Made the switch... A7V8X-X to A7N8X... Now my graphics card fan is annoying.

[hupjack]

I've now switched from the A7V8X-X that I accidentally purchased to a A7N8X
that I had intended to get. The refurb nvidia board from newegg came *with*
the I/O shield and all the extras. Certainly an unexpected plus from what
I was told to expect.

On the A7V8X-X the PSU fan speed was reporting low.. On the A7N8X, it's not
reporting anything at all for the PSU fan speed, not in the BIOS Hardware
Monitor, and not with Motherboard Monitor.. just straight up zero. But I
guess I'll just let that slide since I don't seem to be alone in this.

So with my new retail 2500+ in this new board, the AMD supplied fan is WAY
quieter than that low end hunk of junk that was stuck on top of my OEM
2600+. Not only was that speeze setup loud, but the chip wasn't even kept
cool.. 56C idle with the case off. 60-62C idle with the case on.

With the new.. new setup, I'm getting 29C case, 42C diode, and 35C socket,
with the case off, as reported by motherboard monitor. I had no idea there
were different socket and diode temperatures. Apparently Motherboard
Monitor with the ini files I downloaded for the A7N8X detects an additional
temperature that isn't reported even in the BIOS hardware monitor?

The system was definitely hotter with the case on.. 5C + hotter at idle I
think.. So should I stick a case fan in this thing?
Now that my processor fan is sufficiently quiet, the loudest annoying thing
is the fan on my V7100 GeForce2 MX 400 graphics card. It must have a
dynamically controlled fan speed because it hits certain speeds where it
creates a fair amount of buzzing..

I was all set to keep this graphics card, but its noisiness might be enough
to make me switch it, or at least look at switching its fan. Funny I never
heard it before. I guess my PSU fan and processor fan were always drowning
it out or at least contributing so that the humming symphony couldn't be
directly attributed to the graphics card.. Any suggestion for quieting down
the graphics card fan buzzing / vibration?

-Ethan

 

[Wooducoodu]

replace it with a nvidia 5600xt or ati 9600 (not the pro or xt) neither of
them have fans. if you want a better card sapphire has some better ati based
cards they put giant heatpipes on or gainward is supposed to be doing the
same with a 5700 soon.

 

[kony]

I've now switched from the A7V8X-X that I accidentally purchased to a A7N8X
that I had intended to get. The refurb nvidia board from newegg came *with*
the I/O shield and all the extras. Certainly an unexpected plus from what
I was told to expect.

On the A7V8X-X the PSU fan speed was reporting low.. On the A7N8X, it's not
reporting anything at all for the PSU fan speed, not in the BIOS Hardware
Monitor, and not with Motherboard Monitor.. just straight up zero. But I
guess I'll just let that slide since I don't seem to be alone in this.

So with my new retail 2500+ in this new board, the AMD supplied fan is WAY
quieter than that low end hunk of junk that was stuck on top of my OEM
2600+. Not only was that speeze setup loud, but the chip wasn't even kept
cool.. 56C idle with the case off. 60-62C idle with the case on.

With the new.. new setup, I'm getting 29C case, 42C diode, and 35C socket,
with the case off, as reported by motherboard monitor. I had no idea there
were different socket and diode temperatures. Apparently Motherboard
Monitor with the ini files I downloaded for the A7N8X detects an additional
temperature that isn't reported even in the BIOS hardware monitor?

The bios should report the diode temp when using a Palomino or newer CPU
(that is, any CPU that has it's own integral temp sensor). On past Asus
boards there was either a jumper or at least a spot for one on the board
(that went unused) which would give the option of which temp was reported.
At some point I think they hard wired it with the assumption that anyone
would be running the newer CPUs with the board.

The system was definitely hotter with the case on.. 5C + hotter at idle I
think.. So should I stick a case fan in this thing?

Forget about idle temp. The system needs to be adequately cooled to
remain stable while running for extended period of time at full load.
Popular programs like CPUBurn or Prime95 can be used to attain a load
temp, after 30+ minutes of running. THEN note the temp, and whether it's
stable, particularly in Prime95 after running for at least several hours.
Also check the temp of other components like the hard drive, video card.
They may need the additional case airflow more than actively, well cooled
parts like a CPU. Sometimes a fan isn't even needed, merely increasing
the front bottom air intake vents (and the metal behind them if necessary)
is enough. Take care not to divert too much airflow away from the hard
drive bay, usually a decent case is designed to have a large portion of
the air flowing past the drive(s).

Now that my processor fan is sufficiently quiet, the loudest annoying thing
is the fan on my V7100 GeForce2 MX 400 graphics card. It must have a
dynamically controlled fan speed because it hits certain speeds where it
creates a fair amount of buzzing..

It's probably wearing out and needs lubed. Such a thin fan with a tiny
bearing needs a relatively thick oil, better a very thin grease, else
it'll just run out and need lubed again sooner. If it's overly worn from
extended periods of running "dry" the bearing may be too far gone to last
much longer regardless of lubing it... generally the best way to handle
those crap video card fans is to lube them before they ever start making
the excess noise, else replace the whole heatsink/fan assembly instead.

I was all set to keep this graphics card, but its noisiness might be enough
to make me switch it, or at least look at switching its fan. Funny I never
heard it before. I guess my PSU fan and processor fan were always drowning
it out or at least contributing so that the humming symphony couldn't be
directly attributed to the graphics card.. Any suggestion for quieting down
the graphics card fan buzzing / vibration?

There are lots of ways. It's only a GF2MX400, you could probably find a
large passive heatsink and do without the fan entirely, in which case it'd
help to leave the PCI slot under the AGP empty, and keep that PCI slot's
case bracket cover off, which will allow a stream of airflow past the
card.

There are also several aftermarket video card coolers available, but
frankly I wouldn't spend a lot on it for that video card. If you can
leave the PCI slot emtpy you could instead use epoxy to attach an old
socket 7 CPU heatsink or cut down whatever 'sink you have handy. IIRC the
fan on those are usually 40mm, you could buy just a fan but typically the
fan sits recessed in the metal portion of the 'sink instead of having
it's own framework, so a standard replacement fan would have to sit atop
the metal instead of recessed inside it, so you'd need longer screws too,
and either splicing the old fan's power plug onto the new fan (if you
can't find a fan that comes with a suitable plug) or possibly slipping the
fan's metal contacts out of it's plug and inerting them into the plastic
portion of the old fan's plug... depends on the size and design of the
plugs and contacts whether that's possible.

 

[~misfit~]

replace it with a nvidia 5600xt or ati 9600 (not the pro or xt)
neither of them have fans.

That depends on who manufactured the card. It has nothing to do with the
model. Some manufacturers put fans on the 5600XT.

 

[Wooducoodu]

it has something to do with the model. the nvidia 5600xt and ati 9600 don't
need fans but a few manufacturers put them on anyway. other models require a
fan but for a couple of those sapphire and now gainward put giant heatpipes
on them.

 

[Alan Walker]

it has something to do with the model. the nvidia 5600xt and ati 9600
don't need fans but a few manufacturers put them on anyway. other
models require a fan but for a couple of those sapphire and now
gainward put giant heatpipes on them.

I put an Arctic Cooling VGA Silencer on my 9700 Pro and it's significantly
quieter for only £6, see what they do for your card.

Alan

 

[hupjack]

Kony.. YOU, my friend, provide answers and clarity in a sea of confusion.
Thank you so much for sharing your wisdom. You ROCK!

Forget about idle temp. The system needs to be adequately cooled to
remain stable while running for extended period of time at full load.
Popular programs like CPUBurn or Prime95 can be used to attain a load
temp, after 30+ minutes of running. THEN note the temp, and whether it's
stable, particularly in Prime95 after running for at least several hours.
Also check the temp of other components like the hard drive, video card.
They may need the additional case airflow more than actively, well cooled
parts like a CPU. Sometimes a fan isn't even needed, merely increasing
the front bottom air intake vents (and the metal behind them if necessary)
is enough. Take care not to divert too much airflow away from the hard
drive bay, usually a decent case is designed to have a large portion of
the air flowing past the drive(s).

With ASUS's Smart Doctor Software I get to detect the GPU temperature, so
that's handy for doing the load test. The monitoring software also has a
field for displaying graphics card RAM temp, but it's blank, so I assume the
feature isn't available for this card. But how does one gauge hard drive
temp. The drive felt like it was getting pretty toasty after just running
close to idle for a while. It's a Barracuda ATA IV - ST380021A. I'm
assuming there is no thermal sensing software for the HD.

It's probably wearing out and needs lubed. Such a thin fan with a tiny
bearing needs a relatively thick oil, better a very thin grease, else
it'll just run out and need lubed again sooner. If it's overly worn from
extended periods of running "dry" the bearing may be too far gone to last
much longer regardless of lubing it... generally the best way to handle
those crap video card fans is to lube them before they ever start making
the excess noise, else replace the whole heatsink/fan assembly instead.

I took a peak on USENET looking for more info on how (and with what product)
to lube these fans. Do you agree with the 3rd post at this thread
http://tinyurl.com/2dwek? regarding the technique and product
"non-conductive silicon based lubricant"?

Thanks for the other suggestions. I'd like to try giving it a shot of lube
first. I don't think it's too far gone. Hopefully a little lube will do
the trick.

 

[kony]

Kony.. YOU, my friend, provide answers and clarity in a sea of confusion.
Thank you so much for sharing your wisdom. You ROCK!


With ASUS's Smart Doctor Software I get to detect the GPU temperature, so
that's handy for doing the load test. The monitoring software also has a
field for displaying graphics card RAM temp, but it's blank, so I assume the
feature isn't available for this card. But how does one gauge hard drive
temp. The drive felt like it was getting pretty toasty after just running
close to idle for a while. It's a Barracuda ATA IV - ST380021A. I'm
assuming there is no thermal sensing software for the HD.

"Pretty toasty" sounds too hot. Then again, some people are willing to
make compromises, have a hot running system to avoid the work of
increasing airflow or cost of different case or fans. That's a personal
choice, but in such a situation it would be wise to have everything backed
up (which is prudent regardless) so if the drive fails, it's just a matter
of replacing it, not the data or time to reconfig everything. I'd see
about increasing the airflow past the drive though, not worrying about
getting a temp readout on it but using the touch-test is enough. In a
well ventilated case the drives won't feel more than mildly warm, if that.

I took a peak on USENET looking for more info on how (and with what product)
to lube these fans. Do you agree with the 3rd post at this thread
http://tinyurl.com/2dwek? regarding the technique and product
"non-conductive silicon based lubricant"?

No, not silicone. It should be compatible with the current lube, what
residue is left in it. A common lubricating oil is ok for a fan bearing
that's not too worn but the smaller and more worn it is, and also because
it's horizontally mounted, the thicker it should be. By the time a video
card fan is making noise it's ideally lubed with a very light grease.
Silicone is inferior in any application where the parts aren't vulnerable
to petroluem based products, and also is typically just a coating, not
thick enough to fill the voids now present from the bearing degradation,
the fan would still wobble and wear.

However, if you're continuing to use the system, with the fan still
running dry and further wearing out every moment it's running "dry", you
should add 1/2 drop of any oil you have available asap, even if it isn't
very thick, to temporarily stop the wear until the proper lube can be
used, with an exception - Note that "some" video card fans do have a dual
bearing, one sleeve and one ball, and that type should not be lubed with
oil, may get VERY loud if the oil gets into the ball bearing (which
gravity will tend to do, the oil will run down into the ball bearing).
So, if it's an all-sleeve bearing, any oil is a temporary fix. If it has
a ball-bearing, or for longer lasting fix, it needs be very thin grease.
Unfortunately an [all-sleeve] or [one sleeve and one ball-bearing] fan
bearing can look the same from the back, with the sticker removed, because
when there are both types of bearings, the first, visible bearing is the
sleeve bearing.

If you need to rely on this system, video card, you might just lube it as
a temporary measure and begin looking for a replacement fan or fan-sink
combination, though some 'sinks are epoxied onto the GPU, are possible to
pry off but that must be done carefully, is risky.

 

[hupjack]

So I dripped a couple drops of oil into the graphics card fan and it appears
to have quieted down. I guess I'll call it fixed until I hear the offensive
noise again.

I ran the prime95 torture test for 30 mins on my system.
The temperature results from motherboard monitor
socket: High-52 AVG-50 Low-39
Diode: High-65 AVG-62 Low-45
Case: High-31 AVG-30 Low-30

Are those temps ok or too high?
Prime 95 doesn't get your HD working. I suppose a different test should be
used to gauge it's thermal reaction to heavy load?

-Ethan

 

[kony]

So I dripped a couple drops of oil into the graphics card fan and it appears
to have quieted down. I guess I'll call it fixed until I hear the offensive
noise again.

I ran the prime95 torture test for 30 mins on my system.

That should be enough to generate the heat, but to check CPU stability
it's optimal to test for at least several hours, and a specific Prime95
test is much more likely to quickly reveal problems. It's the way I
usually start out narrowing down a voltage/speed curve when evaluating
overclockability of CPU.

Prime95
Options | Torture Test | In-place large FFTs

That setting may find problems in 1/10th the time of the others, and for
most problems it'll be apparent within 1-10 minutes, but final
determination still requires hours of running.

The temperature results from motherboard monitor
socket: High-52 AVG-50 Low-39
Diode: High-65 AVG-62 Low-45
Case: High-31 AVG-30 Low-30

Are those temps ok or too high?

Socket temp can be ignored.

Case temp is actually a chip, which creates it's own heat (case ambient
air temp is actually lower than 30C _IF_ that reading is at least accurate
for the chip temp).

Diode temp is a bit on the high side. You might do a bit of research on
your board to see if it (and the particular bios revision you're running)
is accurate or off by a bit. It's more significant to know that with your
temp being reported at 65C, as that's around the upper limit for
stability... won't damage the cpu but with some variations like a change
in ambient room temp that CPU temp may go even higher. Ideally the temp
would be under 50C, practically up to 60C is usually stable at stock
speed, sometimes even higher but there may be other factors that come into
play with reduced thermal margins, like minor power fluctuations.

Prime 95 doesn't get your HD working. I suppose a different test should be
used to gauge it's thermal reaction to heavy load?

Higher heat is a progressive problem for drives, they don't generally just
"crash" and then you'd know to get them cooler... once they've crashed
it's usually the end-of-life for a drive. So, there isn't really a stress
test you can perform to gauge their heat tolerance, rather a simple
touch-test is usually enough, because body temp and sensatively are
roughly coinciding with the temp range of hard drives... if it feels too
hot it probably is.