Sign in with
Sign up | Sign in
Your question

7V Fan slowdown mod!

Last response: in Overclocking
Share
February 26, 2003 10:49:00 PM

This is a very easy to implement and very effective mod if you have some loud fans you want to slowdown.
I did it with one of my 50mm buzzers and its now nice and tame.

<A HREF="http://www.viperlair.com/articles/howto/7vmod.shtml" target="_new">http://www.viperlair.com/articles/howto/7vmod.shtml&lt;/A>

In addition to the 12v - 5v = 7v mod, you can also use 5v as active and a neutral as ground, getting 5v if you want a REALLY slow running fan!

highly reccomended!

<i>"Revenues were less than robust"</i> - QWEST
<i>"The company applied its accounting policies incorrectly"</i> - WORLDCOM
<i>"Certian financial adjustments may be required"</i> - AOL+TW.

More about : fan slowdown mod

Anonymous
a b K Overclocking
February 26, 2003 11:34:52 PM

Warning! Some 12V fans have been known to no start at 5V.

That would be weak if it were on something that REALLY needed it.
February 27, 2003 3:35:01 AM

<b>Be warned</b>... not all fans will start reliably on 7 volts!

As a rule, you can quiet most fans down considerably by putting a 10 ohm 2 watt resistor in series with the +12volt wire (usually red). Cut the red wire. Strip the ends, wrap the two stripped ends, one on each side of the resistor close to the resistor body, cut away the excess resistor leads, solder, tape... and you're off to the races. The result is a fan that runs on about 10 volts and you can still plug it into your motherboard and monitor it's speed through software...

I've done this on a number of the fans on Spire coolers. It doesn't seem to change it's cooling effect at all and it quiets them down considerably... Their 4300rpm 70mm fan ends up spinning at about 3750 with the inline resistor.

On 2800 rpm case fans, the result is usually about 2500rpm and virtually no noise at all.



<b>(</b>It ain't better if it don't work.<b>)</b>
Related resources
February 28, 2003 4:57:04 AM

ive done that, on several fans, they were all about 7000 rpm fans. now they are much quieter!

my computer is so fast, it completes an endless loop in less than 4 seconds!
March 2, 2003 4:06:23 AM

well i have a nice passthru connector... so ill try this 7v mod on a whole range of fans... see what works and what doesnt. il keep u posted.

<i>"Revenues were less than robust"</i> - QWEST
<i>"The company applied its accounting policies incorrectly"</i> - WORLDCOM
<i>"Certian financial adjustments may be required"</i> - AOL+TW.
March 2, 2003 5:05:50 AM

I've seen this 7 volt mod before, on pentium 233s, it's not new... and I don't recommend it.

Be aware that since you are lifting the negative lead of the fan away from ground the tachometer will stop functioning which will prevent you from monitoring fans in software. Trying to connect the tach to your motherboard will almost certainly damage the SMB chips... The tach will be producing 5 volt pulses, 5 volts above ground and putting 10volts into a ttl level input will hurt it, might even damage the fan.

There are far better ways of slowing down a fan...

I put up a simple circuit using a resistor and capacitor, which will cost about 75cents to build. It's 100% reliable and doesn't interfere with the tachometer's function. (See "Simple Fan Quieting Circuit" in this forum)

If you want to do it super dirt cheap, leave out the capacitor. Really it's only there to make sure balky fans start everytime. Just cut the red wire and solder a 25cent resistor in series... 10 ohms, 2 watts is usually about right.

There is another mod using 3 diodes in series. Total cost for this one is also well under a dollar.

For a buck and a half, you can build a really simple variable speed controller with a small potentiometer, one resistor and a transistor... The range is about 8 volts to 11.5 volts and it also does not interfere with the tachometer. For the added price of 25cents for a capacitor, you can add a kick start that guarantees the fan will start.

For 2 bucks you can have temperature sensing using a thermister, one resistor and a transistor.


Why bother with something that's so obviously bad science, when, for less than the price of the connectors for the 7volt mod, you can do it right?

But, all that said, no harm in experimenting... :smile:

<b>(</b>It ain't better if it don't work.<b>)</b>
March 2, 2003 5:24:03 AM

EXPERIMENTAL STUDY: FANS STARTING AND RUNNING @ 7V
good = runs smooth and starts multiple times without fuss.
(I tried at least 5 'no rpm' starts for each fan to ensure it works)

(fan rpm's are for standard 12v operation)

120mm 1750rpm panaflo = good. damn quiet.
80mm 2100rpm vantec stealth = good. wow.
60mm fan atop a stock XP1700+ cooler = damn quiet!
60mm fop32-1 fan (4500rpm?) = good.
80mm yum-cha superflower casefan (~1800rpm) = good.
80mm 2000rpm panaflo silent fan = good. UBER silent now hehehe.
50mm yum chaa fan of buzzyness = good. no longer buzzy :) 
TT 2700rpm std fan = good.
Delta 5000rpm focused flow screamer = good. but still obnoxious at 7v. :smile:
Delta 5000rpm screamer + 2 rheostats inline at minimum = STILL STARTS!, but slow speeding up. And at last... no longer obnoxious noises! hehehe :lol: 
80mm 3600 pabst = good. killed alot of its noise.
Enermax 80mm variable on slow = good. so slow u can see then writing on the fan hub spinning :smile:
evercool 2500rpm 120mm silver fan = good. acceptable noise now.
this is fuuuun :lol: 
I must say that for all apart from the uber loud fans 7v is very acceptable. only the Evercool, pabst 3600rpm or delta screamer could need more...
But anyway, doing 5v mod now... results to be updated :smile:

<i>"Revenues were less than robust"</i> - QWEST
<i>"The company applied its accounting policies incorrectly"</i> - WORLDCOM
<i>"Certian financial adjustments may be required"</i> - AOL+TW.
March 2, 2003 8:15:44 AM

EXPERIMENTAL FAN STUDY: FANS @ 5V!

Well all my fans had no trouble at 7V... so lets try 5V operation & starting. once again, each fan was started from 0rpm 5 times to check that it would start properly. rpm's shown are for normal 12v operation.

50mm yummchaa buzzer = good. very slow spinning. damn silent. damn near useless too LOL.
60mm fop32-1 4500rpm fan = good. almost silent.
60mm generic low profile xp1700+ cooler fan = good. silent.
80mm enermax variable = BAD. Didnt run at all on low. ran fine on medium to high. Even starting at high and going to slow made it stop.
Just tooo low for it
80mm superflower yumchaa fan = good.
80mm panaflo 2000rpm = good. so quiet its not funny.
80mm Vantec stealth 2100rpm = good.
80mm TT 2700rpm = good.
80mm pabst 3600rpm = good. silent (at last)
80mm delta screamer 5000rpm = good. Still not silent though! definate bearing buzz and pushing out decent CFM. I estimate 2000 to 2500rpm.
120mm Panaflo 1750rpm = good. silent. so slow u cansee the blades.
120mm Evercool 2500rpm = Good. Very slow. VERY quiet.

Now i gotta say that 5v on most of those fans was complete overkill. specially the ones that are allready silent. How much more silent than silent can one get???
But its good to know that all but 1 ran fine at 5v... and the one that didnt had built in throttling allready (enermax variable).
the varaible enermax would start from 0rpm at down to around 2/3 of max speed. below that, nada.

And the delta screamer STILL made noise, a lowish buzz... which got me thinking....
Q. Whats better than a 5v fan???
A. A 5v fan on 2 serial rheostats of course!!!

MWOOO HAA HAA HAA

<i>"Revenues were less than robust"</i> - QWEST
<i>"The company applied its accounting policies incorrectly"</i> - WORLDCOM
<i>"Certian financial adjustments may be required"</i> - AOL+TW.
March 2, 2003 8:37:11 AM

EXPERIMENT: SOMETHING JUST STOOPID: FANS @ 5V with 2 inline Rheostats on sloooooooow.

50mm yumchaa = would only start spinning with a push.
60mm fop32-1 fan = yep it starts good!
60mm low profile stock cooler fan = this starts too!
80mm enermax variable = NOPE on low, YES on high.
80mm superflow = yes
80mm vantec stealth = no, only started with a push.
80mm panaflo = no, only started with a BIG push.
80mm pabst 3600rpm = Very slow starting... but it did start by itself. right on the edge
80mm delta screamer = yes! at last it is silent running LOL. it took half of the first rheo to finally be truly silent running.
120mm panaflo = yes. slow starting but still goes. VERY slow spinning!
120mm evercool = no... only started with a decent push. also extreemly slow spinning.

So there u have it.
Even with 5v and two rheostats in serial most of the fans still managed to run!
Course for most they were silent a long time ago... and some started up extreemly slowly.

that was fun :smile:

<i>"Revenues were less than robust"</i> - QWEST
<i>"The company applied its accounting policies incorrectly"</i> - WORLDCOM
<i>"Certian financial adjustments may be required"</i> - AOL+TW.<P ID="edit"><FONT SIZE=-1><EM>Edited by LHGPooBaa on 03/02/03 05:38 AM.</EM></FONT></P>
March 2, 2003 9:46:51 AM

Could you explain the thermistor controlled cicuit with one
or is it two transistors. I've been looking at voltage regulators (LM317) comparitors (L200c) and pulse-width controllers (MIC502) to control the speed of any DC brushless fan with a thermistor. The speed of course depends on the thermistors ambient temperature. I know enough to know that all is not as simple as it seems.Your
advice (technical) so far has been very good. If you can explain to me (us) how this all works, I thank you.


avandonk
March 2, 2003 12:54:00 PM

I always use that mod if a fan it to noisy.

But if you do this to a CPU fan be carefull to check if the fan keeps puttting out enough air on 7V.

Anyway a resistor in serie will do the same thing. The amperage on the total line will keep the same, but the voltage on the resistor and fan differ.

My dual-PSU PC is so powerfull that the neighbourhood dims when I turn it on :eek: 
March 2, 2003 12:56:04 PM

A thermister is simply a resistor that reacts to heat. As the temperature rises, it's resistence falls off. Replace any resistor in a transistor's bias chain and you end up with a temperature controlled circuit... :smile: whether you want one or not.

It's way too difficult to be drawing schematics on this board, so I can't offer you that...

Basically, the circuit consists of a 10k (at 20c) thermistor in series with a 22k resistor, operating as a voltage divider, hooked across the 12volt supply to the fan so that the thermistor is on the positive and the resistor is grounded. An NPN transistor is used in an emitter follower configuration, with it's base connected to the junction of the thermistor and resistor, it's collector on the 12 volt supply and it's emitter feeding the fan's positive power lead. The fan's negative lead is hooked to ground, like usual.

As the temperature changes, the thermistor's changing resistence alters the characteristics of the voltage divider so that the voltage to the base of the transistor is increased. The emitter of the transistor will follow the voltage at it's base... increasing voltage to the fan. The transistor provides current gain, to drive the fan.

Obviously, how well this works depends on the characteristics of the thermistor. It must go from it's base resistence of 10k at 20c to almost nothing at about 40c. (And yes they do make them :smile: ). You are also well advised to hook this up to a high speed fan, to ensure you can get enough airflow.

It's a painfully simple circuit, I just wish I could draw it for you.


<b>(</b>It ain't better if it don't work.<b>)</b>
March 2, 2003 1:04:41 PM

It's about time we can all post pictures and/or diagrams on
this place.

avandonk
March 2, 2003 1:13:17 PM

I hear ya! Even the PRE markup isn't working, so I can't even give you a crude picture...




<b>(</b>It ain't better if it don't work.<b>)</b>
March 2, 2003 8:25:45 PM

yeah... i was just mega curious as to what would start on 5v... or less :smile:

The resistor way is good... but u have to cut things and wire up resistors that get hot then cover over the connection so it cant short etc.

<i>"Revenues were less than robust"</i> - QWEST
<i>"The company applied its accounting policies incorrectly"</i> - WORLDCOM
<i>"Certian financial adjustments may be required"</i> - AOL+TW.
Anonymous
a b K Overclocking
March 2, 2003 9:15:14 PM

LHGPooBaa,

It looks like you have some interesting data. I would say though it is still no statistical sampling, and would advise caution to any one who goes the voltage divider route for a CPU fan.

You could probably trust the results if you could do your testing at V-10% (where v is the voltage to run the fan), and verified multiple startups on each winding. I still prefer a statistical sampling though. Also this kind of test would only be valid for the current power supply loading as well, add abnother hard drive or something and you should retest.

Kevin

edit: this doesn't consider wear and tear, which will surely occur in most if not all cases. That said, the above testing is really only valid for the brief window of time whe it can be assumed there is "negligible" loss of performance due to wear and tear.

Props to Teq in a later post in this thread.<P ID="edit"><FONT SIZE=-1><EM>Edited by knewton on 03/02/03 05:06 PM.</EM></FONT></P>
March 2, 2003 9:26:57 PM

true enough.
The PSU is a enermax 550W.
the molex lead i was using was connect to 2 system fans, no other components like hard drives.

Its what i normally do. have the CD's and floppy on one lead, hard drives on their own and fans on the third. (enermax has lots of leads)

<i>"Revenues were less than robust"</i> - QWEST
<i>"The company applied its accounting policies incorrectly"</i> - WORLDCOM
<i>"Certian financial adjustments may be required"</i> - AOL+TW.
March 2, 2003 10:17:52 PM

Not to mention the effects of wear and age on the fan itself.


<b>(</b>It ain't better if it don't work.<b>)</b>
March 2, 2003 10:19:17 PM

Ever followed those leads back inside the case?

They all go to the same place on the PCB.




<b>(</b>It ain't better if it don't work.<b>)</b>
Anonymous
a b K Overclocking
March 2, 2003 10:24:40 PM

very good point.
March 5, 2003 12:46:13 PM

Thanks mate,
I ended up building a voltage control circuit
using an LM317 .Of course it will not get above 10.5 V,but
there are many screamer fans that just need taming, and they
come nearly for free. The control over the said fan is only
what one desires. The unit (circuit) I put together can handle 1.5 Amp and the fan(s) respond not to bad mate.Or is it the thermistor?
bertvandonk


avandonk
March 5, 2003 1:22:52 PM

Yep, I've seen those scattered around the web. Not necessarily the best way of doing it, but it works well enough.



<b>(</b>It ain't better if it don't work.<b>)</b>
March 5, 2003 2:25:35 PM

Of course the correct way is pulse modulated etc. Please enlighten me if there is a better way.
I just want to leave my home at lets say 7AM when it is about 18degC and by the time I get home it's 35degC or42degC) and hope all my systems are still running. This is not an exercise in overclocking ,but an exercise in maintaining
a system that will run and look after itself no matter what the temperature gets to. My computers run 24/7. Anyway
all advice is welcome no-matter how irrelevant.



avandonk
March 5, 2003 3:54:03 PM

Please don't take me wrong, the circuit with the LM317 is a perfectly good fan control. Use it with high speed fans, place the thermistor where the case gets hottest and it should give you excellent results. It even has enough juice to operate multiple fans, so you only need one homebrewed board to control all the case fans in your system. Nothing wrong with that.

There are alternatives... simple 1 transistor switches that turn fans on and off at a preset temperature... 1 transistor thermal controls that run on individual fans... etc. Each has advantages and disadvantages.

I wasn't saying there's anything wrong with the way you're doing it... :smile:


<b>(</b>It ain't better if it don't work.<b>)</b>
March 7, 2003 2:25:16 PM

I got some real bad experience with PWM... it would create a very irritating whining noise with my Delta HP120 fan.

My dual-PSU PC is so powerfull that the neighbourhood dims when I turn it on :eek: 
March 7, 2003 3:01:49 PM

Well, yes... looked inside a fan lately?

They aren't just an electric motor. There's a couple of active components in there to replace the normal brushes of a DC electic motor. Brushless fans need clean DC current to work correctly. The PWM method was probably driving them nuts.



<b>(</b>It ain't better if it don't work.<b>)</b>
March 7, 2003 3:23:20 PM

Ofcourse... now I know what that brushless means.

My dual-PSU PC is so powerfull that the neighbourhood dims when I turn it on :eek: 
March 7, 2003 3:46:25 PM

If you have the chance, take a dead fan apart. It's really some very interesting technology, it's actually an inside out motor where the windings are on the stator and the commutator is a magnetic ring with several pole reversals around it's circumfrence.

What the chip does is reverse the polarity of the stator windings to cause the magnetic ring commutator to jump towards the winding's opposite pole, reverse it often enough and you get a spinning motor.

Fans with a tachometer in them also have a second small chip on the internal circuit board. This works by monitoring the stator windings for reverse pulses caused by the magnetic commutator, which are then cleaned up and sent back down the third wire.

Kewl stuff!




<b>(</b>It ain't better if it don't work.<b>)</b>
March 8, 2003 12:00:30 PM

The last time I disassembled a fan was before I learned how electric engine's work at school. Luckily I don't have any dead fans at the moment.

My dual-PSU PC is so powerfull that the neighbourhood dims when I turn it on :eek: 
!