Ping Spajky: MSI 6163 Q again!

[~misfit~]

Hi Spajky,

I'm about to put a celly 900 coppermine in a slocket in one of my 6163 Pros.
(It has the blue sticker to denote coppermine compatability) I'll be using
the 'MS6905 Master ver:2.3' slocket. Should I just leave the jumpers on the
slocket set to 'Auto' for CPU voltage and let the BIOS handle it?

Also, the guy who is sending me the CPU says it will do 1200Mhz with a small
over-volt, does that sound right to you? If I set the FSB to 90 to keep the
PCI bus low it will be 1215Mhz. In your experience, is this CPU likely to
run OK at that speed?

I'm also getting a Celly 600 coppermine to put in another. That one will be
using a MS6905 ver:1.1 slocket as I don't have another 'Master'. I don't
know what Mhz that CPU will do. Both these machines are currently running
Mendicino 400s.

Thanks Spajky,

Regards,

 

[~misfit~]

Hey Dave,

?
What benefit would there be to running the FSB @ 90?
Isn't that one of the mid-to-later BX boards that implements a 1/4 PCI
divider at >= 124MHz FSB?

I'm not sure at what point the 1/4 divider kicks in, I know the 1/3 divider
kicks in at 90. I'm also not sure of the AGP divider, Spajky did tell me
that at 133 the AGP is running at 89Mhz so I guess it's running at spec at
100.

See if the northbrige stays cool enough and try for 133MHz FSB.
I'd try leaving the slotket voltage set to auto IF voltage can be set
in the bios, and on the slotket set the FSB to 133. I've had a Celery
900 that needed 1.85V for 1.2GHz, other Celerons needed around the
same, 1.8-1.9. I'd try 1.8 first.

At the moment, with the Mendicino cellies the voltage is set to auto on the
slockets and I have the option in the BIOS of Auto, 2.0, 2.05, 2.1, 2.2, 2.3
and 2.4 volts vcore.

I'm right in thinking the coppermines are 1.75v default?

The cA0, cB0 steppings are borderline at 900+MHz (IIRC) but cC0 should
easily run at 900MHz or higher. It "might" even run at 1200MHz too, I
had a Celery 600 cC0 that would, though I think the "average" Celery
600 o'c to a bit less, 1.0-1.1GHz, unless using extreme voltage, but
in my experience volting it over 1.95 has little to no benefit unless
using extreme cooling too, which doesn't seem worthwhile today.

I'm not sure how to tell what stepping the CPU is. I don't have these CPUs
in my hand yet, they're in the post. Is there a number I can look up on them
when I get them? Or do I just fit them and use something like Aida32?

I don't want to go for extreme overclocks on these, I'd just like to get a
little more out of them without sacrificing stability. Also I'm on a tight
budget so extreme cooling doesn't come into the equation either. As well, I
don't want to endanger my peripherals by having really out-of-spec busses. I
already had a HDD fail on one of these boards from running the PCI bus at
41.5Mhz (83FSB) although it could have been that the drive just wasn't very
robust. I don't want to risk damaging the craphics cards especially, they're
both running GeForce 2 MX400/64Mb cards which, although they are old tech, I
can't afford to replace.

I guess trying for 100FSB with both of them may be OK. The PCI and AGP
busses will be running to spec and it's a reasonable overclock. (I've just
Xposted this to a.c.h.o, I may get some more feedback by having it there as
well). I'll be running the standard HSFs on them, although I do have a bit
of AS3 I'll use.

Thanks for your input Dave, anything else you (or anyone else) think(s) may
be relevant would be appreciated.

Cheers,

 

[~misfit~]

Hey Dave,

Ok, just realised from a post by David Maynard (thanks David) in a.c.h.o
that the 900 runs on a 100Mhz FSB anyway, now I see why the '?'. I was
labouring under the mistaken belief that all pre-Tualatin cellies ran on
66Mhz.

 

[kony]

..

I'm not sure at what point the 1/4 divider kicks in, I know the 1/3 divider
kicks in at 90. I'm also not sure of the AGP divider, Spajky did tell me
that at 133 the AGP is running at 89Mhz so I guess it's running at spec at
100.

I"d make a backup of the HDD and give it a go... if the PCI bus is out
of spec you'd probably see HDD problems, then perhaps USB and NIC.
It's true that the AGP bus would be at 89MHz, but IMHO, that's
actually a good thing IF the video card tolerates it, since at 2X mode
it's a bit bottlenecked for newer video cards. I believe Spajky has
on his website a few methods to attempt overcoming the AGP bus speed
issue. In my own experience one of the most significant details is to
disable sideband addressing, as it seem to cause problems at the
higher FSB speeds. Sometimes the BIOS will allow disabling it, but
other times it's necessary to mod the video card bios, if you're
fortunate enough to be using a card supported by the needed bios
utilities (especially nVidia Geforce and latter).

The AGP bus would indeed be in spec again at 100MHz FSB. There were
early BX boards that would sync the AGP bus to the FSB when CPUs like
a Celeron < 800MHz was o'c, since it'd have default 66MHz FSB. This
"issue" was largely nonexistant with latter BX boards, or at least
there was a setting to choose.

At the moment, with the Mendicino cellies the voltage is set to auto on the
slockets and I have the option in the BIOS of Auto, 2.0, 2.05, 2.1, 2.2, 2.3
and 2.4 volts vcore.

Then I'd expect you'll have a +.4V range over the default of whichever
CPU you install, OR over the voltage set by the slotket.

I'm right in thinking the coppermines are 1.75v default?

I "think" around the time that Intel released those faulty PIII
1.13GHz chips, at that point the PIII at 1.0-1.13 used 1.75V, but
otherwise they all used lower voltage. Celery in the cA0 and cB0 used
1.5V, then moved up in voltage to 1.7V at the end. So you can
basically tell the stepping just by checking the voltage on the label.
If it's a 1.5V chip it'll probably not o'c much past 900MHz even if
you do increase the voltage except in rare cases.

I'm not sure how to tell what stepping the CPU is. I don't have these CPUs
in my hand yet, they're in the post. Is there a number I can look up on them
when I get them? Or do I just fit them and use something like Aida32?

Intel has a Celeron datasheet with a table, or to get a rough idea
just note the voltage as mentioned above.

I don't want to go for extreme overclocks on these, I'd just like to get a
little more out of them without sacrificing stability.

1.2Ghz is "usually" about the upper limit, it might not be 100% stable
at that speed... some are, others aren't.

Also I'm on a tight
budget so extreme cooling doesn't come into the equation either.

Well, extreme is relative... these days even the worse Athlon 'sinks
are plenty good enough for a Coppermine, even o'c and overvolted.
Whatever you have handy that's post-Pentium 1 is probably good enough
unless you exceed 1.8V or have poor chassis airflow. The main issue
is the same as always, how much fan noise you can stand. I've used
$1 K6 heatsinks on ~1GHz celerons in SFF systems, they mainly had a
nice thick Al base which compensated for their overall smaller size.

As well, I
don't want to endanger my peripherals by having really out-of-spec busses. I
already had a HDD fail on one of these boards from running the PCI bus at
41.5Mhz (83FSB) although it could have been that the drive just wasn't very
robust. I don't want to risk damaging the craphics cards especially, they're
both running GeForce 2 MX400/64Mb cards which, although they are old tech, I
can't afford to replace.

Well, we both know that to a certain extent you're always choosing to
risk hardware when overclocking... Within margins the hardware is
designed to support, like running the Celery 600 @ 900, is pretty
safe, but with the Celery 900 you're either: A) Pulling a lot of
power though a motherboard that was never intended to run 1.2GHz
CPUs... it may work fine, or might have decreased lifespan. B)
Running at somewhat lower and having out-of-spec PCI bus. If the
board supports 124MHz FSB with 1/4 PCI divider then that "might" be an
ideal setting. The cC0 and later Celerons needed little to no voltage
increase for ~1.0GHz, but the voltage needed starts to climb after
that. This seemed to be the trend, but it can hardly be applied to a
single specimen of CPU.

I guess trying for 100FSB with both of them may be OK. The PCI and AGP
busses will be running to spec and it's a reasonable overclock. (I've just
Xposted this to a.c.h.o, I may get some more feedback by having it there as
well). I'll be running the standard HSFs on them, although I do have a bit
of AS3 I'll use.

I wouldn't even use the AS3 if you have generic compound avaiable.
After you've ran an Athlon you become a lot more comfortable with 45C
temps from a Coppermine, it's really not hard at all to keep them
cool enough.

Thanks for your input Dave, anything else you (or anyone else) think(s) may
be relevant would be appreciated.

Cheers,

I expect the MSI Master slotket will work great, but is the older one
even Coppermine-compatible? It could be, I just don't remember.


Dave

 

[~misfit~]

I don't remember off hand if the v1.1 is coppermine compatible (I've even got
one here somewhere and can't remember for sure but my gut tells me it wasn't)
but you can't deduce anything from the 66/100 jumper as they ALL had that for
the overclocking appeal. (I mean, think about it. If the processor was SUPPOSED
to run 100 MHz FSB you wouldn't need a jumper to 'force' it.)

Agreed. But then again, why would there be a jumper position to
"automatically detect 66 or 100" if it was never intended to run 100Mhz FSB
CPUs? To force 100FSB you remove the jumper. Not that I doubt you, far from
it.

To answer your question, no, there weren't any PPGA processors (all Celerons)
that were spec'd to run 100 MHz FSB in stock form. Some, like the famous Celeron
300a did, when overclocked, though. Socket 370 went 100MHz FSB with the
flip-chip coppermine FC-PGA P-IIIs and later Celerons of 800 MHz and up.

Wait a minute. I think the MSI v1.1 was the one I modded for my P-III so no, it
isn't.

So what do you think is the likely outcome if I try to run my
soon-to-be-here Celly 600 Coppermine in the 1.1 slocket? Will I kill
anything? It'll annoy me to have a 600 coppermine I can't use while a
machine is running a 400 Mendicino. Especially when the board is
Cu-mine-compatable and is running the latest BIOS released for it.

How do I modify the slocket to run a Cu-mine?

Thanks David.

 

[Spajky]

I'm about to put a celly 900 coppermine in a slocket in one of my 6163 Pros.
(It has the blue sticker to denote coppermine compatability) I'll be using
the 'MS6905 Master ver:2.3' slocket. Should I just leave the jumpers on the
slocket set to 'Auto' for CPU voltage and let the BIOS handle it?

YES !

Also, the guy who is sending me the CPU says it will do 1200Mhz with a small
over-volt, does that sound right to you? If I set the FSB to 90 to keep the
PCI bus low it will be 1215Mhz. In your experience, is this CPU likely to
run OK at that speed?

Who knows, depends on the sample!

Why you rather do not take a Slot-T & lowest Tuallie & OC it a bit?
and slotket v.2.3 use for another Cu-Mine?

I'm also getting a Celly 600 coppermine to put in another. That one will be
using a MS6905 ver:1.1 slocket as I don't have another 'Master'. I don't
know what Mhz that CPU will do. Both these machines are currently running
Mendicino 400s.

This slotket is only PPGA one, to run it for FcPga, you have to mod
it! or use the previous one
http://www6.tomshardware.com/cpu/00q1/000229/socket370-to-coppermine-04.html

PS for Tuallies & this MoBo Bios Vcore: (v.3.3 bios at least is like
this)
Vcore jumpers or Vid pins shorting:
if set to 1,5 - 1,65, you will have there 1,65
if set to 1,475 or lower, you will have there 1,47 or less in bios!
I did the last: Vss-Vid0-Vid1 to get 1,45 default & than rise it for a
step....



-- Regards, SPAJKY
- http://freeweb.siol.net/jerman55/HP/Spajky.htm
Celly-III OC-ed,"Tualatin on BX-Slot1-MoBo!"
E-mail AntiSpam: remove ##

 

[~misfit~]

Pci 1/4 divider kicks in @ 122fsb,124fsb -depends if spreadSpectrum is
on or Off & Agp 2/3 kicks in @ 75fsb ... on my setup according to
Sandra (checked right now); but MAY differ with another
slotket/CPU/Bios ver. combo !!!! Check with BusRace proggy-on MoBo
drivers CD & Sandra!

Thanks again. It's really handy knowing someone who is knowledgeable, likes
to play with tweaking, and runs the same board as I do.

 

[~misfit~]

YES !

Who knows, depends on the sample!

Why you rather do not take a Slot-T & lowest Tuallie & OC it a bit?
and slotket v.2.3 use for another Cu-Mine?

I'm simply unable to afford it. The only reason I'm doing what I am is that
I can sell the Celly 400s for about the same as I'm paying for the 900 and
600. (I got a good deal from friends)

This slotket is only PPGA one, to run it for FcPga, you have to mod
it! or use the previous one

http://www6.tomshardware.com/cpu/00q1/000229/socket370-to-coppermine-04.html

Thanks for the link. I've bookmarked it. It doesn't look too difficult, I'll
just need to take a file or grinder to my soldering iron tip and pick up a
couple of resistors maybe.

PS for Tuallies & this MoBo Bios Vcore: (v.3.3 bios at least is like
this)
Vcore jumpers or Vid pins shorting:
if set to 1,5 - 1,65, you will have there 1,65
if set to 1,475 or lower, you will have there 1,47 or less in bios!
I did the last: Vss-Vid0-Vid1 to get 1,45 default & than rise it for a
step....

Thanks Spajky, you've been a big help as always.

 

[~misfit~]

Simply because that's the purpose of the signal per Intel specs and it needs to
be connected (the 'normal' configuration) for the PPGA BSel line to tell the
motherboard it's 66 MHz (pulled low). Disconnect the pin (remove the jumper) and
the motherboard pull-up pulls the pin high and that selects 100 MHz FSB. That is
the same line (literally) called the "B21 trick" for the slot-1 Celerons; it's
just that slotkets provide a jumper to do the 'trick' rather than you having to
tape the pin.
Remember that the slot-1 board you're plugging it into IS designed to accept a
slot-1 cartridge that could be either 66 or 100 MHz, depending on that pin, and
the slotket is simply emulating a slot-1 cartridge, with the added 'feature' of
you being able to manually change the B21 wiring.

Ok, thanks, understood.

It shouldn't hurt anything to try it un modded. It just won't run because the
reset lines were modified in the FC-PGA pinout, except there's a potential 'good
news' twist. Depending on the stepping of your 600 it might not need a mod as
the early FC-PGA celerons were PPGA compatible (so they could be used in
'celeron' PPGA motherboards.). For example, My 566 (that I overclocked to 1020
didn't need the mod. Of course, odds are it's a newer stepping.

So I'll try it first in the slocket un-modded if it won't damage the CPU on
the off-chance mine doesn't need the mod.

Note: there were never any 'compatible' FC-PGA P-IIIs because there was no
such thing as a "PPGA P-III Motherboard." PPGA socket motherboards were always
Celeron only (no socket P-IIIs existed when PPGA came out) so there was nothing,
in Intel's mind, for an FC-PGA P-III to be reverse 'compatible' with.


Here's a link http://www.3feetunder.com/krick/370mod/ and another link
http://www6.tomshardware.com/cpu/20000229/index.html

I pulled the AM2 pin on the processor package itself but you can insulate it
other ways.

Once again, thanks very much. I'm sure I can do that mod, I just need to
take a file/grinder to my old soldering iron to make the point finer.

Cheers,

 

[~misfit~]

It shouldn't hurt anything to try it un modded. It just won't run because the
reset lines were modified in the FC-PGA pinout, except there's a potential 'good
news' twist. Depending on the stepping of your 600 it might not need a mod as
the early FC-PGA celerons were PPGA compatible (so they could be used in
'celeron' PPGA motherboards.). For example, My 566 (that I overclocked to 1020
didn't need the mod. Of course, odds are it's a newer stepping.

Note: there were never any 'compatible' FC-PGA P-IIIs because there was no
such thing as a "PPGA P-III Motherboard." PPGA socket motherboards were always
Celeron only (no socket P-IIIs existed when PPGA came out) so there was nothing,
in Intel's mind, for an FC-PGA P-III to be reverse 'compatible' with.


Here's a link http://www.3feetunder.com/krick/370mod/ and another link
http://www6.tomshardware.com/cpu/20000229/index.html

I pulled the AM2 pin on the processor package itself but you can insulate it
other ways.

Update on the Celly 600:

I tried it in the ver 1.1 slocket and it didn't work. So I modified the
slocket, soldered the wire on the back, then went to break off the pin on
the CPU. Well, you're gonna think I'm a right idiot but I broke off the
wrong pin!!! I'm used to looking at Pentium 1 era CPUs were there is only
one corner that has a pin missing so I looked at this corner, pin missing so
it must be the right corner, and proceeded to wriggle it until it broke off.
I then held the CPU up to check I hadn't bent any other pins too much and...
DOH!! there were two corners with a pin missing and I'd broken off the wrong
one!

I decided that maybe the pin I'd broken off maybe wasn't needed so proceeded
to break off the correct pin. I then fitted it in the modified slocket, put
it in the mobo and..... Nothing.

So, I decided I could fix it. I applied the hot soldering iron briefly to
the pit where the pin came from to deposit a bit of solder there (Thank God
it is an easy-to-access pin) tinned the pin, held it in place with my
shaking 40-something hands with tweezers and applied the soldering iron to
the side of the pin. It held!! *And* it works!!

Currently running at 810Mhz (Didn't like 900) and 1.75vcore and running
Prime95 torture test, SETI and playing mp3s at the same time. It has been
running like that for a couple of hours with Prime95 not reporting any
errors yet. If it goes well for a while I may try dropping the vcore a bit,
maybe it doesn't need 1.75v.

(It booted at 900Mhz/1.8vcore but had artifacts on the screen and Prime95
crashed on the first test)

So, all-in-all I think I'm damn lucky. The 900 should arrive tomorrow, thank
God I have an MSI Master slocket for that one and shouldn't have to modify
anything.

Cheers,

 

[Spajky]

then went to break off the pin on
the CPU. Well, you're gonna think I'm a right idiot but I broke off the
wrong pin!!! I'm used to looking at Pentium 1 era CPUs were there is only
one corner that has a pin missing so I looked at this corner, pin missing so
it must be the right corner, and proceeded to wriggle it until it broke off.
I then held the CPU up to check I hadn't bent any other pins too much and...
DOH!! there were two corners with a pin missing and I'd broken off the wrong
one!

LOL, you were not the first one... heard about others too..

So, I decided I could fix it.

Obviously, there was nothing to loose anymore!

So, all-in-all I think I'm damn lucky.

Yes, Yes ... , now do not touch it anymore ...

-- Regards, SPAJKY
- http://freeweb.siol.net/jerman55/HP/Spajky.htm
Celly-III OC-ed,"Tualatin on BX-Slot1-MoBo!"
E-mail AntiSpam: remove ##

 

[~misfit~]

LOL, you were not the first one... heard about others too..

That's a relief.

Obviously, there was nothing to loose anymore!


Yes, Yes ... , now do not touch it anymore ...

Not a chance! That CPU is never coming out of that slocket again. It works
as it is but it wasn't exactly a robust repair.

Do you think I should try dropping it back from 1.75vcore? It's running fine
at the moment and doesn't seem to be very hot (touching the 'sink, It's
spec'ed at 1.5v) Prime95 torture-test has been running for 7 hours now with
no errors. (FSB of 90Mhz for CPU speed of 810 instead of 600 spec'ed.)

Cheers,

 

[~misfit~]

I vaguely recall having a Celery 566 cB0 (SL3W7)(also the older 1.5V
spec'd CPU) that'd do 85MHz @ 1.75, but when reduced to 1.7V it was
slightly instable... YMMV.

Thanks for that Dave, I'll probably leave it is it is then, this is a cB0 as
well according to Aida32. SL3W8 on the chip.

 

[David Maynard]

Update on the Celly 600:

I tried it in the ver 1.1 slocket and it didn't work. So I modified the
slocket, soldered the wire on the back, then went to break off the pin on
the CPU. Well, you're gonna think I'm a right idiot but I broke off the
wrong pin!!! I'm used to looking at Pentium 1 era CPUs were there is only
one corner that has a pin missing so I looked at this corner, pin missing so
it must be the right corner, and proceeded to wriggle it until it broke off.
I then held the CPU up to check I hadn't bent any other pins too much and...
DOH!! there were two corners with a pin missing and I'd broken off the wrong
one!

I decided that maybe the pin I'd broken off maybe wasn't needed so proceeded
to break off the correct pin. I then fitted it in the modified slocket, put
it in the mobo and..... Nothing.

So, I decided I could fix it. I applied the hot soldering iron briefly to
the pit where the pin came from to deposit a bit of solder there (Thank God
it is an easy-to-access pin) tinned the pin, held it in place with my
shaking 40-something hands with tweezers and applied the soldering iron to
the side of the pin. It held!! *And* it works!!

Currently running at 810Mhz (Didn't like 900) and 1.75vcore and running
Prime95 torture test, SETI and playing mp3s at the same time. It has been
running like that for a couple of hours with Prime95 not reporting any
errors yet. If it goes well for a while I may try dropping the vcore a bit,
maybe it doesn't need 1.75v.

(It booted at 900Mhz/1.8vcore but had artifacts on the screen and Prime95
crashed on the first test)

So, all-in-all I think I'm damn lucky. The 900 should arrive tomorrow, thank
God I have an MSI Master slocket for that one and shouldn't have to modify
anything.

Cheers,

I know what you mean. I sweat bullets every time I pull pins. I didn't 'pull'
the wrong one, came close though, but I've knocked a couple off (more than one
processor) when trying to connect to the power pin on Tualatins. Fortunately,
that's not a critical pin (which is why I picked it).

I am impressed you got it soldered back on but you've probably used up your good
luck for at least a decade with that one. <g>

Sorry I missed your desperation search for celeron pinouts but I was over here
trying to get the blasted Windows 2000 system that was on the BH6 running on a
chaintec via mobo because the BH6 just spontaneously blinked and died last
night. Poof... all she wrote. Firewall got screwed up, 'lost', in the maze of
hardware changes and I had a heck of a time getting that straightened out but I
just got networking operational again.

Looks like I overheated the BH6 regulator. It was running in a hot room nestled
under a cabinet, just to make sure the heat couldn't get out. I think I got a
spare FET so maybe I can bring that dinosaur back to life.

 

[~misfit~]

I know what you mean. I sweat bullets every time I pull pins. I didn't 'pull'
the wrong one, came close though, but I've knocked a couple off (more than one
processor) when trying to connect to the power pin on Tualatins. Fortunately,
that's not a critical pin (which is why I picked it).

I am impressed you got it soldered back on but you've probably used up your good
luck for at least a decade with that one. <g>

I reckon. I didn't think it was possible to put pins back!

Sorry I missed your desperation search for celeron pinouts but I was over here
trying to get the blasted Windows 2000 system that was on the BH6 running on a
chaintec

Ug! Chaintec? Not my favourite mobo. Had one once, got rid of it pretty
quickly.

via mobo because the BH6 just spontaneously blinked and died last
night. Poof... all she wrote. Firewall got screwed up, 'lost', in the maze of
hardware changes and I had a heck of a time getting that straightened out but I
just got networking operational again.

Can be a bitch huh? My network all hangs off a Celly 400 (and another
MS-6163 Pro) and I don't overclock it and try to keep it simple. I'm
probably gonna put a PIII katmai in it when my Celly 900 arrives (should
have been here by now, the guy who sent it to me just emailed me and asked
if I'd got it yet, said he sent it days ago) I'd be lost without it, at
least for the best part of a day until I set up something else. I have a
spare Cyrix PR333 with XP Pro installed that I keep all set up in the
cupboard for just such an emergency. Not a great machine by any means but
would do at a pinch for a gateway PC.

Looks like I overheated the BH6 regulator. It was running in a hot room nestled
under a cabinet, just to make sure the heat couldn't get out. I think I got a
spare FET so maybe I can bring that dinosaur back to life.

I have mine mounted on a pull-out mobo tray on the wall, no case at all.
Hopefully that gets around any potential heat problems. Good luck with
replacing the FET.

Cheers,

 

[~misfit~]

than

I reckon. I didn't think it was possible to put pins back!
over running
on a

Ug! Chaintec? Not my favourite mobo. Had one once, got rid of it pretty
quickly.
maze out
but I

Can be a bitch huh? My network all hangs off a Celly 400 (and another
MS-6163 Pro) and I don't overclock it and try to keep it simple. I'm
probably gonna put a PIII katmai in it when my Celly 900 arrives (should
have been here by now, the guy who sent it to me just emailed me and asked
if I'd got it yet, said he sent it days ago) I'd be lost without it, at
least for the best part of a day until I set up something else. I have a
spare Cyrix PR333 with XP Pro installed that I keep all set up in the
cupboard for just such an emergency. Not a great machine by any means but
would do at a pinch for a gateway PC.

Make that a Cyrix MII333, just reminded from a post above.

 

[David Maynard]

your good



I reckon. I didn't think it was possible to put pins back!

I wouldn't have bet on it either. Good work ya done there.

on a



Ug! Chaintec? Not my favourite mobo. Had one once, got rid of it pretty
quickly.

I understand. It isn't in my 'top 10,000' list either.

It is, however, working and was 15 bucks.

but I



Can be a bitch huh? My network all hangs off a Celly 400 (and another
MS-6163 Pro) and I don't overclock it and try to keep it simple. I'm
probably gonna put a PIII katmai in it when my Celly 900 arrives (should
have been here by now, the guy who sent it to me just emailed me and asked
if I'd got it yet, said he sent it days ago) I'd be lost without it, at
least for the best part of a day until I set up something else. I have a
spare Cyrix PR333 with XP Pro installed that I keep all set up in the
cupboard for just such an emergency. Not a great machine by any means but
would do at a pinch for a gateway PC.

Oh, well, the BH6 wasn't my gateway, nor the domain controller, or any of that
part. It was my old workhorse machine. I just meant *it's* networking got
scrambled. Problem being that's where all my current email and newsgroups setup
was/is, plus a ton of other 'working' data.

I suspect it was due to how the hardware is 'identified', I.E. the network card
was no longer on 'that' PCI bus anymore, although I woulda thought the 'repair'
would have sorted it out.

I first brought a backup machine in sync with email, and some other critical
data, so I wasn't 'out of business'. I was just arms and elbows involved with
getting things back to the configuration I like.

I know what you mean about the gateway. My BP6 runs that (plus some other
things) normally but I have the wireless router setup and ready to takeover with
just a routing activation (special ports already in, etc) if it's needed.

got a



I have mine mounted on a pull-out mobo tray on the wall, no case at all.
Hopefully that gets around any potential heat problems. Good luck with
replacing the FET.

Thanks. Yes, the location I had it in was about as bad as it gets, for the mobo.
It's convenient for me, though. hehe