The "7V trick" on fans with adjustable voltage - RPM gone!

Control: management of fans, temp/rpm monitoring via soft/hardware

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
runeks
Posts: 26
Joined: Wed Jun 18, 2008 1:07 am
Location: Denmark

The "7V trick" on fans with adjustable voltage - RPM gone!

Post by runeks » Thu Mar 17, 2011 9:28 am

Hello all

In my case I have three Scythe Slipstream 1200RPM (SY1225SL12M) fans connected to an Abit IP35 Pro XE motherboard. This motherboard allows me to adjust the voltage of the fan connectors from between 8V to 12V. When the Scythes are running at 8V - instead of 12V - their speed is reduced to around 1000 RPM, which reduces the noise significantly.
I am, however, not quite satisfied with the noise level, I'd like to go lower! So I thought about the following: instead of connecting the ground wire of the 3 pin fan connector to the motherboard's ground pin, connecting it to the 5V pin of a Molex connector. So instead of the fan having 0V on the ground wire, it'd have 5V. The "plus" wire of the fan would still be 8V to 12V, adjustable in the BIOS of my motherboard, but since the ground wire is now 5V, the actual voltage drop would be five volts lower, and I would be able to adjust the fan voltage from 3V to 7V.

So I did just that. Bought some 3 pin fan extensions cables, cut off the ground wire from the female end that's supposed to connect to the motherboard, and connected this wire to the 5V wire of a Molex connector.
Does it work? I'm not sure... I lost the RPM "signal", apparently - when I connect the fan to this modified fan extension cable, and then connect this extension cable to the motherboard's fan connector, the rotational speed of the fan no longer shows in the BIOS. The fan is spinning fine, no problem there. I haven't touched the yellow "RPM"-wire of the modified fan extension cable.

Any ideas why this would happen?

EDIT: I just performed the necessary test, to see if it's the RPM-wire of the fan extension cable that is somehow broken: I disconnected the ground wire of the fan extension cable from the 5V Molex wire, and connected it at its "rightful" place, where I cut it off, so that it's now just a normal 3 pin fan extension cable, unmodded. Result: RPM reading works again.
So it seems that the fan speed reading simply doesn't work when the low voltage wire is 5V instead of 0V. Does anyone know why? And, possible, how to get around this? I'd really like to see the rotational speed of my fans, especially when running at 3V.
You do not have the required permissions to view the files attached to this post.

washu
Posts: 571
Joined: Thu Nov 19, 2009 10:20 am
Location: Ottawa

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by washu » Thu Mar 17, 2011 10:34 am

When you do the 7 volt trick it changes the low voltage on the RPM pin from 0V to +5V. As the RPM pulse never goes back to 0V as far as the motherboard is concerned it no longer works.

dhanson865
Posts: 2198
Joined: Thu Feb 10, 2005 11:20 am
Location: TN, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by dhanson865 » Thu Mar 17, 2011 11:07 am

some motherboard also won't detect or display RPMs below a threshold. It varies but I've seen cases where anything below 300RPM or below 500 RPM or some random number like that wouldn't display even though the fan was on a normal 12v connection and I was varying the fan by way of speedfan.

I'd lower speedfan's percentage 1% at a time until the RPM went to 0 and look inside the case to see the fan still spinning. Not terribly user friendly when a board won't report low RPMs but not much you can do about it when it happens.

Most fans won't even turn at 3v here are some speeds at 5v

SY1225SL12SL 170 RPM
SY1225SL12L 420 RPM

The SPCR review of the SY1225SL12M you have shows

720 RPM at 5V
620 RPM at 4V

So I'm guessing your issue is the signal not going to 0V not the low RPM not reported issue I mentioned.
.
Please put a country in your profile if you haven't already.
This site is international but I'll assume you are in the US if you don't tell me otherwise.
RAID levels thread http://www.silentpcreview.com/forums/viewtopic.php?p=388987

dhanson865
Posts: 2198
Joined: Thu Feb 10, 2005 11:20 am
Location: TN, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by dhanson865 » Thu Mar 17, 2011 11:59 am

Maybe the fix is to avoid the 7V trick and go for a resistor or Zener diode in line with the 12v. That's beyond my knowlege so hopefully someone else will chime in with the suggested part(s).
.
Please put a country in your profile if you haven't already.
This site is international but I'll assume you are in the US if you don't tell me otherwise.
RAID levels thread http://www.silentpcreview.com/forums/viewtopic.php?p=388987

MikeC
Site Admin
Posts: 12283
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by MikeC » Thu Mar 17, 2011 12:19 pm

A Zalman fanmate will work fine to reduce that 8V level lower.
Mike Chin, SPCR Editor/Publisher
Support SPCR by buying your gear through this link: Amazon

Plekto
Posts: 398
Joined: Tue Feb 19, 2008 2:08 pm
Location: Los Angeles

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by Plekto » Fri Mar 18, 2011 7:07 am

The problem with the fanmate is that it eventually will burn out - it gets quite hot and is from what I can tell, a pretty miserable design. A resistor is of course the best solution, but it takes careful planning since most systems react poorly to 5v cooling levels for any extended period of time. I just turn all of the fan management off as a result and have the fans running at fixed speeds. That also took a bit of experimenting to get right.

MikeC
Site Admin
Posts: 12283
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by MikeC » Fri Mar 18, 2011 8:40 am

I've been using umpteen fanmates (both v1 and 2) since they first appeared ~7 yrs ago & have never had a single one burn out. They don't run hot, imo, and they simply cannot draw any more power than the fan by itself. Basically, a voltage fan controller works hardest when the fan speed is most reduced -- it absorbs the power than otherwise would go to the fan. So if a fan is rated for 0.3A @ 12V, the power is 3.6W; a voltage controller which slows the fan by 50% dissipates ~1.8W.

For the OP who wants the voltage dropped further from 8V, the amount of power dissipation required in the fanmate (or any voltage controller) is trivial, probably not more than ~1W.
Mike Chin, SPCR Editor/Publisher
Support SPCR by buying your gear through this link: Amazon

andymcca
Posts: 404
Joined: Mon Jan 11, 2010 8:19 am
Location: Boston, MA, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by andymcca » Fri Mar 18, 2011 9:31 am

Assuming DC, the most it could draw would be half the max fan load, since Pmax would be with resistance matching. :D

Edit: that said, I can certainly go buy a 1W pot :)
Edit2: Looked, if you care, and Fan Mate 2 has a rating of 6W. I'm assuming thats fan rating of 6W, so internal rating of 3W?
Edit3: Just realized: Do we know that it adds series resistance, and is not some more complicated scheme?
Currently running Mint 14 MATE on all machines.
Desktop/Gaming: E8400 / 650ti + Accelero S1 / Antec Solo
Myth Backend + Home Server: D945GSEJT / WD15EADS / Hauppage 950Q / Home made case
Myth Frontend + HTPC: Sempron 140 / Asus M4A785-M
Notebook:AMD A4-3300M

MikeC
Site Admin
Posts: 12283
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by MikeC » Fri Mar 18, 2011 11:36 am

Fanmate is an active device, not a resistor. Insertion loss is ~1V. A detailed review: http://www.silentpcreview.com/article85-page1.html
Mike Chin, SPCR Editor/Publisher
Support SPCR by buying your gear through this link: Amazon

Lensman
Posts: 147
Joined: Thu Jul 26, 2007 1:15 am

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by Lensman » Sat Mar 19, 2011 10:36 am

If you decide to move away from using your MB controls and just want a static solution, I've been happy with the "Deluxe Multi Power Port" from FrozenCPU or Microcenter. It provides 12V, 7V and 5V outputs.

I think you can pull the power leads to this unit and fan indicator lead to your MB and still get RPM indications from your MB. I didn't bother to do this myself, though.

runeks
Posts: 26
Joined: Wed Jun 18, 2008 1:07 am
Location: Denmark

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by runeks » Sat Mar 19, 2011 11:43 am

Thanks for the input everyone! I must admit I'm kind of hooked on the idea of being able to control the voltage from the BIOS. Basically, I have no idea how my system will react when the fans are running at only ~3V, so it'd be optimal for me to be able to bump up the voltage half a volt or so, if I can see that the system isn't adequately cooled. Also, the BIOS for my mobo has a feature where it will adjust the fan voltage between two pre-defined values, based on a reference temperature. Ie. if, for example the CPU temperature is below 50°C, use 8V, and if above, use 10V. So - assuming that I am able to drop the voltage below the standard 8-12V range - using this approach, I am able to have a "really silent" mode where the voltage is, say, 3V, and if the CPU temperature rises above a certain mark, the voltage is increased to, for example, 5V. That's a fairly valuable feature, as I see it, and I'd really like to retain this, if possible.
dhanson865 wrote:Maybe the fix is to avoid the 7V trick and go for a resistor or Zener diode in line with the 12v. That's beyond my knowlege so hopefully someone else will chime in with the suggested part(s).
I've seen a couple of posts regarding this, and it sounds really interesting - if I have understood it correctly. I'd love to get some more input on this. Is the voltage drop that the Zener diode delivers, independent of the power draw of the fan, or do I have to calculate what type of Zener diode to buy (a lá how it'd be done if using a resistor) based on the power draw/wattage of the fan? Or is it just as simple as buying, for example, the 1N5338B 5.1V 5W Zener diode, which will drop the voltage by 5.1V, tolerate up to 5W (just to be on the safe side as suggested here) and then solder this to the 8V-12V wire of the fan, and then I have 6.9V-2.9V?

runeks
Posts: 26
Joined: Wed Jun 18, 2008 1:07 am
Location: Denmark

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by runeks » Mon Mar 21, 2011 12:49 am

Apparantly my dad had a ton of Zener diodes that he doesn't need, 5.1V 0.5W (BZX79C5V1) , so onto a fan extension wire I soldered four of them in parallel, on the positive, red wire, with the band facing in the opposite direction of the current, ie. going in to the fan header. Result:

with the BIOS set to 8V (should be 2.9V over the AUX2 fan):

Code: Select all

AUX1 Fan:         1020 RPM  (min =  300 RPM)
AUX2 Fan:          420 RPM  (min =  300 RPM)
AUX3 Fan:            0 RPM  (min =  300 RPM)
AUX4 Fan:          960 RPM  (min =  300 RPM)
Success! It varies between 420 and 480, 450 is probably correct.

Adjusting the voltage in the BIOS to the following values, these fan speeds are achieved:

BIOS @ 8V (2.9V over fan): 450 RPM
BIOS @ 9V (3.9V over fan): 570 RPM
BIOS @ 10V (4.9V over fan): 690 RPM
BIOS @ 11V (5.9V over fan): 810 RPM
BIOS @ 12V (6.9V over fan): 870 RPM
(averaged values, they all flip-flop with 30 RPM below and above the values above)

These values seem to agree (within +/- 10%) with Mike's measurements of the same fan at the same voltages.

So all in all it seems like Zener diodes really are easy as hell, and do drop the current at their specified "Zener voltage" seemingly independently of the fan's current rating, if they are positioned in the opposite direction to the current flow (band facing towards source).

I'm now running the fan with the diodes on at full speed (12V in BIOS, 6.9V over fan) to see how hot the diodes get. Hopefully they won't melt the wires on the extension cable or anything. That's really the only thing left that could cause problems. Fingers crossed!

Good day all!

andymcca
Posts: 404
Joined: Mon Jan 11, 2010 8:19 am
Location: Boston, MA, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by andymcca » Mon Mar 21, 2011 4:55 am

Just a heads up, if you care:
I assume you placed the zeners in parallel to split the current amongst them, but (as with most transistor technology) characteristics vary. The breakdown for each of them could be slightly different, and as such one of them could be sourcing most of the load!

Nice work, though. I guess a zener would dissipate the same power as a perfectly sized resistor, but there would be no need to fiddle with zener characteristics! :D
Currently running Mint 14 MATE on all machines.
Desktop/Gaming: E8400 / 650ti + Accelero S1 / Antec Solo
Myth Backend + Home Server: D945GSEJT / WD15EADS / Hauppage 950Q / Home made case
Myth Frontend + HTPC: Sempron 140 / Asus M4A785-M
Notebook:AMD A4-3300M

djkest
Posts: 766
Joined: Mon Sep 10, 2007 1:05 pm
Location: Colorado, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by djkest » Mon Mar 21, 2011 6:49 am

Good information in this thread. I was wondering why 7V seemed to be so popular.

Plekto
Posts: 398
Joined: Tue Feb 19, 2008 2:08 pm
Location: Los Angeles

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by Plekto » Mon Mar 21, 2011 7:42 am

Those type of diodes are generally able to handle the heat. But they DO get quite hot, so be aware of it. Common sense and others who have done this trick before suggest a larger 5W diode (or as close to the power rating of the fan in case it stalls due to an obstruction or shorts out at some time) if possible due to smaller ones often suffering from thermal issues.

runeks
Posts: 26
Joined: Wed Jun 18, 2008 1:07 am
Location: Denmark

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by runeks » Tue Mar 22, 2011 2:34 am

andymcca wrote:Just a heads up, if you care:
I assume you placed the zeners in parallel to split the current amongst them, but (as with most transistor technology) characteristics vary. The breakdown for each of them could be slightly different, and as such one of them could be sourcing most of the load!
I'm not good at electric circuits, but I figure that if I put four of them in parallel, the current would be split in four, and one fourth (65mA) of the total current (0.26A) that the fan draws, would go through each diode. Is this not the case?
Plekto wrote:Those type of diodes are generally able to handle the heat. But they DO get quite hot, so be aware of it. Common sense and others who have done this trick before suggest a larger 5W diode (or as close to the power rating of the fan in case it stalls due to an obstruction or shorts out at some time) if possible due to smaller ones often suffering from thermal issues.
What would happen if the fan stalled due to an obstruction? The voltage drop over the diodes would still be ~5.1V, right? What would the current going through them be?

andymcca
Posts: 404
Joined: Mon Jan 11, 2010 8:19 am
Location: Boston, MA, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by andymcca » Tue Mar 22, 2011 7:00 am

runeks wrote:if I put four of them in parallel, the current would be split in four, and one fourth (65mA) of the total current (0.26A) that the fan draws, would go through each diode. Is this not the case?
Sadly, no. The way zener diodes work, they block almost all current (when reverse biased, the way you have them) up to a certain voltage, and then allow almost all current after that voltage. If the diodes are slightly different (due to random differences in the material and production process), one will hit its breakdown voltage slightly before the others, and will prevent the voltage from reaching the breakdown for the others.
Physical analogy: Say you have a steam boiler with 4 pressure valves, and one blows. Because it is letting the pressure out, the other three valves probably won't release any steam. It will all exit though the one, slightly more sensitive valve.
Now, your diodes could be from the same run, and may be very similar in character. If you are getting some current splitting, I would bet only one or two are doing the work, though. You could see which ones get hot. That would be a good indicator.
runeks wrote:What would happen if the fan stalled due to an obstruction? The voltage drop over the diodes would still be ~5.1V, right? What would the current going through them be?
The diodes could fail two ways: short or open. Depending on what happens, you could have your fan return to full voltage (with the diode just acting as a wire) or stop spinning. The latter would probably burn the diodes up 1 by 1, but you would have a hard time telling unless they smoke!

If you know the diode voltage and the fan current at 12V, you could calculate the wattage you need (assuming DC, fan as a resistive load):
Pdiode = Vdiode * Cfan * (12V - Vdiode) / 12V

(12V - Vdiode) / 12V is just scaling the normal fan current by however much voltage the diode steals.
For a 5V diode and .5A fan, this works out to ~1.5W

Edit: If the .26A you mentioned is the 12V fan rating, a 5.1V anode would dissipate ~.76W, so you would be about 50% overloaded at full load, if one diode is doing all the work.
Currently running Mint 14 MATE on all machines.
Desktop/Gaming: E8400 / 650ti + Accelero S1 / Antec Solo
Myth Backend + Home Server: D945GSEJT / WD15EADS / Hauppage 950Q / Home made case
Myth Frontend + HTPC: Sempron 140 / Asus M4A785-M
Notebook:AMD A4-3300M

runeks
Posts: 26
Joined: Wed Jun 18, 2008 1:07 am
Location: Denmark

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by runeks » Wed Mar 23, 2011 3:06 am

OK. Just to get this straight. I'm going to buy some diodes, as I'm not confident that one of the 0.5W diodes that I already have, won't have a slightly different Zener voltage than the others.

If I want to be sure that the 5.1V Zener diodes that I buy can handle a fan that uses 0.26A at 12V - even in case it stalls due to some cable inside my case getting into the blades or whatever - what should the power rating of the Zener diode be?

Is it that which is calculated by the formula you posted below?
andymcca wrote:Pdiode = Vdiode * Cfan * (12V - Vdiode) / 12V
Without knowing better, I'd think the formula would just be Vdiode * Cfan, since I've read somewhere that Zener diodes only cause a voltage drop, and not a drop in the current. But again, I suck at electrical theory, so I'm just asking this to be absolutely sure.

I can see that it makes sense that the current would drop too, since - according to Ohm's law (I=V/R) - the resistance of the fan in question would be 46Ω (12V/0.26A). I presume this resistance of the fan is a constant, and if it is, the current would have to change too when the voltage changes, in order for the equation I=V/R to be true.

andymcca
Posts: 404
Joined: Mon Jan 11, 2010 8:19 am
Location: Boston, MA, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by andymcca » Wed Mar 23, 2011 4:43 am

For DC fans you can treat them as a resistance (ie no capacitance or inductance), and you are correct that Ohm's law applies. Since you know you are reducing the voltage by 5.1V, you would expect the current to drop proportionally (V = I*R, R being constant).

If you wanted to be super safe, just grabbing a 5W diode will work fine, since it's really just a question of heat. If you found some wild 1kW Zener it might not operate well at 1W, but higher ratings are fine within reason!

Edit: Also, I realized: Does the fan mate have active components? I didn't realize if claims to output 5-12V, which sounds like it has a vreg? I may have been wrong about power consumption above.
I just ordered, one, I will open it up if no one answers!
Currently running Mint 14 MATE on all machines.
Desktop/Gaming: E8400 / 650ti + Accelero S1 / Antec Solo
Myth Backend + Home Server: D945GSEJT / WD15EADS / Hauppage 950Q / Home made case
Myth Frontend + HTPC: Sempron 140 / Asus M4A785-M
Notebook:AMD A4-3300M

runeks
Posts: 26
Joined: Wed Jun 18, 2008 1:07 am
Location: Denmark

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by runeks » Thu Mar 24, 2011 2:32 am

andymcca wrote:For DC fans you can treat them as a resistance (ie no capacitance or inductance), and you are correct that Ohm's law applies. Since you know you are reducing the voltage by 5.1V, you would expect the current to drop proportionally (V = I*R, R being constant).

If you wanted to be super safe, just grabbing a 5W diode will work fine, since it's really just a question of heat. If you found some wild 1kW Zener it might not operate well at 1W, but higher ratings are fine within reason!

Edit: Also, I realized: Does the fan mate have active components? I didn't realize if claims to output 5-12V, which sounds like it has a vreg? I may have been wrong about power consumption above.
I just ordered, one, I will open it up if no one answers!
Cool. Thanks for clarifying. I just ordered some 1.3W diodes, as they seems to be the only ones that are available here (except for 0.5W diodes).

I am noticing something odd though, with to the fan that has the diodes attached to it. After the machine has been shut off, and it's turned on again, it takes about 2 minutes for this fan to spin up to the speed it should be spinning at, at the set voltage. I have the fan in question set to 12V in the BIOS, which should mean that it's at ~7V, but the first two minutes the fan spins at ~450 RPM, instead of the ~900 RPM that I figure it should spin at at ~7V. It could also be the actual detection of the RPM rate that's at fault, because the rotational speed the first two minutes seems to be just around one half of the speed that is later reported, and one half of the speed I've measured before at that voltage.

I have attached a graph of the rotational speeds of the fans. The X-axis is the number of seconds since logging in, which is about 30 seconds after pressing the power button.

Edit: I've updated the graph to include descriptions of the three plots. AUX1 is the exhaust fan, AUX2 is the fan attached to the CPU heat sink (in line with the exhaust fan), AUX4 is the fan in the lower chamber of the case (Antec P182) which has the diodes attached to it.
You do not have the required permissions to view the files attached to this post.
Last edited by runeks on Thu Mar 24, 2011 9:36 am, edited 1 time in total.

andymcca
Posts: 404
Joined: Mon Jan 11, 2010 8:19 am
Location: Boston, MA, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by andymcca » Thu Mar 24, 2011 6:16 am

What do the different colors represent?

That is definitely some odd behavior... Does it occur in the absence of the diode?
Currently running Mint 14 MATE on all machines.
Desktop/Gaming: E8400 / 650ti + Accelero S1 / Antec Solo
Myth Backend + Home Server: D945GSEJT / WD15EADS / Hauppage 950Q / Home made case
Myth Frontend + HTPC: Sempron 140 / Asus M4A785-M
Notebook:AMD A4-3300M

Plekto
Posts: 398
Joined: Tue Feb 19, 2008 2:08 pm
Location: Los Angeles

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by Plekto » Thu Mar 24, 2011 7:19 am

That's likely the motherboard increasing the rpm as its components heat up. The idea being that since stock coolers were all downward-flowing, turning on the CPU fan to a higher rate would cool the board as well. Myself, I just hard-wired it to ~9V straight from the PSU.

runeks
Posts: 26
Joined: Wed Jun 18, 2008 1:07 am
Location: Denmark

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by runeks » Thu Mar 24, 2011 9:48 am

andymcca wrote:What do the different colors represent?

That is definitely some odd behavior... Does it occur in the absence of the diode?
I edited the post to show what the different plots are. Since the fan connector is in a bit of an odd place on the mobo, I'm waiting to try it out with the diodes detached. I'll post when I try it out.
Plekto wrote:That's likely the motherboard increasing the rpm as its components heat up. The idea being that since stock coolers were all downward-flowing, turning on the CPU fan to a higher rate would cool the board as well. Myself, I just hard-wired it to ~9V straight from the PSU.
DOH! I just realized what the cause of this is while writing a reply to the above comment.

All the fans are set to use the SYS (system temperature) as the reference temperature - the temperature they use to decide which voltage to apply to the fan. But since I'd like to run the fan in the lower chamber (AUX4) at 12V (to effectively achieve 7V), I have chosen to use the lowest possible values for the minimum temperature and the maximum temperature which is 25 and 26 degrees Celsius, respectively. When the SYS temperature is 25 degrees or below it's at 8V (effectively 3V) and when it's at 26 degrees or above it's at 12V (7V effectively). But of course this isn't low enough to make the fan always use 12V, I didn't realize this until now, I just chose the lowest possible values.
So to put it shortly, what we're seeing is the time it takes for my system to heat up to 26 degrees Celsius. Oops, my bad... :) everything is actually working as it should. :)

andymcca
Posts: 404
Joined: Mon Jan 11, 2010 8:19 am
Location: Boston, MA, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by andymcca » Sun Mar 27, 2011 6:25 pm

Just a little update on my end:
The Fan Mate 2 arrived, and appears to have been built in a way that makes it difficult to open. I don't have my smallest tools handy at the moment, but I may end up cutting it apart to look inside.

From the outside I see a heatsink taking up most of the enclosure. The wacky thing is that the bottom of the heatsink appears to be floating in air (other than the screw holding it in position). I'm not sure if its connected to something on the end with the pot, or if maybe a device is inside it?

Edit: I am pretty sure after looking at it some more that its a TO-220 package or similar bent parallel to the board, with the heatsink screwed to it. (Which makes it almost definitely a vreg). Also, printed on the side is an "Output: 5V - 11V", which definitely sounds like they are including a transistor drop for the max output :)
Currently running Mint 14 MATE on all machines.
Desktop/Gaming: E8400 / 650ti + Accelero S1 / Antec Solo
Myth Backend + Home Server: D945GSEJT / WD15EADS / Hauppage 950Q / Home made case
Myth Frontend + HTPC: Sempron 140 / Asus M4A785-M
Notebook:AMD A4-3300M

Tephras
Posts: 1140
Joined: Tue Sep 07, 2004 11:03 am
Location: Europe

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by Tephras » Sun Mar 27, 2011 7:55 pm

FWIW, the inside of the Fan Mate 1 can be seen here.

andymcca
Posts: 404
Joined: Mon Jan 11, 2010 8:19 am
Location: Boston, MA, USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by andymcca » Mon Mar 28, 2011 4:35 am

Tephras wrote:FWIW, the inside of the Fan Mate 1 can be seen here.
Thanks! Yeah, that looks just like the FM2 heatsink I'm seeing.
Currently running Mint 14 MATE on all machines.
Desktop/Gaming: E8400 / 650ti + Accelero S1 / Antec Solo
Myth Backend + Home Server: D945GSEJT / WD15EADS / Hauppage 950Q / Home made case
Myth Frontend + HTPC: Sempron 140 / Asus M4A785-M
Notebook:AMD A4-3300M

DrCR
Posts: 538
Joined: Sun Jun 13, 2004 11:55 am

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by DrCR » Thu Mar 31, 2011 5:02 am

I would recommend running the fan(s) from a 4-pin molex and running the yellow tac wire by itself directly from the fan to the motherboard on it's own 3-pin molex. This way you can play with what you wire up the fan(s) to run at without worrying about the tac wire.

For Slipstream 1200RPM, I like to run two of them together in series, rather than parallel, to run them at 6v.

SHODAN
Posts: 16
Joined: Sat Sep 22, 2007 6:02 am
Location: USA

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by SHODAN » Sun Jun 19, 2011 1:02 am

On the IP35 Pro, all the fans can be run down to 6 volts, but it requires a trick.

You cannot do this through BIOS; you must use the μGuru Windows utility.

Go to Setup, Fan EQ, and then select the fan header. Check "FanEQ Enable" if it is not already, and then click "Optimize." The software will spin down the fan, and if it operates at 6 volts, the "Low" limit will magically be changed from 8.0V to 6.0V.
Last edited by SHODAN on Wed Jun 22, 2011 1:15 am, edited 1 time in total.

runeks
Posts: 26
Joined: Wed Jun 18, 2008 1:07 am
Location: Denmark

Re: The "7V trick" on fans with adjustable voltage - RPM gon

Post by runeks » Tue Jun 21, 2011 6:13 am

Here's an update from me. I finally got around to attaching each of the three 5.1V Zener diodes to a fan extension cable, and mounting them in my case. Ah! What a difference! When I spin down the HDDs, I have a hard time hearing if the computer is on when lying in my bed where my head is ~2 meters from the case, and there are no objects between me and the case.

As the RPM is reported in 60 RPM increments, which is quite a lot when the fan is moving at ~500 RPM, I created a little script that reads the fan speeds every second, and calculates the average over an hour. I did this twice to make sure the measurings we reliable and the second measuring reported almost exactly the same as the first (within 1%). These are the results from the two tries:

Code: Select all

CPU fan:                    523
Upper chamber exhaust fan:  507
Lower chamber fan:          468

CPU fan:                    521
Upper chamber exhaust fan:  501
Lower chamber fan:          469
So it seems the Zener diodes, the fans, or the RPM measuring devices for each fan connector are not quite the same (all fans are SY1225SL12M though), but still within 10% of each other. This is when the BIOS is set to 8.0V, which should result in the fans getting 2.9V each.
Temperatures have risen in my case of course. Before, average idle CPU temperature was around 38C, now it's 45C. GPU was ~35C, is now ~45C and HDDs were around 30C, are now ~35C.

But all in all it works great. I can now control the voltage from ~2.9V up to ~6.9V (though I always have them at the lowest setting) through the BIOS.

Now it's just the hard drives that make the computer audible. I'll be creating some suspension system soon, to see if I can get it completely quiet :).

Cheers!

Post Reply