Changing 'default' FSB speed with Tualatin?

Spajky wrote:
On Thu, 22 Jul 2004 00:17:50 +1200, "~misfit~"
wrote:

caps bulging slightly
(The two biggest ones near the ATX connector and power FET's) and
won't run Coppermines anymore. Must get some replacement caps and
try to replace them.

there are a lot of dead MoBos around to canibalize them & use parts ..
:-)

I have done this before also.

Didn't you replace a cap or two on your 6163 Pro Spajky? Was it one
of those ones?

I was replacing caps few times partially till now, because I was
replacing only bulged ones & now almost all are replaced behind the
slotket & some others too!

The two caps that are bulging on the board I'm looking at now are EC5 and
EC6, 1,500mfd 10v and 1,000mfd 10v respectively. (Hard to read though, gold
on green)

Some of them for the second time & maybe tomorrow will replace again
one the third time, because its top is heating to much @ heavy tasks
(up to 60°C).

Maybe if you used good-quality new caps from a reputable company you
wouldn't need to replace them so often? I was speaking to a guy who repairs
mobos often yesterday and he said that, about one in twenty times, the
'through-plating' that the cap is soldered into pulls out of the mobo,
effectively destroying the board in most cases.

My experience for replacing caps: take phisically bigger (thicker &
taller if you can put them in) ones with greater capacity & voltage
for one value, especially the first left (back slash situated) two
after a toroidal coil (seen from the front way). So will heat less!!!
& last longer!

I *think* those are the ones that are failing on this board.

ALSO solder on these 2 ones terminals on the back bottom of a MoBO a
small (thin) 100nF block capacitor, so will heat even less!

There between 1st ram slot & slot 1 on the left from chipset close to
unused MoBo mounting hole is a small mosfet transistor which heats up

This is Q47?

to 70°C (remember I have fans @5V , case closed & is summer here); so
I soldered on its metal base soldered to MoBo additional small piece
of metal to act as an additional HS

I did not wanted to open my setup, but 3 weeks ago here was a storm &
I encountered a close lightning hit & had not much damage because I
have few surge protectors, but could not handle this time fully so
strong strike (was a real BigBANG). My neighbors around my house had
much much more damage! w/o that my surge protectors, now I would not
have anything to repair.

After a week or two later problems with my machine started to manifest
(illegal ops with Agent often for example) ...(bulged caps, one almost
exploded) Still have some problems with my PSU (in the meantime I
repaired it, than tried some mod & made a short & there was a true :-)
firework in my kitchen, repaired it again, but I think I would get a
new one soon cause IMHO it heats more than it used too sometimes

Yep, this July for me was quite an andrenaline fullfiled month ... :-)

Keeps you busy though. ;-)

I like these boards and, if possible would like to replace the defective
caps that were originally fitted with high-quality ones and hopefully not
have to keep replacing them in the future. It is written on the board that
it supports down to 1.3v for 'future CPU' so, with good components, surely
it shouldn't need regular replacement of caps.

Cheers,
--
~misfit~

~misfit~ wrote:

Spajky wrote:

On Thu, 22 Jul 2004 00:17:50 +1200, "~misfit~"
wrote:


caps bulging slightly
(The two biggest ones near the ATX connector and power FET's) and
won't run Coppermines anymore. Must get some replacement caps and
try to replace them.

there are a lot of dead MoBos around to canibalize them & use parts ..
:-)


I have done this before also.


Didn't you replace a cap or two on your 6163 Pro Spajky? Was it one
of those ones?

I was replacing caps few times partially till now, because I was
replacing only bulged ones & now almost all are replaced behind the
slotket & some others too!


The two caps that are bulging on the board I'm looking at now are EC5 and
EC6, 1,500mfd 10v and 1,000mfd 10v respectively. (Hard to read though, gold
on green)

If they're bulging then they're bad.

David Maynard wrote:
~misfit~ wrote:

Spajky wrote:

On Thu, 22 Jul 2004 00:17:50 +1200, "~misfit~"
wrote:


caps bulging slightly
(The two biggest ones near the ATX connector and power FET's) and
won't run Coppermines anymore. Must get some replacement caps and
try to replace them.

there are a lot of dead MoBos around to canibalize them & use parts
.. :-)


I have done this before also.


Didn't you replace a cap or two on your 6163 Pro Spajky? Was it one
of those ones?

I was replacing caps few times partially till now, because I was
replacing only bulged ones & now almost all are replaced behind the
slotket & some others too!


The two caps that are bulging on the board I'm looking at now are
EC5 and EC6, 1,500mfd 10v and 1,000mfd 10v respectively. (Hard to
read though, gold on green)

If they're bulging then they're bad.

Yep, for sure. That's why this board is sitting on my desk and isn't in a
PC.
--
~misfit~

~misfit~ wrote:

David Maynard wrote:

~misfit~ wrote:


Spajky wrote:


On Thu, 22 Jul 2004 00:17:50 +1200, "~misfit~"
wrote:



caps bulging slightly
(The two biggest ones near the ATX connector and power FET's) and
won't run Coppermines anymore. Must get some replacement caps and
try to replace them.

there are a lot of dead MoBos around to canibalize them & use parts
.. :-)


I have done this before also.



Didn't you replace a cap or two on your 6163 Pro Spajky? Was it one
of those ones?

I was replacing caps few times partially till now, because I was
replacing only bulged ones & now almost all are replaced behind the
slotket & some others too!


The two caps that are bulging on the board I'm looking at now are
EC5 and EC6, 1,500mfd 10v and 1,000mfd 10v respectively. (Hard to
read though, gold on green)

If they're bulging then they're bad.


Yep, for sure. That's why this board is sitting on my desk and isn't in a
PC.
--
~misfit~



Seeing you post about it did remind me to get off my butt and order the
replacement FET for the BH6 board, though.

On Thu, 22 Jul 2004 14:56:11 +1200, "~misfit~"
wrote:

The two caps that are bulging on the board I'm looking at now are EC5 and
EC6, 1,500mfd 10v and 1,000mfd 10v respectively.

yes, that are the ones I replaced them once, now Ok

Maybe if you used good-quality new caps from a reputable company you
wouldn't need to replace them so often?

they were L.E.S.R type & Ok;
the ones I replaced the second time were problematic after recent
lightning strike !!!! (not before!)

My experience for replacing caps: take phisically bigger (thicker &
taller if you can put them in) ones with greater capacity & voltage
for one value, especially the first left (back slash situated) two
after a toroidal coil (seen from the front way). So will heat less!!!
& last longer!

EC7 & EC9

ALSO solder on these 2 ones terminals on the back bottom of a MoBO a
small (thin) 100nF block capacitor, so will heat even less!

again will replace today EC7 just with bigger (is not bulging, just
prevention

& will use the old noisy PSU, which was not hit by a lightning strike
temporary to see if any probs still later

There between 1st ram slot & slot 1 on the left from chipset close to
unused MoBo mounting hole is a small mosfet transistor which heats up

This is Q47?

YES

It is written on the board that
it supports down to 1.3v for 'future CPU' so, with good components, surely
it shouldn't need regular replacement of caps.

shouldn`t if no more than Cu-Mines were run on the board & using
quality PSU etc & surge protectors for PC & case normally ventilated .

mine developed bad caps one year & half after using Tualatin :-(
--
Regards, SPAJKY ®
& visit my site @ http://www.spajky.vze.com
"Tualatin OC-ed / BX-Slot1 / inaudible setup!"
E-mail AntiSpam: remove ##

Spajky wrote:
On Thu, 22 Jul 2004 14:56:11 +1200, "~misfit~"
wrote:

The two caps that are bulging on the board I'm looking at now are
EC5 and EC6, 1,500mfd 10v and 1,000mfd 10v respectively.

yes, that are the ones I replaced them once, now Ok

Cool, I'll order replacements.

Maybe if you used good-quality new caps from a reputable company you
wouldn't need to replace them so often?

they were L.E.S.R type & Ok;
the ones I replaced the second time were problematic after recent
lightning strike !!!! (not before!)

Ok, I wondered if it was due to recycling caps.

My experience for replacing caps: take phisically bigger (thicker &
taller if you can put them in) ones with greater capacity & voltage
for one value, especially the first left (back slash situated) two
after a toroidal coil (seen from the front way). So will heat
less!!! & last longer!

EC7 & EC9

I see them, they're fine on my board but I've only used Mendicino and
Coppermines in it so far. I've just ordered a Slot-T adapter. I also have
another 6163 not in use at the moment, fitted with a Celly 600@900 that was
working well. I'll use the Slot-T in that one with a Tui 1.4Ghz I have until
I get this board fixed. Plus I have one other 6163 that is in-use with a
Mendicino 500. Three in all. I like these boards.

ALSO solder on these 2 ones terminals on the back bottom of a MoBO a
small (thin) 100nF block capacitor, so will heat even less!

again will replace today EC7 just with bigger (is not bulging, just
prevention

& will use the old noisy PSU, which was not hit by a lightning strike
temporary to see if any probs still later

There between 1st ram slot & slot 1 on the left from chipset close
to unused MoBo mounting hole is a small mosfet transistor which
heats up

This is Q47?

YES

Thought so. I haven't temperature-tested mine.

It is written on the board that
it supports down to 1.3v for 'future CPU' so, with good components,
surely it shouldn't need regular replacement of caps.

shouldn`t if no more than Cu-Mines were run on the board & using
quality PSU etc & surge protectors for PC & case normally ventilated .

But Cu-mines never went below 1.5v (AFAIK). I was assuming that, because
that is written on the board then the circuitry is designed to be capable of
supplying reasonably high current at low voltage and shouldn't be a problem
supplying Tui's.

mine developed bad caps one year & half after using Tualatin :-(

If all goes well with the Slot-T and the cap replacement then I'm thinking
of switching all three of my 6163's over to Tui's on Slot-T's. I have a 1.3
and a 1.4 already with a 1.2 arriving sometime over the next week hopefully.
I haven't checked the board that is currently running the Mendicino for
bulging caps recently but it wouldn't surprise me if a few are bulging, it
didn't want to run a Cu-mine last time I tried one in it.

I'm also expecting a P3 1Ghz on an MSI board next week. That's all I know
about it. I don't know if it's a Cu-mine or Tui, (Did they make a P3 Tui
1Ghz?) although I think it's a Cu-mine. I have no idea what the board is,
whether it's a Slot 1 or a Socket 370.

It seems a shame to be retiring my Cu-mine 600's, they've done me well,
running at 900 for a while now. Also it's a shame that the BX only supports
ATA33 IDE. Maybe I should be looking out for some PCI ATA100 IDE boards too.

I have an Abit BX133 RAID that has a HighPoint ATA100 RAID controller
on-board. I'm not using it in RAID configuration, just single ATA133 drive.
It makes a huge difference to responsiveness compared with other BX boards.
Very fast to load programs. It used to run a Tui 1.3 in an Upgradeware 370GU
adapter but when I got it it had *10* caps bulging/leaking and would only
get halfway through POST. I replaced the caps with recycled ones but it
still won't run the Tui anymore. It's running a Celly 900 at stock. Maybe I
should look at replacing the caps again with new ones. However, sourcing the
exact caps needed and fitting them is a bit tricky. Not to mention the
expense. Good low ESR caps are expensive in small quantities, plus the only
company that seems to carry the requried range of stock in NZ only accepts
payment by credit card, something I don't have.

Cheers Spajky.
--
~misfit~

~misfit~ wrote:

P2B wrote:

Spajky wrote:


On Tue, 20 Jul 2004 10:14:52 +1200, "~misfit~"
wrote:



Also, Spajky, what speed do you have your Tui at and how many marks
does it get with CPUMark99 on a BX?


1.0A@1,35G
I tested it with PCMark2002 & I get 3400 points ...

CPUMark99, 1.0A@1500Mhz = 132
CPUMark99, P3-S@1575Mhz = 143

(both on a P2B-S)

More Tualatin on BX benchmarks at tipperlinne.com/benchmark.htm


Wow! Those are good numbers. I might have to have another attempt at getting
one of these Tuis to work in a BX board. I wish I could find a couple of
P2B's.

I checked your page, your CPUMark figure for a Tui Celeron 1400 (14 x 100)
of 123 is amazing. I tried the same CPU in this board I have and got 87.3.
Huge difference.

Note that my figure of 123 was for 10 x 140Mhz, and is consistent with
Spajky's 118 for 10 x 135Mhz. Your 87.3 would have been for 14 x 100Mhz
- so when Spajky says "Fsb doesn´t matter much" I must disagree :-)

FSB rules on a P2B, 'cos memory is the bottleneck.

P2B

On Fri, 23 Jul 2004 01:14:04 +1200, "~misfit~"
wrote:

shouldn`t if no more than Cu-Mines were run on the board & using
quality PSU etc & surge protectors for PC & case normally ventilated .

But Cu-mines never went below 1.5v (AFAIK). I was assuming that, because
that is written on the board then the circuitry is designed to be capable of
supplying reasonably high current at low voltage and shouldn't be a problem
supplying Tui's.

The caps for Tuallies on board should be around 50% greater value than
for older Cpus, because Tualies suck more current & is advisable to
mount some HS on original voltage regulators ....

Also it's a shame that the BX only supports
ATA33 IDE. Maybe I should be looking out for some PCI ATA100 IDE boards too.

not much advantage over ATA33 ! in real life!
Test your drive on ATA33 & ATA 100 setting on some board with
PCmark2002 & compare the index; you will see there is not much of
difference !!! in real life; its mostly marketing trick like AGP8x
stuff ...

PS.: maybe my Mobo is a bit damaged; Ram sticks are heating a bit & I
do not remember before being warm (damn lightning strike :-(
.... maybe its time to get another one for replacement in the future ..
--
Regards, SPAJKY ®
& visit my site @ http://www.spajky.vze.com
"Tualatin OC-ed / BX-Slot1 / inaudible setup!"
E-mail AntiSpam: remove ##

~misfit~ wrote:
[snip]

It seems a shame to be retiring my Cu-mine 600's, they've done me well,
running at 900 for a while now. Also it's a shame that the BX only supports
ATA33 IDE. Maybe I should be looking out for some PCI ATA100 IDE boards too.

That's why I prefer the P2B variants with onboard U2W SCSI - very few
SCSI drives are capable of fully saturating the 80MB/s SCSI bus on these
boards, but U160 drives with sustained throughputs around 65MB/s can now
be had at fairly reasonable prices on the used market.

I was lucky a few weeks back and got a dead Adaptec 3400S raid
controller (4 channel U160) thrown in with a lot of very reasonably
priced U160 drives. I fixed it and built a RAID 5 file server, with a
P2B-S and Tualeron 1A@1500 - now I need gigabit NICs to take advantage
of the throughput, because it's more than capable of saturating my
100meg home network!

Long live BX and the P2B series!

P2B

On Thu, 22 Jul 2004 22:28:19 -0400, P2B wrote:

Note that my figure of 123 was for 10 x 140Mhz, and is consistent with
Spajky's 118 for 10 x 135Mhz. Your 87.3 would have been for 14 x 100Mhz
- so when Spajky says "Fsb doesn´t matter much" I must disagree :-)

in that particular bench does not: (math bench is that!) mine bench
result (even if on 135fsb) is smaller than on yours at 100Fsb a bit
less than 4% like the final frequency (1,4 vs 1,35) difference is ...
Yes, fsb metters much 'cos memory is the bottleneck, but in this case
looks like CPUMark99-the bench does not use external memory; whole
bench file is loaded in L2 as it seems!

But his 87 points show some problem with CPU (partially ? damaged L1
or L2 ? or disabled in Bios or **** chipset like Ali ?) ...

or, LOL he has popped Tuallie on a non Tuallie MoBo & defaults were
set in Bios & runs really at 66-fsb (933MHz) & did not check the speed
in Win with Sandra for instance! :-))))) My bios does this with
garbage at Post !!! - I have to manualy set Fsb ... & since there was
no older Cu-Mine at 100 Fsb & 933MHz clock with 14 multiplier, he did
not see that !!!! With "ordinary" work, there is cumulative only 25%
loss in performance of a system going downClocking Tuallies as I
benched time ago; surfin´the web & reading Usenet?-he wouldn`t notice
it... LOL

If so, no wonder that result (if calculated from yours CPU to his
DC-ed, result would be quite similar ... :-))))
--
Regards, SPAJKY ®
& visit my site @ http://www.spajky.vze.com
"Tualatin OC-ed / BX-Slot1 / inaudible setup!"
E-mail AntiSpam: remove ##