[SOLVED] Can I connect the GT Scythe 5400 RPM Fan (3-Pin) on the CPU Pump Header (4-Pin)?

KJLH

Reputable
Oct 16, 2016
9
0
4,510
Hi, I just got a Scythe Fan that runs at high RPM. Unfortunately it's only a 3-pin so no PWM to control the speed of the fan so it will run at 5400 RPM when I connect it using the Molex. However, I'm thinking of plugging it into the my unused 4-pin CPU Pump Header on my motherboard. My motherboard does come with a DC Mode for all its 4-pin connectors. I can't really connect it to the usual case fan 4-pin since those usually can't provide enough current (usually 1.00 A). Details shown below for reference:

Model: from Scythe's website
Gentle Typhoon 120 mm High Speed Fan Series D1225C12BBAP-31
Power Consumption: 10.5 W
Rated Current: 1.14 A
Starting Current: 2.69 A
Operating Voltage: 12 V (DC)

My Motherboard:
MSI B550 A PRO powered by Seasonic Focus+ Gold 550 Watt PSU

4-Pin specifications referenced from the online manual page 36 (12 V pins)

ConnectorDefault fan
mode (Can Switch if needed)
Maximum CurrentMaximum Power
CPU_FAN1Auto mode2A24W
PUMP_FAN1PWM mode (will switch it to DC)3A36W
SYS_FAN1~6DC mode1A12W
Manual was from their website: https://www.msi.com/Motherboard/support/B550-A-PRO#down-bios

Given the information above, what can you say? Is this safe? or will it burn the pins? Hoping to learn from you guys. Thanks
 
Solution
Do not worry about those posts. Poster does not understand the header specs, but you do. The specs for PUMP_FAN1 say max current 3 A. ALL fans (and AIO pumps) run on a power supply of max 12 VDC. The AMPS flowing out of the header is determined by the impedance (complicated version of resistance) of the load attached. The header does NOT force a particular current out through its load. But if the load impedance it too low (say, a very large motor), the header cannot supply more than 3 Amps, and MIGHT be damaged trying to do so. That header will NOT damage a 12 VDC fan motor.

The only way to control the speed of a 3-pin fan is to vary the VOLTAGE supplied to it by the header. That is why you must configure the header to use the DC Mode...

KJLH

Reputable
Oct 16, 2016
9
0
4,510
dont connect on pump fan, connect on SYS_FAN1~6 will work fine.see spec fof pump fan 36w 3a is too much. 3pin fan u can control over DC
I'm no electrical or electronics engineer though but I should ask this. Why would providing 36W 3A pin be too much considering I'm going to control/limit its voltage? Wouldn't the 36W 3A pin give me "headroom" to adjust at higher RPMs? Also, thanks for being first to respond.
 
Last edited:

dimtodim

Reputable
no u cant control on pump header fan speed, he will broken intermediately or soon...u control fan speed basically when u connect fan on sys fan he will work on 100% speed max voltage...reducing voltage u will get less speed....pump header is for aio or liquid cooling systems, not for basic fan
 

KJLH

Reputable
Oct 16, 2016
9
0
4,510
no u cant control on pump header fan speed, he will broken intermediately or soon...u control fan speed basically when u connect fan on sys fan he will work on 100% speed max voltage...reducing voltage u will get less speed....pump header is for aio or liquid cooling systems, not for basic fan
But I can set DC Voltage control on that pump fan header from the MSI BIOS like the other system fan 1-6 pins. I'm only doing it because it can supply max 3 Amperes (my fan needs 2.69 A starting current). The 1 A current on system fan 1-6 might not be enough for this fan.

I do get what you're saying that running a fan at 100% RPM is not healthy. That's why I plan on using DC mode with fan curve on the pump fan pin so it wont run at 100% all the time(similar to sys fan 1-6).

So my question really is what other concerns should I be looking at if im going to do this? What is the major difference between the pump fan 4-pin vs Sys Fan 1-6 pin that I need to consider?
 

Paperdoc

Polypheme
Ambassador
Do not worry about those posts. Poster does not understand the header specs, but you do. The specs for PUMP_FAN1 say max current 3 A. ALL fans (and AIO pumps) run on a power supply of max 12 VDC. The AMPS flowing out of the header is determined by the impedance (complicated version of resistance) of the load attached. The header does NOT force a particular current out through its load. But if the load impedance it too low (say, a very large motor), the header cannot supply more than 3 Amps, and MIGHT be damaged trying to do so. That header will NOT damage a 12 VDC fan motor.

The only way to control the speed of a 3-pin fan is to vary the VOLTAGE supplied to it by the header. That is why you must configure the header to use the DC Mode, rather than PWM Mode. It will send out to the fan motor a Voltage between 12 VDC (max speed) and about 5 VDC - any lower and the fan motor may stall and fail to re-start without some corrective action. Fed 12 VDC, that fan's specs say it will consume 10.5 W (so that means 0.875 A current), and it may allow up to 1.14A current in some situations. At most, under start-up conditions for a second or so, it may draw a current of up to 2.69A, and the PUMP_FAN1 header can provide that with NO strain.

There is one small downside to your plan. Both the CPU_FAN header and the PUMP_FAN1 header use as their guide for fan speed adjustment the temperature measured inside the CPU chip by a sensor built into that chip. That is ideal for controlling cooling of the CPU chip. The mobo SYS_FAN headers all use a different sensor built into the mobo, and that is ideal for case cooling fan control. What you propose will control that fan according to the wrong temperature sensor. BUT there is a pretty good correlation between workload and cooling needs of the CPU, and the workload and cooling needs of the rest of the mobo components. So it will work out OK, and that really is the ONLY header you have that can do the job.

Just a small note for you to check. SOME mobos do not actually let the PUMP header do any speed control of its device, since actual pumps often are designed to run at full speed all the time. So just check whether you actually do have an option in configuration of the PUMP_FAN1 header to do automatic speed control of whatever is plugged in there. The way the manual describes it indicates that you will not have a problem there.
 
Solution

KJLH

Reputable
Oct 16, 2016
9
0
4,510
Do not worry about those posts. Poster does not understand the header specs, but you do. The specs for PUMP_FAN1 say max current 3 A. ALL fans (and AIO pumps) run on a power supply of max 12 VDC. The AMPS flowing out of the header is determined by the impedance (complicated version of resistance) of the load attached. The header does NOT force a particular current out through its load. But if the load impedance it too low (say, a very large motor), the header cannot supply more than 3 Amps, and MIGHT be damaged trying to do so. That header will NOT damage a 12 VDC fan motor.

The only way to control the speed of a 3-pin fan is to vary the VOLTAGE supplied to it by the header. That is why you must configure the header to use the DC Mode, rather than PWM Mode. It will send out to the fan motor a Voltage between 12 VDC (max speed) and about 5 VDC - any lower and the fan motor may stall and fail to re-start without some corrective action. Fed 12 VDC, that fan's specs say it will consume 10.5 W (so that means 0.875 A current), and it may allow up to 1.14A current in some situations. At most, under start-up conditions for a second or so, it may draw a current of up to 2.69A, and the PUMP_FAN1 header can provide that with NO strain.

There is one small downside to your plan. Both the CPU_FAN header and the PUMP_FAN1 header use as their guide for fan speed adjustment the temperature measured inside the CPU chip by a sensor built into that chip. That is ideal for controlling cooling of the CPU chip. The mobo SYS_FAN headers all use a different sensor built into the mobo, and that is ideal for case cooling fan control. What you propose will control that fan according to the wrong temperature sensor. BUT there is a pretty good correlation between workload and cooling needs of the CPU, and the workload and cooling needs of the rest of the mobo components. So it will work out OK, and that really is the ONLY header you have that can do the job.

Just a small note for you to check. SOME mobos do not actually let the PUMP header do any speed control of its device, since actual pumps often are designed to run at full speed all the time. So just check whether you actually do have an option in configuration of the PUMP_FAN1 header to do automatic speed control of whatever is plugged in there. The way the manual describes it indicates that you will not have a problem there.

Thank you for the detailed response. I came here to learn and did learn new things especially the 5 VDC minimum as a precaution and impedance as a form of "resistance" to look out when connecting with 4-pins.

You did mention the the pump header referencing on the CPU socket for temperature. Fortunately, I can set the temperature source from the MSI BIOS to other sensors such as PCIE, Thermistors or the VRMs, so that's one downside crossed-out. I will do this cautiously still though, probably set my 4-pin to start at 5-6V before shutting down to connect the fan. Will post an update once I'm done setting up my PC with the Scythe Fans (I really don't care about noise due to high RPMs haha, I use headsets most of the time). Currently, this is the best help i got so thank you again, sir.
 

Paperdoc

Polypheme
Ambassador
First, thanks for Best Answer.

You do NOT need to make any adjustment for a 5 VDC minimum voltage output. There is no such thing if you are using PWM Mode, But you need DC Mode, and in that Mode the mobo already has a limit set up for that, and you cannot change it.

I assume the fan you are adding is to increase the general case ventilation. If the PUMP_FAN1 header allows you to select the temperature sensor it uses, I suggest the general motherboard one. Things like PCIe and VRM are particular mobo components. Those choices are only used if you place a particular fan aimed right at that mobo area and then use one header to control that fan dedicated to that area.

We're familiar with the way that a simple resistor impacts the current that can flow through it given a particular voltage applied. A motor is a more complex device, and the impact on current flow for any specifiec input voltage is determined by the motor's inductance and speed which produces a phenomenon called "back-emf" that appears to "resist" current flow. But a motor in motion is not a simple resistor. I just used the term to indicate that this property of the motor will limit how much current flows through it for a specified supply voltage, just as a resistor would do. For a mobo fan header, the real limit is the maximum voltage the header can provide; it does NOT push out whatever voltage is required to make a particular current flow.