🚀 go-pugleaf

On Fri, 21 Sep 2001 23:21:57 GMT, "Karl Dieter" <nospam@mydomain.com>
wrote:

>I'd appreciate other people's thoughts on this as I can't seem to locate any
>web pages on the subject...
>
>I have built a PC using the following:
>
>1) Intel D850GB
>2) 2 x Nidec 120x120x25mm fans
>3) ...and other things
>
>The problem (if it is a problem), is that the rear fan spins fine, to pull
>out the hot air from inside the case.  The fan at the front never spins to
>draw in cool air.  Now I know why, it's because the sensor for the front fan
>is so low down in the case, it doesn't detect any heat and doesn't spin.
>Where-as, the rear fan has it's sensor at the top of the motherboard, where
>the processor & RAM are, and there-fore spins all the time.
>
>Now is it normal to have this setup?  I basically have the rear fan spinning
>all the time, and the front fan never spinning, as it's sensor doesn't
>detect any heat.  I would have thought to have a constant in-flow of cool
>air would be a good thing.

Sure it would. Of course, you actually *do* have an in-flow of cool
air at the front, but it's all being *driven* from the one fan at the
back. Having some extra push on the way in would help keep the density
of air in the PC up (allowing it to absorb more heat, i.e. provide
more cooling, while it's there), and speed up the airflow ('cos the
rear fan won't be fighting such a big pressure differential). But for
this, you need either a "dumb" fan with no temperature sensor
controlling it, or one that actually senses the temperatures of the
sensitive components in the system, and turns on the extra cooling
when it really is required.

I have to say, though, just one of those big fans is probably plenty,
unless you have a seriously hot/over-crowded system. And don't forget
the noise an extra fan makes... I'd be inclined to go for a more
modest (i.e., slower, smaller and quieter) "dumb" fan at the front, or
just make do without.

There are some really good low-noise models at www.quietpc.com .


>Anyway, thoughts and comments appreciated.  The system works perfectly, I'm
>just wondering why I have a front case fan that appears to be never used.
>
>Regards.
>
Is the sensor part of the fan, or on the MOBO, or a seperate plug-in,
or what? Can you override it, e.g. by plugging in the fan differently,
or by a BIOS/Jumper setting?


	Paul B.  =:o}

(Discard unreality before replying...)

The sensor is on the MOBO.  Basically, the fan sends it's RPM to the MOBO,
then to the SMSC controller.  The sensor is placed near the fan plug header,
and the sensor controls the voltage that gets sent to the fan.  What is
happening is the sensor for the front fan never gets hot enough to send any
voltage to the fan.  Where-as the sensor for the rear fan, is at the top of
the case where the heat is, and naturally, gets hot and kicks off the rear
fan to exhaust the air.

This is fine and I can see how it's meant to work, but it's pointless having
a fan in the front when it rarely ever starts.  Most of the time, my system
is using the rear fan to exhaust air.  I do have a couple of Panaflo fans
(120x120x38mm) that plug directly into the PSU, they work tremendously, but
they are loud.  I can't hear the TV 5-feet away with these things.

So because of that, I opted for the current setup (Nidec 120x120x25mm),
these fans plug directly into the MOBO.

Bit annoying...

~ K ~

Karl Dieter wrote:
> The sensor is on the MOBO.  Basically, the fan sends it's RPM to the
> MOBO, then to the SMSC controller.  The sensor is placed near the fan
> plug header, and the sensor controls the voltage that gets sent to
> the fan.  What is happening is the sensor for the front fan never
> gets hot enough to send any voltage to the fan.  Where-as the sensor
> for the rear fan, is at the top of the case where the heat is, and
> naturally, gets hot and kicks off the rear fan to exhaust the air.

We that's dumb fuck stupid you sould use 1 sensor so that both of the 
fans are synced (havin' 2 fans at 1/2 speed gives better cooling that 1 
of them at full speed), that sensor should be on a header so you can 
place the sensor where you can get a decent idea of what the case temp 
(ie. the exaust air temp ;)


> This is fine and I can see how it's meant to work, but it's pointless
> having a fan in the front when it rarely ever starts.  Most of the
> time, my system is using the rear fan to exhaust air.  I do have a
> couple of Panaflo fans (120x120x38mm) that plug directly into the
> PSU, they work tremendously, but they are loud.  I can't hear the TV
> 5-feet away with these things.

Well go to someone like www.TekHeads.co.uk and get your self a 3-pin 
(male) to 4 pin molex (male) connector and plug you fan in.

If you want the speed sensor to work you've got to do a little work, 
you need to take the earth and the singnal wire to the 3 pin plug, to 
do this I genraly cut earth 1/2 way between the fan and the 3 pin and 
remove the 12v line from the 3 pin plug, get a 4pin molex (male) 
connect up the 12v straignt into the PSU 12v, then split the earth one 
going to the molex and the other going to the 3 pin, this'll give you 
both 12v constent and speed sensor (oh 5v on PSU's is yellow).

Ani

-- 
Disclaimer:
I reserve the right to be 100% wrong/have malformed &/or biased 
opinions, feel free to politely correct me. If however you do not wish 
to be courteous go {the remainder of this disclaimer has been censored}!

Tue, 25 Sep 2001 05:49:40 GMT, "Karl Dieter" <nospam@mydomain.com>
wrote:

>The sensor is on the MOBO.  Basically, the fan sends it's RPM to the MOBO,
>then to the SMSC controller.  The sensor is placed near the fan plug header,
>and the sensor controls the voltage that gets sent to the fan.  What is
>happening is the sensor for the front fan never gets hot enough to send any
>voltage to the fan.  Where-as the sensor for the rear fan, is at the top of
>the case where the heat is, and naturally, gets hot and kicks off the rear
>fan to exhaust the air.
>
>This is fine and I can see how it's meant to work, but it's pointless having
>a fan in the front when it rarely ever starts.  Most of the time, my system
>is using the rear fan to exhaust air.  I do have a couple of Panaflo fans
>(120x120x38mm) that plug directly into the PSU, they work tremendously, but
>they are loud.  I can't hear the TV 5-feet away with these things.
>
>So because of that, I opted for the current setup (Nidec 120x120x25mm),
>these fans plug directly into the MOBO.
>
>Bit annoying...
>
>~ K ~
>
>

Hi Karl.

If the system runs cool enough, enjoy it, there's no need to have that
front fan going, especially if it's sensor is correct, that the lower
case temps are fine.  Then again, the reason your front fan isn't
turning, could be that it's a 120mm... is that the size the
motherboard was engineered to control?  Possibly an 80mm would be
turning?

If you really want to use those panaflows, you could try them with the
ground connected to the PSU 5V (red) for a potential of 7V.  The
easiest way would be to move the fan's ground contact in the fan
connector, so it mates with the PSU connector 5V line.  Whether they
will start at 7V may depend on the specific model.

How loaded is this system?  It's very uncommon for a PC to need two
120mm fans.


Dave
___________________________
Please Reply to Newsgroup,
Google ate my email address.

> 
> Hi Karl.
> 
> If the system runs cool enough, enjoy it, there's no need to have that
> front fan going, especially if it's sensor is correct, that the lower
> case temps are fine.  Then again, the reason your front fan isn't
> turning, could be that it's a 120mm... is that the size the
> motherboard was engineered to control?  Possibly an 80mm would be
> turning?
> 
> If you really want to use those panaflows, you could try them with the
> ground connected to the PSU 5V (red) for a potential of 7V.  The
> easiest way would be to move the fan's ground contact in the fan
> connector, so it mates with the PSU connector 5V line.  Whether they
> will start at 7V may depend on the specific model.
> 
> How loaded is this system?  It's very uncommon for a PC to need two
> 120mm fans.
> 
> Dave
> ___________________________
> Please Reply to Newsgroup,
> Google ate my email address.

Connecting a fan between 12V and 5V is not a good idea. If the fan 
motor shorts out you got big problems to more expensive  hardware 
than a $10 fan.
If you more less noise, buy a lower rpm fan.

Rick Baker wrote:

> Connecting a fan between 12V and 5V is not a good idea. If the fan
> motor shorts out you got big problems to more expensive  hardware
> than a $10 fan.
> If you more less noise, buy a lower rpm fan.

I agree. If the +5v has a very fast OVP circuit, the damage might be
minimal. But it isn't worth the risk.

I use a 50 or 100 ohm chassis mount pot.

-- 
Better than hearing "Lady Day", or checking in at Monterey...
[Remove bodies from address for email.]

Eric wrote:
> 
> Rick Baker wrote:
> 
> > Connecting a fan between 12V and 5V is not a good idea. If the fan
> > motor shorts out you got big problems to more expensive  hardware
> > than a $10 fan.
> > If you more less noise, buy a lower rpm fan.
> 
> I agree. If the +5v has a very fast OVP circuit, the damage might be
> minimal. But it isn't worth the risk.
> 
> I use a 50 or 100 ohm chassis mount pot.
> 
> --
> Better than hearing "Lady Day", or checking in at Monterey...
> [Remove bodies from address for email.]

An excellent inexpensive alternative.

Tue, 25 Sep 2001 16:18:43 -0500, Rick Baker <rb2b@ti.com> wrote:


>Connecting a fan between 12V and 5V is not a good idea. If the fan 
>motor shorts out you got big problems to more expensive  hardware 
>than a $10 fan.
>If you more less noise, buy a lower rpm fan.

Is it a common mode of failure (or even possible) for a brushless DC
computer fan to short out the power leads?  I'm not aruging that this
isn't remotely possible, but it seems unlikely, without introducing a
power drill, pond, or some other overt act of destruction.

Are there any relevant schematics you could link to?


Dave
___________________________
Please Reply to Newsgroup,
Google ate my email address.

Probably not common at all, but why risk it when there are inexpensive 
alternatives? A lower rpm fan, a modified fan with a low ohm pot, etc.

kony wrote:
> 
> Tue, 25 Sep 2001 16:18:43 -0500, Rick Baker <rb2b@ti.com> wrote:
> 
> >Connecting a fan between 12V and 5V is not a good idea. If the fan
> >motor shorts out you got big problems to more expensive  hardware
> >than a $10 fan.
> >If you more less noise, buy a lower rpm fan.
> 
> Is it a common mode of failure (or even possible) for a brushless DC
> computer fan to short out the power leads?  I'm not aruging that this
> isn't remotely possible, but it seems unlikely, without introducing a
> power drill, pond, or some other overt act of destruction.
> 
> Are there any relevant schematics you could link to?
> 
> Dave
> ___________________________
> Please Reply to Newsgroup,
> Google ate my email address.

Wed, 26 Sep 2001 16:15:22 -0500, Rick Baker <rb2b@ti.com> wrote:

>Probably not common at all, but why risk it when there are inexpensive 
>alternatives? A lower rpm fan, a modified fan with a low ohm pot, etc.
>

Well, I've had fans running fine at 7V for years now, never a problem
with them, though I've gotten ambitious a few times, and converted
some boxes to using power resistors instead, which i "sized" using a
pot like you suggested.

One reason i suggested that approach, is that not everyone is familiar
with electronics, has an iron, or wants to use it, while it's pretty
easy to just swap a couple tab-locked wires.


Dave
___________________________
Please Reply to Newsgroup,
Google ate my email address.

I wouldn't suspect that you would have a problem. It's a risk/reward
factor. The risk, although a very remote risk, is present, and the fact
that there are easy, inexpensive alternatives are a good reason why
it doesn't make much since to subject the system to even a remote risk.
As for the non-technically inclined, the purchase of an relatively 
inexpensive low RPM fan is a good solution.
I think this issue has been beaten to death, so for those still
willing to do this mod even with the knowledge of a remote risk, then
more power to you. I only wanted to give them the information to make
an educated decision.

kony wrote:
> 
> Wed, 26 Sep 2001 16:15:22 -0500, Rick Baker <rb2b@ti.com> wrote:
> 
> >Probably not common at all, but why risk it when there are inexpensive
> >alternatives? A lower rpm fan, a modified fan with a low ohm pot, etc.
> >
> 
> Well, I've had fans running fine at 7V for years now, never a problem
> with them, though I've gotten ambitious a few times, and converted
> some boxes to using power resistors instead, which i "sized" using a
> pot like you suggested.
> 
> One reason i suggested that approach, is that not everyone is familiar
> with electronics, has an iron, or wants to use it, while it's pretty
> easy to just swap a couple tab-locked wires.
> 
> Dave
> ___________________________
> Please Reply to Newsgroup,
> Google ate my email address.

> Hi Karl.

Hallo.

>
> If the system runs cool enough, enjoy it, there's no need to have that
> front fan going, especially if it's sensor is correct, that the lower
> case temps are fine.  Then again, the reason your front fan isn't
> turning, could be that it's a 120mm... is that the size the
> motherboard was engineered to control?  Possibly an 80mm would be
> turning?

The fan appears OK, even at 120mm.  I think it's just that the sensor never
gets hot enough to start the fan.  If I lightly blow a hair-dryer on the
sensor, the fan starts up as supposed.

>
> If you really want to use those panaflows, you could try them with the
> ground connected to the PSU 5V (red) for a potential of 7V.  The
> easiest way would be to move the fan's ground contact in the fan
> connector, so it mates with the PSU connector 5V line.  Whether they
> will start at 7V may depend on the specific model.

I'm not over willing to use the Panaflos, I'd just like a steady airflow
through-out the case.

>
> How loaded is this system?  It's very uncommon for a PC to need two
> 120mm fans.

P4 1.7, 512MB RDRAM, 100GBs Ultra160 SCSI, 3Com 990SVR ARM NIC, Creative
AUDIGY, Adaptec FireConnect.  Fairly loaded, not excessive though.

Danke.

Sat, 06 Oct 2001 04:08:38 GMT, "Karl Dieter" <nospam@mydomain.com>
wrote:

>> Hi Karl.
>
>Hallo.
>
>>
>> If the system runs cool enough, enjoy it, there's no need to have that
>> front fan going, especially if it's sensor is correct, that the lower
>> case temps are fine.  Then again, the reason your front fan isn't
>> turning, could be that it's a 120mm... is that the size the
>> motherboard was engineered to control?  Possibly an 80mm would be
>> turning?
>
>The fan appears OK, even at 120mm.  I think it's just that the sensor never
>gets hot enough to start the fan.  If I lightly blow a hair-dryer on the
>sensor, the fan starts up as supposed.
>
>>
>> If you really want to use those panaflows, you could try them with the
>> ground connected to the PSU 5V (red) for a potential of 7V.  The
>> easiest way would be to move the fan's ground contact in the fan
>> connector, so it mates with the PSU connector 5V line.  Whether they
>> will start at 7V may depend on the specific model.
>
>I'm not over willing to use the Panaflos, I'd just like a steady airflow
>through-out the case.
>
>>
>> How loaded is this system?  It's very uncommon for a PC to need two
>> 120mm fans.
>
>P4 1.7, 512MB RDRAM, 100GBs Ultra160 SCSI, 3Com 990SVR ARM NIC, Creative
>AUDIGY, Adaptec FireConnect.  Fairly loaded, not excessive though.
>
>Danke.
>
>

If there aren't any components getting too hot, you could just leave
it alone.

You could get an adaptor to power it at full speed from a PSU line if
you really want it running, or try newsgroups like sci.electronics (or
similar) for the possibility of making a thermal-sensing fan speed
controller, or just buy a fan with this feature, and use the adaptor I
mentioned.


Dave
___________________________
Please Reply to Newsgroup,
Google ate my email address.