I set the Slot-T for 1.5v and 100Mhz FSB, and the mainboard jumpers for
133Mhz FSB. CPU is a retail SL5ZF, Malay, Week 23. I used Arctic Silver
instead of the supplied thermal interface.
Worth noting that the 1014beta3 BIOS was *required* - 1013 did not post
with the Tualatin CPU, although my P2B-DS runs PIII 1Ghz CPUs on 1013.
The BIOS reports 'Pentium III 1329 Mhz Processor'.
Memory is only PC100 CAS3, but was previously running at 2-2-2 on 103Mhz
FSB. No post at 2-2-2 on 133Mhz FSB, but ran 5 full Memtest86 passes
without error at 3-3-3. The system ran CPUStres in Windows at 100%
without a glitch for 2 hours until my son demanded his machine back.
Comparative benchmarks are about what I would expect - CPU up just over
3x (1A has twice the L2 cache) and memory bandwidth up 30%
Looks like a successful upgrade to me - I'll try the same thing on my
daughter's system tomorrow.
Hope others find the information useful.
/Simon
>I just upgraded my son's P2B-S rev 1.04. from Celeron 300A @ 464Mhz to
>Celeron 1A @ 1330Mhz using a Slot-T adapter. It proved to be a cheap and
>straightforward process - less than $100 CDN for the CPU and adapter.
>
>I set the Slot-T for 1.5v and 100Mhz FSB, and the mainboard jumpers for
>133Mhz FSB. CPU is a retail SL5ZF, Malay, Week 23. I used Arctic Silver
>instead of the supplied thermal interface.
>
>Worth noting that the 1014beta3 BIOS was *required* - 1013 did not post
>with the Tualatin CPU, although my P2B-DS runs PIII 1Ghz CPUs on 1013.
>The BIOS reports 'Pentium III 1329 Mhz Processor'.
>
Hi
Nice work ! You are just on the right side of the edge !
Good info to some that beta bios is required.
Obviously you have a voltage regulator chip that can handle 1.5 volt.
Chips that can provide voltages down to 1.3V:
HIP6019BCB
HIP6004CB
HIP6004BCB
US3007CW
http://n.ethz.ch/student/sroland/p2b_procupgrade_faq.html
P2B with different chip cannot be upgraded to tualatin unless the
powerleap adapter with its own voltage regulator is used (1.8 volt is
definitely too much for a tualatin in the long run).
best regards
John
Keith
"o&u" <johnkn...@mail.dk> wrote in message
news:lmia1v0qjri8c4mn4...@4ax.com...
>I have also heard of someone who modified his good old coppermine sloket
>(eg. MSI 6905 Master 2.3) and got a Tualatin working. Can anyone confirm
>that?
>
>Keith
>
Sure ! possible.
the mod:
http://www.overclockers.com/tips965/index04.asp
patching of the bios:
http://www.froggy.com.au/frogge/pepper/bmreport1.html
success stories for p2b:
http://www.geocities.com/_lunchbox/tualeron_success_table.html
Just more comfortable with the slot-T and further older coppermine
slockets are very hard to find.
best regards
John
Intel changed the processor signalling specification between .18
(Coppermine) and .13 (Tualatin) CPUs.
Modified Coppermine slotkets work, but the processor sees higher signal
voltages than it was designed for - which may affect stability and/or
longevity.
The Slot-T adapter has a voltage clamp chip which ensures all CPU
signals meet Intel's specification.
/Simon
I have a P2B-LS rev 1.02 and have a HIP6004ACB - won't do 1.3 volts. But I
want to get adventurous and replace it with a HIP6004BCB. While I'm there
I'm going to do the JEN and Photoshop bug reworks on it as well.
Newark Electronics sells the HIP6004BCB in singles, that's how I'm going to
get it.
Now where did you buy that Slot-T from, Simon?
Keith
"o&u" <johnkn...@mail.dk> wrote in message
news:lmia1v0qjri8c4mn4...@4ax.com...
Slot-T adapter comes from Upgradeware via StrattonComputer.com. I
know the link is on our site in the news section.
http://www.geocities.com/_lunchbox/
Cheers!
LunchBox wrote:
> Replacing the volatge regulator has been done before with success.
>
How about the clock generator?
I have a P2B-DS rev 1.04 which could run a pair of PIII-S CPUs on
Slot-Ts if I swapped the voltage regulators (there is one for each CPU)
but it wouldn't be worth the cost unless I could run them at (at least)
their rated speed.
I'm currently limited to 112Mhz FSB by the clock generator, but there's
a pin-compatible replacement available which will go to 150Mhz.
Anyone done a clock chip swap?
/Simon
The second one *should* have been a repeat performance since the
hardware is pretty much identical, but it wasn't quite...
Every time I set the FSB to 133Mhz, the BIOS complained 'Can't write
ESCD'. It would boot DOS fine, but W2K blue screened with inaccessible
boot device, and W98se would display the desktop but the HDD activity
LED and hourglass were permanent.
I Googled on the error message found several possible causes:
Bad BIOS flash - not likely since it worked fine at 124Mhz FSB, but I
re-flashed anyway. No change.
Flash memory worn out (can only be written ~1000 times) - not likely as
the motherboard is near new, and the flash can be written at any FSB
*except* 133Mhz.
Old NIC cards can cause this, specifically 3COM Etherlink III ISA - I
thought I'd found it! But no change after pulling the Elink III (and
everything else except the video)
I tried nailing up the HDD and memory configs in the BIOS, but it still
insists on trying to update ESCD every time you change the FSB.
I gave up, set the FSB to 124Mhz, booted Windows and installed SoftFSB.
The system runs rock-stable at 150Mhz FSB! Everything (CPU, PC100
memory, 440BX chipset) is overclocked 50% at default voltages and
handles all the torture tests I've thrown at it - unbelievable! I
suspected SoftFSB was lying to me about having set the FSB to 150, but
the benchmarks prove it.
P2B's aint dead yet!
/Simon
That sounds like a fun project. :)
This swap would be a little more complicated than just changing the
chip, which is why I ask if it's been done.
The FSB inputs to the clock generator are bi-directional. During power
on reset, the chip reads the FSB input pins and latches the values into
an internal register. After reset the same pins become clock outputs for
the PCI bus.
This dual-use requires the use of relatively high-value programming
resistors (10K) to minimise loading of the clock signal during normal
operation. The replacement clock generator has an identical pinout and
electrical characteristics, except that an additional PCI clock output
is defined as the 4th FSB input pin. My motherboard is almost certainly
missing the programming resistors for the 4th input pin, but they are
definitely required.
Ideally, I'd have access to a P2B-DS rev 1.06 D03 (with the new chip)
and could compare the two and just update mine - but unfortunately Asus
did not add the 4th FSB jumper on the updated board. However, they did
do so on the updated vanilla P2B... which leads me to suspect they had a
good reason for doing it differently on the S and DS versions. Methinks
I'd better have a clear understanding of this before taking a soldering
iron to my trusty motherboard ;-)
/Simon
I finally replaced the HIP6004ACB on my P2B-LS 1.02 with a HIP6004BCB. I'm
typing this on my box after the mod. While I'm there I did the JEN mod, put
some thermal grease under the greenie, and found out - to my pleasant
surprise - the Photoshop bug rework had been done on it.
Now I just need to mod my MS6905 master 2.3 and get a Tualeron 1.4A, now
that I can jumper it all the way to 1.3V...
I'm beginning to love older Asus boards, after all the mods done and doable
on a T2P4 and P2BLS.
Finally, am I ready for 100mhz slot 1 coppermines now?
Keith
"Simon Edwards" <si...@risca.com> wrote in message
news:y9pT9.7522$Yr2.1...@news20.bellglobal.com...
Do the soldering yourself? Any tips on technique?
/Simon
A 15W iron is enough for all these.
Soak a piece of wick in rosin flux paste and sucked as much solder as
possible off the pins, then take a sewing needle, stick it behind the pins,
and pry them up one by one while it's still being heated.
As the last pin is undone the old chip flips off on me (yeah it fliped me
off :-) Bye bye 6004A. See you in parts bin. :-)
More wick and flux to clean the pads.
Align new chip with pads, put two pieces of blutack on each end to hold it
in place, then solder it. Cleanup. Rinse and repeat with my other P2B-LS,
this time a rev 1.03.
Say hi to 6004B and 1.3V core voltage.
Right. I am a hardcore DIYer... :-)
Keith
"Simon Edwards" <si...@risca.com> wrote in message
news:2fqU9.767$%W2.1...@news20.bellglobal.com...
I finally got my Tualeron 1.4, modded the slotket and has it running
right now!
Only that it runs at 42C idle. Got to get a bigger sink and fan.
I have two Tualerons, one 1.0A SL5ZF, and the new 1.4 SL68G. Can
anyone tell me what are their overclocking potentials?
Keith
"Buurin" <ci259 @ nospace.torfree.net> wrote in message news:<ENLU9.13644$i%.2652138@localhost>...
> I have two Tualerons, one 1.0A SL5ZF, and the new 1.4 SL68G. Can
> anyone tell me what are their overclocking potentials?
Both have the same (older stepping). The 1.0A very likely will easily do
1.33Ghz (133Mhz FSB), likely even at default (1.5V) voltage. Depending
on your luck, might do 1.5Ghz (but don't increase the voltage to more
than 1.7V). The 1.4Ghz will probably reach around 1.6Ghz with sane
voltages. That said, I think your board can't do more than 112Mhz FSB
anyway, so you can't push them that much... (unless you also exchanged
the clock generator chip)
Roland
I now have 4 1.0A SL5ZF, all week 23 Malay, and all do 1.5Ghz at default
voltage - 3 of them only need 1.4V - on P2B-S boards with Slot-Ts.
AFAIK P2B-LS boards are the same as P2B-S, i.e. version 1.04/1.04. can
do 133Mhz FSB via jumpers and 150Mhz via SoftFSB.
I still haven't figured out how to add the missing 4th jumper to allow
the full range of FSB settings via jumpers, but I'll get to it....
/Simon
Now I have a chance to pick up a Thermaltake Golden Orb for US$4 on ebay
plus shipping. The listing says it is the smaller 64mm diameter version and
rated for up to 1.2GHz. Is it enough to cool my Tualeron 1.4 at 1.4V, and
can it cover the entire IHS? Saw it installed on a MS6905 slotket in a
review somewhere so I believe it fits physically.
There's also a 69mm Chrome Orb up for bid. Will it fit?
And while I'm on this subject, can I slip in a Crystal Orb in place of the
greenie and still clear the whatever Orb I put in?
Thanks
Keith
"Roland Scheidegger" <rschei...@gmx.ch> wrote in message
news:b0p2q6$rmv4t$1...@ID-84205.news.dfncis.de...