IP35 Pro and Fan Controls
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
IP35 Pro and Fan Controls
I'm planning to use an Abit IP35 Pro for an upcoming build, but now that it's time to start looking at stuff to buy, I'm trying to make sure that it's still my first choice. While this might not be the best place (I'll ask over at Abit eventually), I trust the people here more than any other board.
I'd like to be able to control the fan speed to drop the idle noise pretty low. Normally, the answer would be SpeedFan, but this build will be used for Linux as well, so that's currently out of the picture. Controlling the fans in BIOS seems to be the best (only?) alternative. However, I have a few questions before I jump in and buy this board:
1: The CPU fan supports PWM speed control, supposedly it also supports 3pin fans via PWM-like low-frequency voltage switching. I'm prepared to get a Scythe PWM fan for the heatsink, but is it needed? What sort of luck have people had with ordinary 3pin fans being controlled by the 4pin header?
2: How are all the other case fans controlled? Voltage? PWM? PWM-for-3pins?
3: Has anyone experienced the clicking in fans that are being controlled by the board?
4: What is the minimum duty-cycle/percentage/voltage that the board allows?
If the fan control just isn't worth it, then the IP35 Pro loses a lot of its advantage over other boards and I'll start doing a bit more looking.
I'd like to be able to control the fan speed to drop the idle noise pretty low. Normally, the answer would be SpeedFan, but this build will be used for Linux as well, so that's currently out of the picture. Controlling the fans in BIOS seems to be the best (only?) alternative. However, I have a few questions before I jump in and buy this board:
1: The CPU fan supports PWM speed control, supposedly it also supports 3pin fans via PWM-like low-frequency voltage switching. I'm prepared to get a Scythe PWM fan for the heatsink, but is it needed? What sort of luck have people had with ordinary 3pin fans being controlled by the 4pin header?
2: How are all the other case fans controlled? Voltage? PWM? PWM-for-3pins?
3: Has anyone experienced the clicking in fans that are being controlled by the board?
4: What is the minimum duty-cycle/percentage/voltage that the board allows?
If the fan control just isn't worth it, then the IP35 Pro loses a lot of its advantage over other boards and I'll start doing a bit more looking.
Re: IP35 Pro and Fan Controls
No problems at all with my 1200rpm SlipstreamWhat sort of luck have people had with ordinary 3pin fans being controlled by the 4pin header?
Voltage - 8 or 12V with nothing in between or below.How are all the other case fans controlled?
I haven't.Has anyone experienced the clicking in fans that are being controlled by the board?
30% for the CPU and 8V for the others. I plan to use in-line resistors to drop the voltage an extra notch or two but, to be honest, even at 8V my S-Flex case fans are usually inaudible.What is the minimum duty-cycle/percentage/voltage that the board allows?
My opinion: I like it but it could be better (0%/0V minimums for a start!) plus I was disappointed to learn it's not Speedfan compatible, but imho it's definitely better than most "enthusiast" boards in this regard. The BIOS control is a major plus, although you need to write down your FanEQ settings as they often get reset when you make a hardware change (presumably a "safety" feature for the hard-of-thinking).If the fan control just isn't worth it, then the IP35 Pro loses a lot of its advantage over other boards and I'll start doing a bit more looking.
Re: IP35 Pro and Fan Controls
Really? That's disappointing. Just two values? Is there any chance that the board scales the voltage between 8 and 12 within the temperature range?blackworx wrote:Voltage - 8 or 12V with nothing in between or below.figment wrote:How are all the other case fans controlled?
Not that I'm doubting you, but now I feel the need to confirm this. I was just reading posts over at uabit.com that were suggesting (very obscurely) that there was a "range" from 12v to 8v. You've got the Pro version correct? Looks like I'll have to register over at uabit.com. It's somewhat troubling.
The wider duty cycle for PWM fans might be enough to make me to buy two PWM fans (CPU and main exhaust). For the others, this is still "acceptable". 8v is about what I would have been shooting for, but I think the best would have been having the board switch between 6v and 10v.blackworx wrote:30% for the CPU and 8V for the others. I plan to use in-line resistors to drop the voltage an extra notch or two but, to be honest, even at 8V my S-Flex case fans are usually inaudible.figment wrote:What is the minimum duty-cycle/percentage/voltage that the board allows?
I'm not disappointed by that. They still haven't gotten a reliable port working in Linux, so I'd stick with the BIOS version only.blackworx wrote: My opinion: I like it but it could be better (0%/0V minimums for a start!) plus I was disappointed to learn it's not Speedfan compatible, but imho it's definitely better than most "enthusiast" boards in this regard.
Thanks for your response. Those were the most direct answers I've been able to find.
I'm gonna give a brief answer cause the former datas are not completely correct.
Using b14 (beta) bios version i got:
- you can tune, in the bios, 8-12V with a step of 0.5V: 8-8.5... while 8V is the minimal and 8.5V the maximal votage regulated with CPU Temperature or SYStem Temperature or else PWM temperature.
- the PWM/CPU is differently regulated: 30%-100% (i'm not sure about 30% but it doesn't be far away from 30%) reugulted by any temperature probe.
Well, I'm not so sure if the PWM can be regulated by any probe other than the CPU's one... but it SURE is can be done with µGuru in Windows.
- you can tune since 6V with µGuru in Windows: 6-6.5V to begin with then... Furthermore you can associate any temperature probe for any fan.
To sum up, if you chose too set up a 6-6.5V regulation with 30-40°C temperature interval: you got 6V when the temperature is at 30°C and 6.5V while the temperature is at 40°C. I cannot gaurantee the way the fan is regulated between those two temperatures.
Of course, once you set up your settings in µGuru those settings are automatically loaded in the bios. I cannot explain the impossiblity to do the same in the bios. I writed to Abit a while back and they didn't dare to answer.
Hope I help and was clear enough,
toki.
Using b14 (beta) bios version i got:
- you can tune, in the bios, 8-12V with a step of 0.5V: 8-8.5... while 8V is the minimal and 8.5V the maximal votage regulated with CPU Temperature or SYStem Temperature or else PWM temperature.
- the PWM/CPU is differently regulated: 30%-100% (i'm not sure about 30% but it doesn't be far away from 30%) reugulted by any temperature probe.
Well, I'm not so sure if the PWM can be regulated by any probe other than the CPU's one... but it SURE is can be done with µGuru in Windows.
- you can tune since 6V with µGuru in Windows: 6-6.5V to begin with then... Furthermore you can associate any temperature probe for any fan.
To sum up, if you chose too set up a 6-6.5V regulation with 30-40°C temperature interval: you got 6V when the temperature is at 30°C and 6.5V while the temperature is at 40°C. I cannot gaurantee the way the fan is regulated between those two temperatures.
Of course, once you set up your settings in µGuru those settings are automatically loaded in the bios. I cannot explain the impossiblity to do the same in the bios. I writed to Abit a while back and they didn't dare to answer.
Hope I help and was clear enough,
toki.
Re: IP35 Pro and Fan Controls
Oops my mistake (I don't have the board hooked up at present as I'm refitting the heatpipe assembly). Checking the manual, it says the hi/lo voltages can actually be set anywhere between 8 and 12V.figment wrote:Really? That's disappointing. Just two values? Is there any chance that the board scales the voltage between 8 and 12 within the temperature range?blackworx wrote:Voltage - 8 or 12V with nothing in between or below.
Unfortunately there's only one PWM header - CPUFAN. The rest (SYSFAN and AUX1-4) are all Voltage controlled.The wider duty cycle for PWM fans might be enough to make me to buy two PWM fans (CPU and main exhaust).
My feelings exactly. If you're up for a bit of soldering you could mod your fans with an inline resistor as I intend to. http://www.blackfiveservices.co.uk/fanspeed.shtml is a handy calculator.I think the best would have been having the board switch between 6v and 10v
NPThanks for your response. Those were the most direct answers I've been able to find.
Yes, on both accounts.toki_c wrote:Hope I help and was clear enough.
In the last day or so, I've heard a number of different things, and I'm beginning to think the BIOS updates and discrepancies between BIOS and the Windows software might be causing a lot of the confusion. I've registered at uabit and I'm going to hit their boards with some of these questions and post the responses here in case other people ever come looking for the same answers.
Alright, here are some answers (which agree with and add to toki's response), and I'm encouraged. These answers are in reference to the Abit IP35 Pro, but I would assume they would apply to other contemporary boards with FanEQ like the IX38 boards. Note that the IP35, IP35-E and IP35-V do not have a full FanEQ implementation and only have controls for the CPU and one other fan (SYS).
The IP35 Pro has one 4-pin CPU header controlled by PWM. If a 3-pin fan is attached, it will try to use approximated voltage. (This is autodetected in IP35 Pro. IP35-E and IP35 have a BIOS setting to control this, but don't have the full FanEQ system). It's recommended to actually get a PWM fan for this header since some people report problems with FanEQ losing control of the CPU header and running it at the full 12v.
There are 5 other 3-pin fan headers which can be controlled via voltage modification (SYS and AUX1 - AUX4). It is possible (likely?) that this is done by smoothing a PWM signal, but the end result is reliable and doesn't seem to have a history of causing problems.
The CPU header allows a range of 30%-100% duty-cycle. The other headers have a range of 8V to 12V at 0.5V increments in BIOS and 6V to 12V if using the Windows uGuru software (Master of None at uabit.com confirms that he does see a lower fan speed when using 6V instead of 8V).
Each fan control is based off a monitored temperature sensor: the CPU temperature, the "ambient" system temperature, and the temperature at the PWM circuitry. The control maps a voltage (or duty-cycle) to use at a low temperature and a high temperature.
At or below the low temperature: The header will use the "low" voltage.
At or above the high temperature: The header will use the "high" voltage.
Between the high and low temperature: The header will use a voltage which roughly scale to the distance between the high and low temperature.
For example:
AUX1 is an exhaust fan.
Monitoring: CPU
Low: 35C = 8V
High: 55C = 12V
CPU Temp = 30C :: AUX1 = 8.0V
CPU Temp = 35C :: AUX1 = 8.0V
CPU Temp = 40C :: AUX1 = 9.0V
CPU Temp = 45C :: AUX1 = 10.0V
CPU Temp = 50C :: AUX1 = 11.0V
CPU Temp = 55C :: AUX1 = 12.0V
CPU Temp = 60C :: AUX1 = 12.0V
(NOTE: This is just a rough illustration, not a declaration of scientific fact. Actual voltages may vary.)
In practice (and hopefully due to actually useful heatpipes), the PWM temperature tends to generally track with the CPU and northbridge (MCH) temperatures, though it is much smoother. Those seeking quiet cases might consider the PWM temperature as the one to base a majority of fans off of in order to keep fan noise smoother and less noticeable.
All of this basically indicates that the IP35 Pro has a pretty sophisticated fan controller on-board (just like Abit claims) and that it should provide enough control to make SPCR users happy. The 8V lower limit is still a bit high, but it's possible that future BIOS revisions will lower the value to the 6V allowed in the Windows utility (which has problems in Vista and no equivalent in Linux). As it stands, at 8V, a 1200rpm Slipstream should run at 800rpm and be nearly inaudible.
DISCLAIMER: I do not have this board (yet). This is information harvested from a number of places and confirmed over at the Abit boards (link to source). Feel free to add corrections and when I finally get the board and start using it, I'll try to update this information to keep it accurate.
The IP35 Pro has one 4-pin CPU header controlled by PWM. If a 3-pin fan is attached, it will try to use approximated voltage. (This is autodetected in IP35 Pro. IP35-E and IP35 have a BIOS setting to control this, but don't have the full FanEQ system). It's recommended to actually get a PWM fan for this header since some people report problems with FanEQ losing control of the CPU header and running it at the full 12v.
There are 5 other 3-pin fan headers which can be controlled via voltage modification (SYS and AUX1 - AUX4). It is possible (likely?) that this is done by smoothing a PWM signal, but the end result is reliable and doesn't seem to have a history of causing problems.
The CPU header allows a range of 30%-100% duty-cycle. The other headers have a range of 8V to 12V at 0.5V increments in BIOS and 6V to 12V if using the Windows uGuru software (Master of None at uabit.com confirms that he does see a lower fan speed when using 6V instead of 8V).
Each fan control is based off a monitored temperature sensor: the CPU temperature, the "ambient" system temperature, and the temperature at the PWM circuitry. The control maps a voltage (or duty-cycle) to use at a low temperature and a high temperature.
At or below the low temperature: The header will use the "low" voltage.
At or above the high temperature: The header will use the "high" voltage.
Between the high and low temperature: The header will use a voltage which roughly scale to the distance between the high and low temperature.
For example:
AUX1 is an exhaust fan.
Monitoring: CPU
Low: 35C = 8V
High: 55C = 12V
CPU Temp = 30C :: AUX1 = 8.0V
CPU Temp = 35C :: AUX1 = 8.0V
CPU Temp = 40C :: AUX1 = 9.0V
CPU Temp = 45C :: AUX1 = 10.0V
CPU Temp = 50C :: AUX1 = 11.0V
CPU Temp = 55C :: AUX1 = 12.0V
CPU Temp = 60C :: AUX1 = 12.0V
(NOTE: This is just a rough illustration, not a declaration of scientific fact. Actual voltages may vary.)
In practice (and hopefully due to actually useful heatpipes), the PWM temperature tends to generally track with the CPU and northbridge (MCH) temperatures, though it is much smoother. Those seeking quiet cases might consider the PWM temperature as the one to base a majority of fans off of in order to keep fan noise smoother and less noticeable.
All of this basically indicates that the IP35 Pro has a pretty sophisticated fan controller on-board (just like Abit claims) and that it should provide enough control to make SPCR users happy. The 8V lower limit is still a bit high, but it's possible that future BIOS revisions will lower the value to the 6V allowed in the Windows utility (which has problems in Vista and no equivalent in Linux). As it stands, at 8V, a 1200rpm Slipstream should run at 800rpm and be nearly inaudible.
DISCLAIMER: I do not have this board (yet). This is information harvested from a number of places and confirmed over at the Abit boards (link to source). Feel free to add corrections and when I finally get the board and start using it, I'll try to update this information to keep it accurate.
Last edited by figment on Fri Apr 18, 2008 5:54 am, edited 1 time in total.
-
- Posts: 1386
- Joined: Fri Jun 18, 2004 6:53 pm
I have IP35Pro, but I still haven't tried using on board fan controls (I'm using sunbeam fan controller), so it's a little funny I should ask this question: does anybody know if Abit IP35Pro jump starts fans at 12V for a couple of seconds to ensure they all spun up before settling down to the voltages set in uGuru?
From what I've gathered in my quest:
For PWM fans (the CPU fan): Yes, the fan is supplied with a higher duty cycle at startup to get it running.
For other fans: No, (probably not?) this is presumed to be the reason for the 8V (or 6V in windows) lower voltage limit. I'd expect the fans would be given the lower voltage level set in FanEQ.
For PWM fans (the CPU fan): Yes, the fan is supplied with a higher duty cycle at startup to get it running.
For other fans: No, (probably not?) this is presumed to be the reason for the 8V (or 6V in windows) lower voltage limit. I'd expect the fans would be given the lower voltage level set in FanEQ.
I totaly agree with your post, figment, I don't use the PWM header because of the reported issue. Sometimes it works just fine else the fan spins at full speed.
Withh 4 3-pin header there's no problem using a 3-pin for the CPU.
Else, I recommend no to leave µGuru run/idl for nothing in Vista cause it can become buggy when the bord is oced and freezes Vista... and a Reset is necessary to... Otherwise µG works just fine. I load µG, tweak the board and load my settings in the bios and then shut down µG. It works fine without any issue.
Withh 4 3-pin header there's no problem using a 3-pin for the CPU.
Else, I recommend no to leave µGuru run/idl for nothing in Vista cause it can become buggy when the bord is oced and freezes Vista... and a Reset is necessary to... Otherwise µG works just fine. I load µG, tweak the board and load my settings in the bios and then shut down µG. It works fine without any issue.
Sure - there's AUXFAN1-4 and SYSFAN. The AUX headers are arranged round the edges of the board (2 next to CPUFAN, 1 next to IDE connector and 1 on bottom edge near last PCI connector). SYSFAN header is next to the NB.
In BIOS FanEQ the 6 fans are split into 2 groups - CPU,SYS,AUX1 in one group and AUX2-4 in the other. I don't think the split is anything other than cosmetic, as each fan has its own reference temp anyway.
In BIOS FanEQ the 6 fans are split into 2 groups - CPU,SYS,AUX1 in one group and AUX2-4 in the other. I don't think the split is anything other than cosmetic, as each fan has its own reference temp anyway.
-
- *Lifetime Patron*
- Posts: 1288
- Joined: Sat Oct 25, 2003 3:21 pm
- Location: 15143, USA
- Contact:
FWIW, here's what I've found with my IP35-E. Note that I've used only the CPU & Sysfan1 connectors. They have been tested with Yate Loon SLs & the Scythe Ninja Rev. B's stock fan.
- Both ports run the fans at full speed until POST. No problems spinning up.
- No issues experienced with regular 3-pin fans, and I've heard no unusual clicking.
- Both allow settings down to "30%", but I don't know how low that really runs. This was useful for quieting the system while running Memtest, but I've used Speedfan since installing Windows.
- Both ports work flawlessly with Speedfan, but only if FanEQ is disabled. So my system's loud until Windows boots, which makes it hard to impress people with my PC's noise level...
- The board reliably reports RPM down below 400rpm.
- Both ports run the fans at full speed until POST. No problems spinning up.
- No issues experienced with regular 3-pin fans, and I've heard no unusual clicking.
- Both allow settings down to "30%", but I don't know how low that really runs. This was useful for quieting the system while running Memtest, but I've used Speedfan since installing Windows.
- Both ports work flawlessly with Speedfan, but only if FanEQ is disabled. So my system's loud until Windows boots, which makes it hard to impress people with my PC's noise level...
- The board reliably reports RPM down below 400rpm.
Your hard work is for people impressions... or for your pleasure?HammerSandwich wrote:...which makes it hard to impress people with my PC's noise level...
With my hard work case insulation with damping material for the case, vents, 3.5" and 5.25" bays... I heard almost nothing when powering up my rig. Noises before POST are really lowered.
Your hard work is for people belief... or for your ears pleasure?HammerSandwich wrote:...which makes it hard to impress people with my PC's noise level...
With my hard work case insulation with damping material for the case, vents, 3.5" and 5.25" bays... I heard almost nothing when powering up my rig. Noises before POST are really lowered.
-
- *Lifetime Patron*
- Posts: 1288
- Joined: Sat Oct 25, 2003 3:21 pm
- Location: 15143, USA
- Contact:
I just want to clarify a couple points here: the IP35 pro is THE MOBO from a silencer point of view:
1 4pin CPU fan socket
1 3pin SYS fan socket
4 3pin AUX fan sockets
I dont use the 4pin one. As described above, the others are regulable from 8 to 12 V in the bios, and you can link them to CPU, PWM or ambient temps.
HOWEVER:
There is a uGURU software that runs in winXP (I have a partition just for this and games ). You run it, clic on fanEQ and go to the fan of choice. Then you clic on Optimize, and the program changes the speed to find out the range of the fan (starting voltage-max voltage).
THEN (if the fan allows) it lowers the minimum voltage to 6V AND STORES THE VALUE IN THE BIOS. End of it. No more problem. You close the uGURU program and the BIOS does the job. No more controllers needed. Nothing. No uGURU, no speedfan. I have a 800 rpm noctua case fan now linked to my CPU (E6750, ninja fanless, 40-45ºC normally) running at 480 RPMs. Nice to have the bios taking care of it, because uGURU does not run in Ubuntu 8.04, my actual OS.
This works with the original BIOS (10) and the actual stable one (14). IT DOES NOT with others I tried before. More info:
http://forum.abit-usa.com/showthread.php?t=124502
1 4pin CPU fan socket
1 3pin SYS fan socket
4 3pin AUX fan sockets
I dont use the 4pin one. As described above, the others are regulable from 8 to 12 V in the bios, and you can link them to CPU, PWM or ambient temps.
HOWEVER:
There is a uGURU software that runs in winXP (I have a partition just for this and games ). You run it, clic on fanEQ and go to the fan of choice. Then you clic on Optimize, and the program changes the speed to find out the range of the fan (starting voltage-max voltage).
THEN (if the fan allows) it lowers the minimum voltage to 6V AND STORES THE VALUE IN THE BIOS. End of it. No more problem. You close the uGURU program and the BIOS does the job. No more controllers needed. Nothing. No uGURU, no speedfan. I have a 800 rpm noctua case fan now linked to my CPU (E6750, ninja fanless, 40-45ºC normally) running at 480 RPMs. Nice to have the bios taking care of it, because uGURU does not run in Ubuntu 8.04, my actual OS.
This works with the original BIOS (10) and the actual stable one (14). IT DOES NOT with others I tried before. More info:
http://forum.abit-usa.com/showthread.php?t=124502
On a related note: does anyone know how much power each of the Voltage-controlled headers are likely to have available? The info is not in the specs or elsewhere online. The reason I ask is that I'm toying with the idea of connecting two 1200rpm S-Flexes in parallel to one header for an experimental push-pull arrangement.
I've done a bit of measuring and the peak power required by a pair of these fans is about 3W (@12V) at start-up, dropping to half that after a second or so. I won't be running them at 12V, so I'm assuming there should be plenty of headroom, but don't want to go bungling in and destroy my mobo.
I've done a bit of measuring and the peak power required by a pair of these fans is about 3W (@12V) at start-up, dropping to half that after a second or so. I won't be running them at 12V, so I'm assuming there should be plenty of headroom, but don't want to go bungling in and destroy my mobo.
I don't have any numbers, but I think you'll be fine. I've heard stories about people chaining push/pull Arctic Fans without any problems. One mentioned they'd even chained three together (as Arctic Fans make this simple).
Slipstreams and S-FLEXs seem pretty efficient. I only think you'd have problems if you were trying to run Deltas or high-speed Sunons. Some of those are pulling 0.6 amps, twice what the Slipstreams do.
Slipstreams and S-FLEXs seem pretty efficient. I only think you'd have problems if you were trying to run Deltas or high-speed Sunons. Some of those are pulling 0.6 amps, twice what the Slipstreams do.
-
- Posts: 1386
- Joined: Fri Jun 18, 2004 6:53 pm
Hold on a second, are you saying that fan speed control will not work with any BIOS other than 10 and 14?fjf wrote:This works with the original BIOS (10) and the actual stable one (14). IT DOES NOT with others I tried before. More info:
http://forum.abit-usa.com/showthread.php?t=124502
No.
I think he was specifically talking about the "Optimize" feature for the uGuru Windows software, and probably just for Windows Vista (possibly just Vista/64)
From what I can tell, "Optimize" tries to figure out how much lower it can drop the fan voltage without causing the fan to stall. For most quiet fans, this is pretty low and you end up getting the "second" minimum value of 6.0V (even though Slipstreams and S-Flex's can go down to 5.0V and lower, I think).
Even if you don't use the Windows software, you can still configure FanEQ in BIOS and get the 8.0V to 12.0V range. The one thing you can't do in BIOS is perform the "Optimize" function to drop the voltages lower than 8.0V. For that, you can run the uGuru tool once, set the minimum and shut it down. It has the ability to run in the background much like SpeedFan, but you don't need to keep it running to get the fan control. That's the whole point of fan control in BIOS. This also means that you don't need to be running Windows to get the advanced fan control.
Also, note that the current BIOS revision for this board is actually 16, not 14. This is the revision which should be shipping on most new boards and it has better support for Penryns. There is a beta BIOS 17 which a number of people have also switched to for better backward support of Conroe chips. All versions support the full FanEQ system and allow control of 6 fans.
I think he was specifically talking about the "Optimize" feature for the uGuru Windows software, and probably just for Windows Vista (possibly just Vista/64)
From what I can tell, "Optimize" tries to figure out how much lower it can drop the fan voltage without causing the fan to stall. For most quiet fans, this is pretty low and you end up getting the "second" minimum value of 6.0V (even though Slipstreams and S-Flex's can go down to 5.0V and lower, I think).
Even if you don't use the Windows software, you can still configure FanEQ in BIOS and get the 8.0V to 12.0V range. The one thing you can't do in BIOS is perform the "Optimize" function to drop the voltages lower than 8.0V. For that, you can run the uGuru tool once, set the minimum and shut it down. It has the ability to run in the background much like SpeedFan, but you don't need to keep it running to get the fan control. That's the whole point of fan control in BIOS. This also means that you don't need to be running Windows to get the advanced fan control.
Also, note that the current BIOS revision for this board is actually 16, not 14. This is the revision which should be shipping on most new boards and it has better support for Penryns. There is a beta BIOS 17 which a number of people have also switched to for better backward support of Conroe chips. All versions support the full FanEQ system and allow control of 6 fans.
Could you do me a favour and test the other fan headers, and how many fan control channels pops up in SpeedFan? (a screenshot would be awesome)HammerSandwich wrote:FWIW, here's what I've found with my IP35-E. Note that I've used only the CPU & Sysfan1 connectors. They have been tested with Yate Loon SLs & the Scythe Ninja Rev. B's stock fan.
- Both ports run the fans at full speed until POST. No problems spinning up.
- No issues experienced with regular 3-pin fans, and I've heard no unusual clicking.
- Both allow settings down to "30%", but I don't know how low that really runs. This was useful for quieting the system while running Memtest, but I've used Speedfan since installing Windows.
- Both ports work flawlessly with Speedfan, but only if FanEQ is disabled. So my system's loud until Windows boots, which makes it hard to impress people with my PC's noise level...
- The board reliably reports RPM down below 400rpm.
.. Im looking for a new mobo, but im not sacrificing speedfan control.
-
- *Lifetime Patron*
- Posts: 1288
- Joined: Sat Oct 25, 2003 3:21 pm
- Location: 15143, USA
- Contact:
Right. The Optimize button in the uGuru software (I've tried only winXP, not Vista) lowers the minimum voltage from 8 (the default BIOS minimum) to 6, and then stores it in the BIOS, so it'll work even in linux, without the uGuru soft running. The BIOS alone wont let you do that. And This works in 10 and 14 BIOS versions; others that I tried don't. I haven't tried the 16 (last I checked was still beta), but unless you need it because you have a very recent CPU, the 14 works very well. Maybe someone with the 16 loaded can tell us if this works with it.
figment wrote:No.
I think he was specifically talking about the "Optimize" feature for the uGuru Windows software, and probably just for Windows Vista (possibly just Vista/64)
From what I can tell, "Optimize" tries to figure out how much lower it can drop the fan voltage without causing the fan to stall. For most quiet fans, this is pretty low and you end up getting the "second" minimum value of 6.0V (even though Slipstreams and S-Flex's can go down to 5.0V and lower, I think).
Even if you don't use the Windows software, you can still configure FanEQ in BIOS and get the 8.0V to 12.0V range. The one thing you can't do in BIOS is perform the "Optimize" function to drop the voltages lower than 8.0V. For that, you can run the uGuru tool once, set the minimum and shut it down. It has the ability to run in the background much like SpeedFan, but you don't need to keep it running to get the fan control. That's the whole point of fan control in BIOS. This also means that you don't need to be running Windows to get the advanced fan control.
Also, note that the current BIOS revision for this board is actually 16, not 14. This is the revision which should be shipping on most new boards and it has better support for Penryns. There is a beta BIOS 17 which a number of people have also switched to for better backward support of Conroe chips. All versions support the full FanEQ system and allow control of 6 fans.
I've started my build and I'll be getting the last parts today. With luck, I'll be able to fire up BIOS 16 this weekend and I'll let everyone know.fjf wrote:I haven't tried the 16 (last I checked was still beta), but unless you need it because you have a very recent CPU, the 14 works very well. Maybe someone with the 16 loaded can tell us if this works with it.
I'm really hoping I'll be able to get the voltages down to 6V.
figment, et al.,
I own the Abit IP 35 Pro. I've got it in an Antec P182 case. I am interested in a mild overclock for the q6600 g0 processor I'm using (with 8 Gb of ram - 4 x 2 Gb, DDR2, G.Skill). My video card is an evga 8800gtx.
For cooling I have a Thermalright Ultima 90 on the cpu and a Thermalright HR-03plus mounted on the video card. (The HR-03 is backward mounted: the heat pipes and cooling fins are on top of the card - they are closer to the top of the case. The card has all the chips on the bottom.)
I have 7 fans blowing in it, all on their own header.
CPU fan, 3 pin, to MB CPU header. A nexus 92mm. (I may eventually swap to a PWM fan).
Video card fan, nexus 92mm PWM real silent case fan mounted to HR-03plus, profile controlled by sidewinder's RivaTuner v2.08. (30% default start speed)
I'm using two Noctua 120mm fans for pullling air in. One in the top, MB side, of the case, the other in the bottom PSU part of the the case. These are attached to the AUX 3 and AUX 4 headers, respecively.
I have two Nexus 120mm fans exhausting the case. One in the top fan mount, the other in the top, rear, position. They are connected to AUX 1 and AUX 2.
Finally, I have a nexus 92mm fan mounted on a homebuilt bracket aimed directly at the northbridge. (3/4" aluminum "L", so that the fan is 45 degrees or so angled away from the side of the case. Plenty of rubber isolators and nylon lock nuts. This is specifically to prevent a dead air space near the MCH.) This is connected to the SYS header.
That's the 7 fans. They're quiet, but not silent. Outside sounds are louder than the computer (birds, cars starting at neighbor houses, etc.).
Also, I've performed the bolt mod to the MB's heat pipe, with Arctic Ceramique.
Using Vista64 Ultimate, uGuru crashes. I have not run the optimize function.
I am curious about which of the three MB temp sensors, CPU, PWM, SYS, most accurately tracks northbridge (MCH) temps.
Let me know what you come up with.
Good luck,
Ken
I own the Abit IP 35 Pro. I've got it in an Antec P182 case. I am interested in a mild overclock for the q6600 g0 processor I'm using (with 8 Gb of ram - 4 x 2 Gb, DDR2, G.Skill). My video card is an evga 8800gtx.
For cooling I have a Thermalright Ultima 90 on the cpu and a Thermalright HR-03plus mounted on the video card. (The HR-03 is backward mounted: the heat pipes and cooling fins are on top of the card - they are closer to the top of the case. The card has all the chips on the bottom.)
I have 7 fans blowing in it, all on their own header.
CPU fan, 3 pin, to MB CPU header. A nexus 92mm. (I may eventually swap to a PWM fan).
Video card fan, nexus 92mm PWM real silent case fan mounted to HR-03plus, profile controlled by sidewinder's RivaTuner v2.08. (30% default start speed)
I'm using two Noctua 120mm fans for pullling air in. One in the top, MB side, of the case, the other in the bottom PSU part of the the case. These are attached to the AUX 3 and AUX 4 headers, respecively.
I have two Nexus 120mm fans exhausting the case. One in the top fan mount, the other in the top, rear, position. They are connected to AUX 1 and AUX 2.
Finally, I have a nexus 92mm fan mounted on a homebuilt bracket aimed directly at the northbridge. (3/4" aluminum "L", so that the fan is 45 degrees or so angled away from the side of the case. Plenty of rubber isolators and nylon lock nuts. This is specifically to prevent a dead air space near the MCH.) This is connected to the SYS header.
That's the 7 fans. They're quiet, but not silent. Outside sounds are louder than the computer (birds, cars starting at neighbor houses, etc.).
Also, I've performed the bolt mod to the MB's heat pipe, with Arctic Ceramique.
Using Vista64 Ultimate, uGuru crashes. I have not run the optimize function.
I am curious about which of the three MB temp sensors, CPU, PWM, SYS, most accurately tracks northbridge (MCH) temps.
Let me know what you come up with.
Good luck,
Ken