCAU managed to get a chip of the same week and stepping to 2.2GHz. Perhaps I should scramble for a week 35?Quote:
DHD1400DLV1C Z331557271849
MIXIB 0337TMPW
i just got yet another duron :D
dhd1600dlv1c
2366435202063
mixib0345upmw
i think it runs stable at 2ghz with 1.5v :D
here's mine so far (not modded w/ L2 yet)
http://www.xtremesystems.org/forums/...threadid=24334
too busy to try more volts atm, looks very good so far :)
i couldnt decide what test to run so i installed prime again and ran both tests :D
didt some people manage to enable the cache on their multiplier locked durons?
EDIT: 2ghz is stable as well, at 2.05 it crashed after some time, but i think its not the cpu maxing out. kt266a@170fsb =43mhz pci and 86mhz agp speed...
Looks like mine is a dud, most I can get out of her is 1.9GHz with 1.85v and at only 42c under load. Bummer.
i have an duron 1.6
DHD1600DLV1C TS8808S240124
MIUHB 0529UPMW
now it's @ 2.2 GHZ on 1.85 v and going further
edit: 210*11 @1.9v
My duron was doing 2650Mhz pifast....but now it only does 2560Mhz....
Probably my cpu is going to die...
DHD1400DLV1C Z334997240648 MIXIB 0338VPMW
:stick: I did the cache mod with a pencil.. likt this :)
http://home.c2i.net/madOC/img/pics_m...dduron1400.jpg
This one at 1.6vcore in bios.
http://home.c2i.net/madOC/img/pics_oc/1.6vcore.JPG
And maximum oc at 2.125vcore in bios.
http://home.c2i.net/madOC/img/pics_o...-max-256kb.JPG
And i got 1000 points more in 3dmark01SE after cachemod :banana:
What a drag, my chip is unlocked which was lucky, but what a turd of an OCer. I was hoping it was the cheap 8k9ai that was holding it back, but after dropping it in my 8rda+ and a Shuttle an35n ultra, it still tops out at 1.93GHz.......:(
I think its a T-bred "A"..Quote:
Originally posted by SPQQKY
What a drag, my chip is unlocked which was lucky, but what a turd of an OCer. I was hoping it was the cheap 8k9ai that was holding it back, but after dropping it in my 8rda+ and a Shuttle an35n ultra, it still tops out at 1.93GHz.......:(
how mutch vcore btw?
lemme get this straight before I give it a try...
The duron I'm going to cache-enable is a week 46 greeny. Do I fill in the laser valley or do I connect the contacts. I'm a little confused because I've seen both go by. From what I understand, though, you can't connect the contacts on the greens without scraping them off with a razor blade first?? :confused:
I've got a good result out of them by just raw fsb overclocking (my mobo doesn't seem to want to unlock them because it identifies them as Durons)
http://www.atomicmpc.com.au/forums.asp?s=2&c=6&t=4372
It's a MIXIB Week 37 on default voltage and §§§§ cooling.
i'm running a locked green 0344tpmw 1400 @ 217x10.5 @ 1.825 and an unlocked brown 0334vpmw 1600 @ 205x10 @ 1.7...i haven't tried the L2 mod on the 1400 cause i don't feel like scraping it...
I Might be getting me one of those Applebred's as well, 35euro's for the 1400 isn't much hehe :D
Wich one would be the best oc'er? 1400, 1500 or 1600?
id get a 1400 incase its locked...so you dont have to have like a 12 multiplier
ok :) ty
BUMP :P
Right, any news here? people with higher clocks etc? i killed my duron..... i think my mainboard took it with him when it died >< ... got a new one laying here now, bought some silver-glue and i'll go for the cache mod in a few minutes...
Bad week btw :( 0350