[SOLVED] Silent Base 802 fan controller rotates fans at max speed when using AUTO (PWM) mode

[Platinum_Gamer]

Hi,

Just finished a build with the new bequiet Silent Base 802 case and a ASRock Z370m Pro4 motherboard.

On top of the case there is a fan controller with speeds 3 to 1 and AUTO which is suppose to use the PWM signal from the motherboard's fan header.

I have connected the fan controller to CHASSIS_FAN2 (4pin) header of my motherboard.

This is my current fan profile for CHASSIS_FAN2

I'm expecting that when I'm using the AUTO mode, the fans should spin at around half speed since the motherboard's temperature is around 30C. However currently they are spinning at ~max speed all the time. Using the manuals speeds 3-1 works fine, speed increases/decreases depending on the number.

Anyone know why this is happening? I'm suspecting my little ASRock Z370m Pro4's 4pin CHASSIS_FAN2 header doesn't actually support PWM.

Thanks

 

[Paperdoc]

Actually, the problem is the other way around. The Pure Wings 2 fans supplied with that case are the older 3-pin Voltage Controlled fans, NOT their newer Pure Wings 2 PWM models. If you connect any 3-pin fan to a header using the new PWM Mode to control speed, that fan will always run at full speed. A PWM signal system cannot control its speed. Apparently the three-speed button system built into the case does it the old way - it changes the VOLTAGE sent to the fan - and that works with that fan design. Although the case system allows the option to let a PWM signal from a mobo header be fed to an attached fan, that fan would have to be of the new PWM design in order for that control system to work.

Edit: I did not think to point this out earlier. You can NOT make this work by having the mobo header configured to send a VOLTAGE Control Mode signal (aka DC Mode) to the hub. I am sure that Hub can do only what any simple PWM Fan Hub can do, which is to relay the PWM signal it gets from a header to its fans. The Hub you have certainly does NOT translate a VOLTAGE signal from a mobo header into anything.

 

[Platinum_Gamer]

Actually, the problem is the other way around. The Pure Wings 2 fans supplied with that case are the older 3-pin Voltage Controlled fans, NOT their newer Pure Wings 2 PWM models. If you connect any 3-pin fan to a header using the new PWM Mode to control speed, that fan will always run at full speed. A PWM signal system cannot control its speed. Apparently the three-speed button system built into the case does it the old way - it changes the VOLTAGE sent to the fan - and that works with that fan design. Although the case system allows the option to let a PWM signal from a mobo header be fed to an attached fan, that fan would have to be of the new PWM design in order for that control system to work.

Edit: I did not think to point this out earlier. You can NOT make this work by having the mobo header configured to send a VOLTAGE Control Mode signal (aka DC Mode) to the hub. I am sure that Hub can do only what any simple PWM Fan Hub can do, which is to relay the PWM signal it gets from a header to its fans. The Hub you have certainly does NOT translate a VOLTAGE signal from a mobo header into anything.

Right, so the PWM mode in the fan controller is effectively useless then? (because the attached Pure Wings 2 aren't PWM compatible)

 

[Karadjgne]

That fan hub supposedly can. If set to Auto it uses the pwm signal to adjust the Sata supplied voltage. If using the actual 1-3 speed settings, those are switch controlled Sata power not motherboard header supplied/controlled power.

That's why it cannot be set for DC in the bios, as it relies on pwm to moderate external voltages, it doesn't use the motherboard supplied voltages through the header.

You'll want to use Bios set for PWM, not Auto. Auto will detect no pwm signal change, so will attempt DC voltage instead, which doesn't change anything either as that's not how the hub works.

So physically changing 1-3 moderates the Sata power voltages, speeds get reported by the tach sensor wire to the motherboard. On Auto, the pwm signal from the motherboard changes the Sata power, speeds get reported back to the header by the tach wire. The hub doesn't use the 12v wire from the header, only tach/ pwm/ possibly ground.

 

[Platinum_Gamer]

That fan hub supposedly can. If set to Auto it uses the pwm signal to adjust the Sata supplied voltage. If using the actual 1-3 speed settings, those are switch controlled Sata power not motherboard header supplied/controlled power.

That's why it cannot be set for DC in the bios, as it relies on pwm to moderate external voltages, it doesn't use the motherboard supplied voltages through the header.

You'll want to use Bios set for PWM, not Auto. Auto will detect no pwm signal change, so will attempt DC voltage instead, which doesn't change anything either as that's not how the hub works.

So physically changing 1-3 moderates the Sata power voltages, speeds get reported by the tach sensor wire to the motherboard. On Auto, the pwm signal from the motherboard changes the Sata power, speeds get reported back to the header by the tach wire. The hub doesn't use the 12v wire from the header, only tach/ pwm/ possibly ground.

Ahh, I think I get what you mean. However, I do not have an explicit PWM mode in my BIOS for any of my fan headers.

I have:

• Customize (which I'm currently using, sets temperature and fan speed steps)
• Standard
• Silent
• Performance
• Full speed

I'm not sure if any of those modes provide the PWM signal required. They may all be using voltage modulation. I'll try one of the presets and see what happens

Edit: Using the silent preset on CHASSIS_FAN2 and still the case fans spin at 100% on AUTO mode. Does this suggest that my fan headers aren't actually providing a PWM signal?

 

[Platinum_Gamer]

Connecting the fan hub to my 4pin CPU FAN1 header seems to make the AUTO mode actually work. Now my H115i's tach is sitting on CPU FAN2 and its giving me an inaccurate reading of 4000rpm at the pump.

So, I guess the 4pin chassis fan headers don't support PWM which was my original theory. This is all very confusing, what's the point of having a 4pin chassis fan header if it isn't for PWM??

 

[Paperdoc]

I'm surprised that Karadjgne says the case's fan hub CAN "translate" from PWM to Voltage Control Mode. That is not a common feature is basic fan hubs, but it certainly can be done. Moreover, your last post, OP, says it appears to work that way when connected to your CPU_FAN header.

Almost all mobo fan headers now use a 4-pin configuration. Ideally, and in most modern boards, that means the same header CAN be configured to deliver EITHER 3-pin Voltage Control Mode signals, or 4-pin PWM Mode signals, so the board can be used with either type of fan. However, in earlier days as PWM fan systems were introduced, various board makers implemented the 4-pin header systems in several ways:
(a) 3-pin Voltage Control Mode only;
(b) 4-pin PWM Mode only;
(c) both modes available, with choices made manually in BIOS Setup;
(d) both modes available, plus an "Auto" option that was supposed to test the fan connected at each start-up and adjust the header to what was present.
I note also that several used Option (b) above for the CPU_FAN header only, and Option (a) for all CHA_FAN headers.

Now, 4-pin fans are designed with backwards compatibility features. One of them is that it WILL operate properly under the older Voltage Control Mode, even though that is not ideal. Consider what happens when that fan is connected to a real 3-pin header, OR to a 4-pin header using the older Mode. It receives NO PWM signal from Pin #4, so the fan's internal chip cannot modify the power flow from the power supply line (Pin #2) through the windings; the power available is just passed on unchanged. But that source is a VARYING voltage (as a 3-pin fan would require), not a fixed 12 VDC, so the fan speed IS controlled. So NOTE an easy "Fake Auto" method: use option (a) above and the header does ONLY 3-pin Mode. This DOES control the speed of either fan type connected there. The only way you find out the trick is when you try to use a fan Hub that requires a PWM signal from Pin #4, and it does NOT work because there is no signal there! OP, since your mobo manual does NOT say anything about options for the header MODE, it is entirely possible that they used Option (a) above, or the two-option version I noted below that list.

The two-option combo appears to be what yours does: a PWM signal seems present from the CPU_FAN header, but absent from the CHA_FAN headers. However, that is NOT what the labels in your manual say about those headers. The only real way to check what they are doing is to use two fans - one 3-pin, one 4-pin - and test each header. Plug one fan into the header, and use settings in that header's configuration in BIOS Setup to tell it to change speeds. A header using the older 3-pin Voltage control Mode CAN change the speed of either fan type. A header using the new PWM Mode can change the speed of only a 4-pin fan, and will cause a 3-pin fan to run full speed no matter how it is set. If you really need to know whether a header with 4 pins may NOT be sending out a PWM signal, then you need to do a test with a fan Hub of known design (i.e., one that can only send out the PWM signal from a header, and supplies the +12 VDC power solely from its PSU connection with NO DC power drawn from the mobo header) in combination with those two test fans. In that connection system, the fans will always receive the full +12 VDC supply on Pin #2 from the Hub, and can only receive a PWM signal for speed control IF the mobo header supplies it. So a 3-pin fan in this scheme will always run full speed, and a 4-pin fan will change its speed ONLY if the mobo header is supplying the PWM signal (and hence IS using the new PWM Mode).

 

[Karadjgne]

It's the external power requirement. That hub is pretty neat, it's got an Auto setting to allow a pwm sent signal to control upto 6x 3pin fans. Normally, 3pin fans you get 3x, maybe 4x, before hitting amperage limits from a header. 6x fans on a non header power source. Nice. You basically take a pwm signal and use it to control a rheostat.

Doing it manually is easy, it's just manually shifting power through 3 resistors to lower output voltage.

Op, those are Fan curve settings you are looking at, they control how the fans respond to the sensors. What you want is the the header setting, which will be Auto, pwm or dc/voltage. Might be found under Monitor tab or such, look for anything that says cpu_fan or chassis_fan etc.

4000rpm for a pump is normal range for many of those pumps, it'll be max. If your pump has a USB cable attachment it means it's waiting on you to install iCue software and control is initiated there, not in bios.

 

[Platinum_Gamer]

I'm surprised that Karadjgne says the case's fan hub CAN "translate" from PWM to Voltage Control Mode. That is not a common feature is basic fan hubs, but it certainly can be done. Moreover, your last post, OP, says it appears to work that way when connected to your CPU_FAN header.

Almost all mobo fan headers now use a 4-pin configuration. Ideally, and in most modern boards, that means the same header CAN be configured to deliver EITHER 3-pin Voltage Control Mode signals, or 4-pin PWM Mode signals, so the board can be used with either type of fan. However, in earlier days as PWM fan systems were introduced, various board makers implemented the 4-pin header systems in several ways:
(a) 3-pin Voltage Control Mode only;
(b) 4-pin PWM Mode only;
(c) both modes available, with choices made manually in BIOS Setup;
(d) both modes available, plus an "Auto" option that was supposed to test the fan connected at each start-up and adjust the header to what was present.
I note also that several used Option (b) above for the CPU_FAN header only, and Option (a) for all CHA_FAN headers.

Now, 4-pin fans are designed with backwards compatibility features. One of them is that it WILL operate properly under the older Voltage Control Mode, even though that is not ideal. Consider what happens when that fan is connected to a real 3-pin header, OR to a 4-pin header using the older Mode. It receives NO PWM signal from Pin #4, so the fan's internal chip cannot modify the power flow from the power supply line (Pin #2) through the windings; the power available is just passed on unchanged. But that source is a VARYING voltage (as a 3-pin fan would require), not a fixed 12 VDC, so the fan speed IS controlled. So NOTE an easy "Fake Auto" method: use option (a) above and the header does ONLY 3-pin Mode. This DOES control the speed of either fan type connected there. The only way you find out the trick is when you try to use a fan Hub that requires a PWM signal from Pin #4, and it does NOT work because there is no signal there! OP, since your mobo manual does NOT say anything about options for the header MODE, it is entirely possible that they used Option (a) above, or the two-option version I noted below that list.

The two-option combo appears to be what yours does: a PWM signal seems present from the CPU_FAN header, but absent from the CHA_FAN headers. However, that is NOT what the labels in your manual say about those headers. The only real way to check what they are doing is to use two fans - one 3-pin, one 4-pin - and test each header. Plug one fan into the header, and use settings in that header's configuration in BIOS Setup to tell it to change speeds. A header using the older 3-pin Voltage control Mode CAN change the speed of either fan type. A header using the new PWM Mode can change the speed of only a 4-pin fan, and will cause a 3-pin fan to run full speed no matter how it is set. If you really need to know whether a header with 4 pins may NOT be sending out a PWM signal, then you need to do a test with a fan Hub of known design (i.e., one that can only send out the PWM signal from a header, and supplies the +12 VDC power solely from its PSU connection with NO DC power drawn from the mobo header) in combination with those two test fans. In that connection system, the fans will always receive the full +12 VDC supply on Pin #2 from the Hub, and can only receive a PWM signal for speed control IF the mobo header supplies it. So a 3-pin fan in this scheme will always run full speed, and a 4-pin fan will change its speed ONLY if the mobo header is supplying the PWM signal (and hence IS using the new PWM Mode).

Thank you for such a detailed explanation. Unfortunately I don't have any spare fan hubs so I won't be able to perform the full test but the underlying theory is very interesting!

It's the external power requirement. That hub is pretty neat, it's got an Auto setting to allow a pwm sent signal to control upto 6x 3pin fans. Normally, 3pin fans you get 3x, maybe 4x, before hitting amperage limits from a header. 6x fans on a non header power source. Nice. You basically take a pwm signal and use it to control a rheostat.

Doing it manually is easy, it's just manually shifting power through 3 resistors to lower output voltage.

Op, those are Fan curve settings you are looking at, they control how the fans respond to the sensors. What you want is the the header setting, which will be Auto, pwm or dc/voltage. Might be found under Monitor tab or such, look for anything that says cpu_fan or chassis_fan etc.

4000rpm for a pump is normal range for many of those pumps, it'll be max. If your pump has a USB cable attachment it means it's waiting on you to install iCue software and control is initiated there, not in bios.

Looked through my BIOS twice and there is no mention of changing the fan header setting/mode. My motherboard is like the cheapest you can get for a Z370 board so maybe it lacks that feature.

About the pump RPM, I think it was always showing ~4000rpm in programs like HWInfo but I didn't notice it because I always use iCUE for my pump speed reading. Doing some research, the raw value is normally 2x of the iCUE reading which is what is happening to me too.

I thought I had a really brilliant idea yesterday. My AIO has two 4pin fan connectors coming from the pump. These are definitely PWM headers because when I attach my 3pin case fans, they always spin at 100% no matter the setting on iCUE. However, when I attach my hub to one of these pins, their speed don't change in AUTO mode either and is around ~50% of their max speed, no matter what I set for that fan in iCUE.

 

[Platinum_Gamer]

Thanks guys, I ended up moving all the case fans (3pin) onto a fan splitter which is then connected to CHASSIS_FAN2 (4pin, probs using the old Voltage Control Mode as mentioned by Paperdoc). Now my fan speeds will be dictacted by my mobo. The fan controller now has no fans attached, not sure what I'm going to use that for in the future.

 

[Paperdoc]

Thanks for Best Solution, and good luck.

Just a note about the rad fans. You say the "case fans" are now all moved to a Splitter and mobo header. But you also have fans on the radiator of your AIO system, which obviously is from Corsair. You also comment that there is a cable from the PUMP that has 4-pin male fan connectors. I presume also that there is another cable that runs from the PUMP to a mobo USB2 header. The Corsair design is intended to be connected this way, I believe, if you check its instructions. The 3-hole female connector on one cable from the PUMP goes to the mobo CPU_FAN header, and its main function is to send the pump's speed back to that header. The RAD FANS are supposed to plug into the cable coming out of the PUMP, and the Corsair iCue software controls those fans' speeds that way, using that USB2 cable connection to communicate with the pump. Thus iCue takes over CPU cooling control, and the CPU_FAN header does NOT control those rad fans. iCue can show you separately the speed of the pump, and the speed of one of the rad fans - other rad fans are presumed to be doing the same thing.

 

[Karadjgne]

Agreed. If using iCue, you'll want the rad fans connected to the pump outlets. Rgb can be connected elsewhere, it's seperate from the fan actuation. Using iCue means you can dictate fan settings from within the OS, using bios requires a reboot, which isn't exactly a desirable affect.

 

[Platinum_Gamer]

Thanks for Best Solution, and good luck.

Just a note about the rad fans. You say the "case fans" are now all moved to a Splitter and mobo header. But you also have fans on the radiator of your AIO system, which obviously is from Corsair. You also comment that there is a cable from the PUMP that has 4-pin male fan connectors. I presume also that there is another cable that runs from the PUMP to a mobo USB2 header. The Corsair design is intended to be connected this way, I believe, if you check its instructions. The 3-hole female connector on one cable from the PUMP goes to the mobo CPU_FAN header, and its main function is to send the pump's speed back to that header. The RAD FANS are supposed to plug into the cable coming out of the PUMP, and the Corsair iCue software controls those fans' speeds that way, using that USB2 cable connection to communicate with the pump. Thus iCue takes over CPU cooling control, and the CPU_FAN header does NOT control those rad fans. iCue can show you separately the speed of the pump, and the speed of one of the rad fans - other rad fans are presumed to be doing the same thing.

Agreed. If using iCue, you'll want the rad fans connected to the pump outlets. Rgb can be connected elsewhere, it's seperate from the fan actuation. Using iCue means you can dictate fan settings from within the OS, using bios requires a reboot, which isn't exactly a desirable affect.

Ya, sorry for the confusion. I've been separating the terms for case fans & radiator fans. So when I say case fans, I only meant the fans included with my Silent Base 802 and not my Corsair radiator fans.

That's how I've been running the Corsair radiator fans, each of them is plugged into its own 4pin PWM header from the pump.

My previous idea was that I can treat one of those 4pin PWM headers as the fan hub connector and then I'll manage the hub with iCUE. However that didn't work for some reason and the hub fans would spin at ~50% speed no matter what I set in iCUE and using AUTO on the hub.