thepcgamer099

Distinguished
Dec 1, 2013
261
2
18,785
Hi, I have a 4pin PWM noctua F12 as a radiator fan in the CPU_FAN header and the corsair H60 AIO (3 pin) pump in the CPU_OPT headerm my motherboard is asus z87-pro. My CPU_FAN RPM is controlled by the CPU temps detected on my motherboard (using Fan Xpert). Does the pump AIO receive full voltage reaching max RPM in the CPU_OPT header as it is only 3 pin and therefore cannot be controlled like the 4 pin? I am hoping this is the case, as I want my AIO pump to running at max RPM. The reason I am asking is because I have read that the CPU_OPT is essentially a slave header to the CPU_FAN and receives the same voltage or control that the CPU_FAN receives but I was unsure if this was the case with cpu_fan being 4 pin pwm and cpu_opt being 3 pin.
What is the sort of RPM I should be seeing for a H60 AIO pump, FanXpert says the pump RPM is running at around 4300RPM, is this the right sort of value for an AIO pump at max?
As you can probably tell, I do not know much about this topic but I am hoping to learn more for better control of the fans in my system. Any help is appreciated, many thanks, Tom.
 
Solution
Yes. Do not worry.

Actually, you can't go wrong here as long as the CPU_OPT header IS using the new PWM Mode suited to 4-pin PWM fans. The PUMP is wired just like an older 3-pin fan. When you plug one of those into a 4-pin header that IS using the new PWM Mode for control, that 3-pin fan (or pump) will ALWAYS receive a full +12 VDC power supply from Pin #2 and always run full speed. Most AIO system pumps are desinged to run that way.

Paperdoc

Polypheme
Ambassador
Yes. Do not worry.

Actually, you can't go wrong here as long as the CPU_OPT header IS using the new PWM Mode suited to 4-pin PWM fans. The PUMP is wired just like an older 3-pin fan. When you plug one of those into a 4-pin header that IS using the new PWM Mode for control, that 3-pin fan (or pump) will ALWAYS receive a full +12 VDC power supply from Pin #2 and always run full speed. Most AIO system pumps are desinged to run that way.
 
Solution

thepcgamer099

Distinguished
Dec 1, 2013
261
2
18,785
Yes. Do not worry.

Actually, you can't go wrong here as long as the CPU_OPT header IS using the new PWM Mode suited to 4-pin PWM fans. The PUMP is wired just like an older 3-pin fan. When you plug one of those into a 4-pin header that IS using the new PWM Mode for control, that 3-pin fan (or pump) will ALWAYS receive a full +12 VDC power supply from Pin #2 and always run full speed. Most AIO system pumps are desinged to run that way.

I read that the only 4 pin PWM header on my motherboard is the CPU_FAN header, this means the pump (CPU_OPT) is wired into a 3 pin header. Is this a problem?
 

Paperdoc

Polypheme
Ambassador
You "read that the only 4 pin PWM header on my motherboard is the CPU_FAN header", you say. The manual I found on-line says ALL the fan headers on that mobo are 4-pin. If is not absolutely clear on whether or not the four CHA_FAN headers use the new PWM Mode, but the labels for both the CPU_FAN and CPU_OPT headers are identical, suggesting they both operate the same way. I might have got the wrong manual, but that's what I found.

You also say that the BIOS reports the speed of your pump connected to the CPU_OPT header is 4300 RPM. that is much faster than most fans might do, but some pump units DO run that fast. So your pump is NOT being slowed down by the connection you have made. Further, IF that header were successfully slowing down your pump, that would be most obvious in the speed reading when the system is at LOW temperatures (with the rad fan running slowly), like at start-up when cold. If you never see the pump speed running slower, then it is NOT having its speed reduced by the header.
 

thepcgamer099

Distinguished
Dec 1, 2013
261
2
18,785
You "read that the only 4 pin PWM header on my motherboard is the CPU_FAN header", you say. The manual I found on-line says ALL the fan headers on that mobo are 4-pin. If is not absolutely clear on whether or not the four CHA_FAN headers use the new PWM Mode, but the labels for both the CPU_FAN and CPU_OPT headers are identical, suggesting they both operate the same way. I might have got the wrong manual, but that's what I found.

You also say that the BIOS reports the speed of your pump connected to the CPU_OPT header is 4300 RPM. that is much faster than most fans might do, but some pump units DO run that fast. So your pump is NOT being slowed down by the connection you have made. Further, IF that header were successfully slowing down your pump, that would be most obvious in the speed reading when the system is at LOW temperatures (with the rad fan running slowly), like at start-up when cold. If you never see the pump speed running slower, then it is NOT having its speed reduced by the header.

Yeah sorry for confusing you but you probably DO have the right manual. I read that although the fan headers ARE 4 pin connectors, they are 'pseudo 4 pin connectors' and actually the only header that experiences PWM control is the CPU_FAN header. At least that is what is stated in this article here:
https://silentpcreview.com/forums/viewtopic.php?t=66283
Yeah, Fan Xpert has been showing the pump (CPU_OPT) header to consistently being at around 4300RPM for the past few days, showing no fluctuation between CPU temps, so as long as the RPM is being reported accurately then the pump must be receiving full power.
 

Paperdoc

Polypheme
Ambassador
That is correct, I believe.

FYI, the "pseudo 4 pin connectors " comment may be right. FYI, the design of the new PWM style fans included some backwards compatibility features with the older 3-pin system. Among those is this: if you plug a PWM-type fan into a header using the older Voltage Control Mode of 3-pin headers, it WILL operate properly and have its speed controlled that way, even though this is not completely ideal. So in the early days of introducing this new design, many mobo makers began using ONLY 4-pin headers. But for some of those, they did NOT implement the new PWM Mode of speed control; rather, they simply feed those headers with the older-style Voltage Control Mode electrical signals (which do not use Pin #4 for anything) and rely on this feature of the new design to make it all work, anyway. This fails in certain circumstances, like when you try to use a 4-pin fan HUB. For most of these devices, they REQUIRE that they receive a PWM control signal from Pin #4 of the host header, so any header not doing that cannot work with a HUB. But if you are simply connecting a FAN to such a header, it accepts BOTH older and newer designs, and acts like a "universal" header.

Yours is another case where this little short-cut MAY be a problem. IF that were being done on the CPU_OPT header you are using for the pump, then the signals would cause the pump speed to be changed by the header signals. There is a way to defeat that if necessary. But you do NOT have that problem, because that CPU_OPT header really is using ONLY the new PWM Mode of control.
 

thepcgamer099

Distinguished
Dec 1, 2013
261
2
18,785
That is correct, I believe.

FYI, the "pseudo 4 pin connectors " comment may be right. FYI, the design of the new PWM style fans included some backwards compatibility features with the older 3-pin system. Among those is this: if you plug a PWM-type fan into a header using the older Voltage Control Mode of 3-pin headers, it WILL operate properly and have its speed controlled that way, even though this is not completely ideal. So in the early days of introducing this new design, many mobo makers began using ONLY 4-pin headers. But for some of those, they did NOT implement the new PWM Mode of speed control; rather, they simply feed those headers with the older-style Voltage Control Mode electrical signals (which do not use Pin #4 for anything) and rely on this feature of the new design to make it all work, anyway. This fails in certain circumstances, like when you try to use a 4-pin fan HUB. For most of these devices, they REQUIRE that they receive a PWM control signal from Pin #4 of the host header, so any header not doing that cannot work with a HUB. But if you are simply connecting a FAN to such a header, it accepts BOTH older and newer designs, and acts like a "universal" header.

Yours is another case where this little short-cut MAY be a problem. IF that were being done on the CPU_OPT header you are using for the pump, then the signals would cause the pump speed to be changed by the header signals. There is a way to defeat that if necessary. But you do NOT have that problem, because that CPU_OPT header really is using ONLY the new PWM Mode of control.

Thanks again for a really clear but detailed response!
 

Karadjgne

Titan
Ambassador
Have to seperate the distinction between pin count on the header and wire count on the connector. That's where most ppl get hung up. Nowadays all headers are 4pin. 12v/Gnd/speed/pwm. If you use a 4wired connector it'll be pwm that uses all 4 pins. If you use a 3 wired connector it uses 12v/gnd/speed and speed tells the motherboard to change the voltage instead of the pwm signal.

The cpu_fan header is almost always pwm by default, the other headers that can vary. Some are DC by default and some are pwm by default, so when set to (Auto) the header automatically assumes one or the other regardless of actual pins in use. That's when you'll need to manually change that setting to dc/voltage or pwm to get the correct kind of control.

Cpu_opt is part of cpu_fan, same controlling buss but different address, so is pwm by default. This means when on (PWM) or (Auto), anything plugged into it is treated as a pwm connection. That means full 12v constant, with speed changes made by the pwm signal wire. Plugging in a 3pin means there's no signal to change anything, so everything stays at 12v. So don't be tempted to change that header to DC/V mode, or the header will think its a DC fan, and rpm will be changed according to cpu temp fan curves.

Rpm for the Asetek and CooliT pumps is often 4000rpm ±5% and that can vary a little with how close to 12.00v exactly its receiving.
 
Last edited: