VA-503+ roundup questions and some New questions

Hi!

First of all forgive me my lawzy english

I have some questions about the VA-503+. I've Googled the INET and the
NewsGroups and didn't find an accurate answer, so I hope some of you
maybe may had had the problems that I'll further describe.

-------------------------------------------------------------------------------------------------------------------------------------
I've the folowing setup:
-------------------------------------------------------------------------------------------------------------------------------------

* Mobo : FIC® VA503+ v1.2a with 1MB L2 cache (2 x 512Kb chips)

* BIOS : JE4333 BIOS

* CPU : AMD® K6-2 400 (CXT core)

* Memory : 1 x 256MB Infineon® PC133 CL3-3-3 @ PC100 CL 2-2-2

* VGA : ELSA® GF2MX AGP on AGP

* LAN : Realtek PCI 10/100 LAN card (RTL8139) on PCI #2

* Sound : SB128PCI (model CT4810) on PCI#3

* HDD : WD1200JB (yes, that's an abuse -- 120GB / 8MB / 7200rpm)

* Optical : LG52x32x52

* FDD : 1x Samsung 1.44Mb

*PSU : ATX 300W (fan silent modded by termistor)

*removable USB devices: #1 : DSL modem (500mA) ; #2 : Pen Disk TwinMOS
512Mb (100mA)


It's my old "silent PC": WD1200JB is nearly inaudible; 300W generic no
brand ATX PSU is "fan quiet modded by a termistor", very quiet; and CPU
Fan...well it's the "noisiest" component (CoolerMaster - ball
bearing) but I have an quieter replacement at another location


-------------------------------------------------------------------------------------------------------------------------------------
I've the following software:
-------------------------------------------------------------------------------------------------------------------------------------

I'm running XP Pro SP2 and W98SE (dual boot)

* On W98SE, I've installed
----------------------------------------
- VIA 4-in-1 v4.43 ( AGP + IRQ Routing + INF driver);
- DirectX 9.0c (way faster than the version 7a that came with NFSPU
game)
- Nvidia Detonator 45.23 for W9x (seems to be the fastest drivers I've
seen so far)
- Creative SB128PCI (model CT481x) drivers
- USB Filter Patch v1.10 (to fix VIA USB problems)
- PCI ZeroLatencyPatch (to fix "clicking" sound problems)
- Nero 6 with ImageDrive
- NFS PorscheUnleashed game It's a bit heavy, but I can manage to
play it at 30-40FPS, with another opponent, on another PC

* On XP, I've installed
-------------------------------
(already installed : VIA 4-in-1 on XP SP2)
(already installed : DX9.0c on my distribution of XP SP2)
- Nvidia Detonator 45.23 for XP (seems to be the fastest drivers I've
seen so far)
- Creative SB128PCI (model CT481x) drivers
- PCI ZeroLatencyPatch (to fix "clicking" sound problems)
- Nero 6 with ImageDrive
- NFS PorscheUnleashed game It's a bit heavy, but I can manage to
play it at 20-30FPS, with another opponent, on another PC
- NFSPU patch for XP

-------------------------------------------------------------------------------------------------------------------------------------
Problem 1 (latest Nvidia drivers on XP)
-------------------------------------------------------------------------------------------------------------------------------------
Question 1
-------------------
After I install the latest Nvidia drivers (tried v81.98 and v71.22),
when I try to run NFSPU the computer halts with a BlueScreenOfDeath
(BSOD). If I install the v43.45 or the v45.23 that won't happen.

Any idea why this happens with the latest drivers?
How to correct this problem?
What Nvidia drivers are you using?

-------------------------------------------------------------------------------------------------------------------------------------
Problem 2.1 (USB Filter v1.10 on XP)
-------------------------------------------------------------------------------------------------------------------------------------
Question 2.1
-------------------
USB Filter v1.10 states on its README and on VIAArena that it's for
using on W98, ME, 2K and XP, but I can't manage to properly install it
on XP SP2! It stops the installation with the "Cannot update original
INF" message!

Maybe SP2 includes this FilterPatch and prevents it from being
"re-installed"?
Maybe VIAArena meant XP without no SP1 nor SP2?
* Anyway, the USB seems to work fine, if I connect ONLY one device at
the time

-------------------------------------------------
Problem 2.2 (USB devices on XP)
-------------------------------------------------
Question 2.2
-------------------
Can you connect 2 USB devices SIMULTANELY draining about 600mA on the 2
USB ports and still manage them to work? In W98SE I can do it, but on
XP I only can do that in a certain order, e.g., 1st the 100mA part,
then the 500mA part. On the reverse order (1st 500mA part then 100mA
part), the 1st device --a DSL USB modem-, will stop working! I guess
that XP when I 1st connect the 500mA part "recognizes" that there is no
more power on the 500mA hub and "prevents" and if any more devices will
be connected it will shutdown the 1st device? On W98SE that does not
happens, so i'll assume that XP is "smarter" and prevents "overcurrent"
of the USB hub? Or might be because I can't install USB Filter Patch on
XP? (Question 2.1)



-------------------------------------------------------------------------------------------------------------------------------------
Problem 3 (Overclocking)
-------------------------------------------------------------------------------------------------------------------------------------
I've an AMD K6-2 400 CPU (CXT core, that I suppose its very good
overclocker). I've got it @ 4x100=400MHz/2.2V.


Tried 4 x 112MHz = 448MHz @ 2,4V (I don't want to go above +10% default
Vcore).
@ 4 x 112MHz = 448MHz @ 2,4V the computer only starts reliable if I
disable the onboard L2 cache, so I gave up about the 112MHz setting

But I'm still thinking on a way to circunvent this, because:
- higher FSB = better perfomance,
- because the PCI cards seem to be OK at 1.12 x 33MHz = 37MHz (no
sound distortion),
- AGP card seems OK also at 74MHz (no visual artifacts)
, and the HDD seems also OK (no data corruption so far)

Question 3.1
-------------------
Is there any way to increment the L2 cache voltage? Any resistor mod
trick or something?Anyone knows what voltage feeds those L2 cache
chips? 3.3V? 5V?

Tried 4.5 x 100MHz = 450MHz @ 2.4V but BSOD sometimes on XP, so I guess
I'm stucked with 400MHz with stabilty and 450MHz with minor instability
(still I would prefer the 4 x112MHz setting)

-------------------------------------------------------------------------------------------------------------------------------------
Problem 4 (BIOS ...)
-------------------------------------------------------------------------------------------------------------------------------------
Question 4.1 ...and Standby
----------------------------------------------
On W98SE if I enter the mobo into "Standby mode" (via Shutdown menu >
Standby), it won't resume to desktop if I press the keyboard (event#1)
neither if I move the mouse(event#2), but on XP any of those events (#1
or #2) will resume the VA-503+ from Standby.

On W98SE the only way to "resume" from Standy is by pressing the
Soft_Off_(ATX PSU)_Power_Button, but althought I select the "Delay4Sec"
on BIOS this procedure "resume" the mobo to shutdown it

Did you notice this "strange" behavior on your motherboard, on any of
your current OS's?


Question 4.2 ...and RTC alarm resume
-------------------------------------------------------------
I can't manage to get the "RTC alarm resume" to work! Did you tried it
out? It won't Power Up the motherboard, althought I'm using a brand new
300W ATX PSU. I tried with the accurate date and time I would like the
mobo to Power Up, several tries, but it seems to be a "no go" for this
item.

Question 4.3
-------------------
Because of these problems I remembered why I want to reflash the JE4333
BIOS I want to see if on early BIOS versions the following items will
function correctly (or on another words, these are my "minor"
problems).
I'm looking for tecniques to "unsolder/desolder" (?) VA503+ BIOS chip
(don't know the exact word). That way, if I can unsolder the BIOS chip
I will put a socket to socket (duh) the BIOS chip. Then, case I flash
the BIOS wrong, I'll take BIOS chip off and replace it with a "good"
BIOS chip. But this opperation might be risky and maybe I can damage
the motherboar and the BIOS chip itself, because of the heat needed to
be applied to unsolder... Anyone tried this before?

Well that ALL folks! If you, by now, are not sleeping (because of this
long post)...well I hope to hear from you soon!

Kind regards,
Mikeluz

P.S: I understand your "nostalgia thing" about this board.. I'm
nostalgic also: 386SX16@20; P60@66MHz;P100@133, etc, etc

 

Mikeluz,

Good to see another 503+ enthusiast. The one question I will try
to postulate an answer to would be question 1, the nVidia driver. I
would expect the newest drivers to not support such an old chipset. I
have not used the reference drivers with anything older like the 503+,
and the GF 2 is not the issue. For example, I installed a V770 TNT2 in a
BX motherboard the other day, and I just used the old Detonator drivers.
Works fine. Even if the newer graphics drivers worked okay I seriously
doubt they are still tweeking a GPU like the GF 2 anymore, so there
would no real benefit. Esp if you are not running any 3D apps.

Okay, now it's time for someone to reply to one of the other
questions. BTW I liked the very last line, and I thought I ran some old
stuff. Tell you the truth, I much prefer older hardware simply because
it still works great. There are other reasons, but that is the best.

John

 

You might want to re-consider your overclocking strategy. Overclocking the
FSB will also overclock the PCI bus which can lead to hard drive corruption.
Considering the VA503+ only supports UDMA mode 2, you're hard drive is going
to be running really slowly.

Personally, I've decided that once I upgrade to an Athlon 64 X2 based
system, my 503+ is headed for the scrapheap.

--
Doug
Hi!


* HDD : WD1200JB (yes, that's an abuse -- 120GB / 8MB / 7200rpm)

-------------------------------------------------------------------------------------------------------------------------------------
Problem 3 (Overclocking)
-------------------------------------------------------------------------------------------------------------------------------------
I've an AMD K6-2 400 CPU (CXT core, that I suppose its very good
overclocker). I've got it @ 4x100=400MHz/2.2V.


Tried 4 x 112MHz = 448MHz @ 2,4V (I don't want to go above +10% default
Vcore).
@ 4 x 112MHz = 448MHz @ 2,4V the computer only starts reliable if I
disable the onboard L2 cache, so I gave up about the 112MHz setting

But I'm still thinking on a way to circunvent this, because:
- higher FSB = better perfomance,
- because the PCI cards seem to be OK at 1.12 x 33MHz = 37MHz (no
sound distortion),
- AGP card seems OK also at 74MHz (no visual artifacts)
, and the HDD seems also OK (no data corruption so far)

Question 3.1
-------------------
Is there any way to increment the L2 cache voltage? Any resistor mod
trick or something?Anyone knows what voltage feeds those L2 cache
chips? 3.3V? 5V?

Tried 4.5 x 100MHz = 450MHz @ 2.4V but BSOD sometimes on XP, so I guess
I'm stucked with 400MHz with stabilty and 450MHz with minor instability
(still I would prefer the 4 x112MHz setting)

-------------------------------------------------------------------------------------------------------------------------------------
Problem 4 (BIOS ...)
-------------------------------------------------------------------------------------------------------------------------------------
Question 4.1 ...and Standby
----------------------------------------------
On W98SE if I enter the mobo into "Standby mode" (via Shutdown menu >
Standby), it won't resume to desktop if I press the keyboard (event#1)
neither if I move the mouse(event#2), but on XP any of those events (#1
or #2) will resume the VA-503+ from Standby.

On W98SE the only way to "resume" from Standy is by pressing the
Soft_Off_(ATX PSU)_Power_Button, but althought I select the "Delay4Sec"
on BIOS this procedure "resume" the mobo to shutdown it

Did you notice this "strange" behavior on your motherboard, on any of
your current OS's?


Question 4.2 ...and RTC alarm resume
-------------------------------------------------------------
I can't manage to get the "RTC alarm resume" to work! Did you tried it
out? It won't Power Up the motherboard, althought I'm using a brand new
300W ATX PSU. I tried with the accurate date and time I would like the
mobo to Power Up, several tries, but it seems to be a "no go" for this
item.

Question 4.3
-------------------
Because of these problems I remembered why I want to reflash the JE4333
BIOS I want to see if on early BIOS versions the following items will
function correctly (or on another words, these are my "minor"
problems).
I'm looking for tecniques to "unsolder/desolder" (?) VA503+ BIOS chip
(don't know the exact word). That way, if I can unsolder the BIOS chip
I will put a socket to socket (duh) the BIOS chip. Then, case I flash
the BIOS wrong, I'll take BIOS chip off and replace it with a "good"
BIOS chip. But this opperation might be risky and maybe I can damage
the motherboar and the BIOS chip itself, because of the heat needed to
be applied to unsolder... Anyone tried this before?

Well that ALL folks! If you, by now, are not sleeping (because of this
long post)...well I hope to hear from you soon!

Kind regards,
Mikeluz

P.S: I understand your "nostalgia thing" about this board.. I'm
nostalgic also: 386SX16@20; P60@66MHz;P100@133, etc, etc

 

sell it, somebody can still use it. personally I have no use for
newer computer hardware, with few exceptions they are really just
multimedia toys.

the other thing he could do to improve his hard drive UDMA is use a
controller card, Promise makes several good ones. He could also upgrade
to a K6 3+ since he has the correct BIOS. I have a similar configuration
and it runs very well.

 

Pigdos:
-----------

I'm using 503+ at 100MHz...I'll leave that way... When oveclocking I
can't manage to use the 112 MHz FSB, so I'll leave it at 100MHz FSB...
By the way, when going from 4x100MHz=400MHz to 4.5x100MHz=450MHz, that
12% increment in CPU speed only increase about 1FPS/30FPS on NFSPU
(i.e. an 2-4% increase on FPS), so I'll leave the 400MHz setting --it's
more stable than 450MHz. That little increase on FPS would be
expectable since it's better to increase FSB rather than multiplier as
you probably know


I want to use the FIC VA503+ to play old (network) games.
Since I have coleagues at home occasionaly, I play with them, 1/2 hour
or so. It's only just for play from time to time, so I do not need to
purchase a dedicated (more expensive) solution
I've played on it NFSPU, althought on the low quality @ 640x480 on a
15'' CRT at about 30-80 FPS. Could have better quality, but FPS rate
decrease...to play ocasionally it's enough...

By the way i live on Portugal, and money it's not abundant here, so I
need to keep the budget tight I've an main system, but it's giving
me troubles so I need to backup all files and to work on the 503+ if
necessary. Also it's nice to play with coleagues

Farmuse:
--------------
This computer was offered to me by someone, so I don't intend to spend
money on it Altought you were right about the Promise controler. It's
a waste to have the WD1200JB on the 503+ *per se*, but I don't have any
smaller disk (nor quieter) to put on the 503+ *quiet* system, so I'll
leave it there.

I'm trying to squeeze the most from the computer spending the minimum
money

Unfortunelly there is not much more to squeeze... LOL

Do you know any more tweakings I could use to improve performance?

I noticed as increase of about 10FPS from the 20FPS NFSPU (NFS5) *on
XP* to the 30FPS *on W98SE*. Unfortunely I can't manage to play NFSHS
(NFS4) any faster than NFS5 on this system. Unless there are any more
tweak I can't use.

Kind regards
Mikeluz

 

Mikeluz,

Hmmm, really surprised to hear your game runs faster under XP.
The one thing you still have not mentioned is overclocking the GF 2
card. You could use powerstrip, and I am sure you can squeeze more fps
from that card. I believe the MX200 is 64 bit bus, but not too bad a
card really. Try overclocking the GPU first, then the memory. Of course
you must use active cooling, and a case with good cooling as well. Lower
room temps do help a lot. Other than that, the only other choice is to
rip the top off the K6 2, and mount a lapped heatsink directly to the
core. I bet you could get to 450 then. But is it worth it ? Or you could
just bump the voltage up to 2.4 and see if the 400 can overclock well.

 

Farmuse

I suppose I don't explained myself well : NFSPU runs faster on W98SE
(30FPS) than on XP (20PFS)

With an GF4 Ti4200 AGP card the game (NFSPU) won't give me any more
FPS, it's CPU bottlenecked -- it's not entirely true, but I'll explain
below. I suppose that "car simulation/arcade games" are very CPU
demanding rather than GPU demanding, in general. Shoot'em up games from
the NFSPU era I suppose were less CPU demanding games...

I tell this because with a GF2MX, with game effects on medium, the FPS
rate don't change when I pass the screen resolution from 640x480 to
800x600, so it's the CPU the limiting factor at those resolutions --
this because game effects are CPU demanding;
At 1024x768 the FPS rates goes down -1~ -2FPS, so the GPU starts
becoming ALSO a limiting factor, and overclocking will do it the trick,
but on a slower system like this one, it won't give me real benefits,
because the CPU is the major bottleneck

I bumped the voltage to 2.4V @ 450MHz (I tried also @ 2,5V for
increased stability) but I found out, that for the games that I play
the most (NFS3, 4, 5) --on this system, well understood--, the FPS
increase of about 1FPS on every 25PFS won't make a real difference over
the * base FPS rate @ 400MHz* of 30FPS (i.e. an effective increase of 1
~2FPS), so I give up on the idea of overclocking by bumping the
multiplier from 4x to 4.5x... Case the L2 cache was up to the FSB
112MHz that would make a noticeable difference (about 10% increase)
over the mere 3% increase on FPS gived by bumping only the multiplier.

If anyone knows how to tweak L2 cache voltage or decrease L2 cache
latency, tell us! Maybe that will do the trick to use the 112MHz FSB.
Until then, I'll use the default setting (4x 100MHz @ 2.2V)

Kind regards,
Mikeluz

 

Question 2: (USB)
-------------------------------------------------------------------------
The 1.10 USB filter patch at VIA is meant to fix certain specific
issues. To be honest, there was so much confusion about the patch, that
VIA got sick of answering the questions. Instead, they made the patch
available to all newer versions of Windows, but it only installs if it
is really needed. That way, VIA can say, "Here, run this program. It
will detect your operating system and install only what is needed. Now
leave us alone."

http://www.viaarena.com/default.aspx?PageID=420&OSID=1&CatID=1200&SubCatID=121

A USB hub is only supposed to support 500mA total, maximum. As you
correctly guessed, Windows XP is "smarter" about this.


Question 4: (standby)
-------------------------------------------------------------------------
Since standby works with Windows XP but not Windows 98SE, the problem is
with the drivers in Windows 98SE (or maybe Windows 98SE itself), not
with the motherboard. Unfortunately, there is not really much you can
do about this, except hope for new drivers.

One question, though... is that 100mA USB device a USB mouse? If so,
that could be part of the problem, if you are overloading the USB hub
with 600mA. Remember, Windows XP is smarter about USB power than Win98SE.

I am not sure what to say about the RTC alarm.

Is the VA-503+ BIOS chip really soldered to the motherboard? I did not
know that. I would have thought it was one of those long computer chips
with two rows of teeth (DIP/DIL chip). DIP chips you can extract using
a tool that looks like a gigantic pair of tweezers. If the BIOS chip
really is soldered, then yes, it is possible to desolder/resolder the
chip, but you need to know what you are doing.

Here are a couple of links to places that specialize in BIOS chips and
replacements. Even if you do not use their services, you can at least
see what is available:

http://www.badbiosflash.com
http://www.flashbios.org

--Alex

 

Thank you Alex for your answer.

Well I guess that's why USB Filter Patch v1.10 won't install
(correctly) on XP, althought it says on its "Readme.txt" that will
install on XP (maybe on XP first release before SP1 or SP2).

I was told by coleagues that USB1.1 specification is 2ports/500mA
limited and USB2.0 has no limitation on current (I've to investigate
this latest info on www.usb.org). That's why I made the above guess
about W98SE vs. XP.

The 100mA device that I connect in 2nd place is a USB drive ("pen
disk").

The 100mA device that I connect in 2nd place is a USB drive ("pen
disk"). I understand your põint here but I tested Standby
functionality without any USB devices attached...new drivers for
W98SE... I guess MS won't bother...neither do I...I disabled Standby on
W98SE, i.e., it's on the "Never enter standby" setting

Someone told me that this RTC thing is a plague of the MPV3
implementation, that RTC never had correctly functioned on this
chipset? I don't know. however a friend of mine has an ASUS P3 mobo
that after a BIOS update won't boot from RTC anymore... I might try to
revert the BIOS on this motherboard to some earlier versions, to see if
this problem is still present (MPV3 bug theory) or not(BOS bug
problem). That lead us to my next question...BIOS flashing...

....BIOS flashing I done it for years, and I've learned how to recover
"dead motherboards" via another motherboard with a similar BIOS chip
type (same socket & same memory size wise). But on this case, in fact,
all VA503+ BIOS are not socketed to motherboard, so in case of a bad
flash... kaput! .. well it's not entirely truth, because I think that
if I don't force the programming of the boot block, then I can use the
"Boot block feature" with an FDD drive and an ISA VGA card to see the
procedures to correctly BIOS flash again the mobo. Still, I'm more
confortable with BIOS chip swapping than with this procedure, but I
guess rhis procedure evolves less risk than "hot-swap-programming", so
I guess I'll leave the "socket VA-503+ BIOS chip project" in the
drawer

By the way, had someone tried the recovery BIOS flash procedure? Was it
sucessfull?

Thank again ALL for your feedback!
Kind regards,
Mikeluz

 

Hey, Mikeluz.

I just went to http://www.usb.org myself and downloaded the USB 2.0 spec
at http://www.usb.org/developers/docs/usb_20_02212005.zip . Looking at
Table 7-7 (DC Electrical Characteristics) on page 178, the supply
current for a high-power hub port is listed as 500 mA, minimum. This is
500 mA per port. This means that having a 100 mA device and a 500 mA
device *should* be possible when they are plugged into different ports.

However, the high-power function input (what is actually used by the USB
device) is listed as 500 mA, maximum. In other words, USB 2.0 devices
still have the same 500 mA current limit as before.

There are two basic device classes, low-power (100 mA) and high-power
(500 mA). According to section 7.2.1 on page 171...

"All devices default to low-power. The transition to high-power is under
software control. It is the responsibility of software to ensure
adequate power is available before allowing devices to consume high-power."

In other words, when you first plug in any USB device, it will act like
a low-power 100 mA device. Only after the device and the computer
"talk" to each other will it increase to 500 mA, if it is a high-power
device. It sounds like Win98 and WinXP do not "talk" in the same manner!


Moving onto the BIOS flash, I have done several of those "recovery"
flashes. Sometimes they actually do work. It depends more on why the
old BIOS went bad in the first place. If the boot block is still
working, then it will probably work. Of course, this also assumes that
other bad things have not happened. For example, if the floppy disk
drive is bad, then the recovery method will not work very well.

--Alex

 

Hi Alex!

Alex: "In other words, when you first plug in any USB device, it will
act like
a low-power 100 mA device. Only after the device and the computer
"talk" to each other will it increase to 500 mA, if it is a high-power
device. It sounds like Win98 and WinXP do not "talk" in the same
manner!"

Yup, that's look like the problem. Still, W98SE will function with 2
USB devices (1x High Power (500mA DSL modem) and 1x Low Power ("pen
disk drive") draining 600mA from the 500mA hub, strange isn't it?
I would assume that XP driver is somewhat "bad" because it shouldn't
allow any more devices to function while one 500mA device is draining
the "maximum" allowed current from the 500mA USB1.1 hub... Never heard
of such problem before, at least I'm not aware of that.

It's a "bummer" the case when I'm surfing the INET (via USB DLS modem)
and this one gets "disabled" when I connect the USB pen disk... INET
goes down, and the only way to "reconnect" the modem (and to INET) is
by reseting computer and dettaching the pen disk.

I'll google the net a little bit more to see if someone is getting the
same problem.

And yes. it's true that 2x USB devices could be attached to a USB2.0
hub. I've verified that myself

Good to know about your sucessfull BIOS recoveries. That way the
"socketed-BIOS-project" goes directly to the drawer

Thank you!

Kind regards,
Mikeluz

 

I socketed the BIOS chip on my 503+. It was fiddly, and it was necessary
to use a slim, low socket because there is very little clearance (2mm?)
for any card installed above.

Yes, and yes it was.
Has Uniflash been mentioned? I can't remember, and this thread has
mostly expired here. It's a very useful tool.
http://www.uniflash.org/

 

Hi Roger!

Thanks for replying!I

How did you unsoldered the BIOS chip from the mobo? Did you use a 40W
(or so) soldering iron? I was thinking of levering up the BIOS chip at
the same time I applied the heat to one row of pins.

I've tried this procedure first on a mobo with the BIOS chip socket
that I want. But I saw that this procedure was not to good in the first
place to harvest the socket so I gave up on that idea, because I don't
want to apply the same procedure on the BIOS chip itself and might ruin
it...

Did you follow some tutorial, or so? What technique did you use? By the
way, I'm recapping my P4 mobo and I was sucessfull so far, so I can say
that my soldering skills are good, though maybe my unsoldering
techniques are no so good.

Kind regards,
Mikeluz

 

Hey, Mikeluz.

In Windows XP, look through the Device Manager for your USB DSL modem
(or any other USB device). Somewhere in the settings, there is a little
checkbox that says something like, "Allow Windows to turn this device
off to save power." Try unchecking this box. See if it makes any
difference.

--Alex

 

You could either use solder wick, which is quite neat and tidy.
Or, it is possible to mask off the rest of the motherboard and use an
air duster aerosol to blast the solder out of the way.
This is very messy and carries a real risk of creating a short somewhere
if a bit of spatter lands in the wrong place, but it worked for me.

Have fun!

 

Ah yes, and my friends wonder why I would rather game under 98,
several reasons but it uses less system resources right ?

To be honest, I would still overclock that GF2, you might be
surprised. You won't really know until you try. If still no happiness
then save up for a K6 2 550 or a K 6 III plus and overclock that, they
can often reach 600 MHz. Then you would be happy.

Cheers from America, John