Question AIO Pump not registering in BIOS or elsewhere

[BoomerD]

After I re-routed my various fan cables following the wi-fi antenna issue which required me to pull the motherboard, I see that the Arctic Cooling Liquid Freezer II pump no longer registers any RPMs in the BIOS, in Fan Control, or any other program. ALL other fans register. The pump appears to be working...temps seem fine, and the tiny VRM fan in the pump is spinning happily away.
Any ideas? (yes, the pump is connected to the CPU PUMP_FAN header.)

 

[Seba]

Check in BIOS (UEFI) if that fan header is set in PWM mode (and not in DC mode). That change might have happened if you performed a Clear CMOS at one point or reset BIOS to default (depending on Default settings for that motherboard).

Also test with a known good PWM fan in that header.

 

[BoomerD]

Check in BIOS (UEFI) if that fan header is set in PWM mode (and not in DC mode). That change might have happened if you performed a Clear CMOS at one point or reset BIOS to default (depending on Default settings for that motherboard).

Also test with a known good PWM fan in that header.

Everything is in PWM mode...first thing I checked. Recently upgraded the BIOS to the latest version with the 0x129 microcode. I wondered it that had changed the PWM status. No CMOS done.

 

[Paperdoc]

There are only a few things that can cause this. The pump and its VRM fan are working and you report it appears that cooling is correct, so the only issue is the pump speed signal.That signal, generated by the pump motor, is fed back to the mobo CPU_FAN header on Pin #3 of the header. On the header and fan female connector the "tongue" beside the header is next to Pins 1 to 3, and Pin 4 is beyond the tongue width. For some reason the signal is NOT getting to your header.

Inspect the female connector and header to look for any indication that there is a poor connection of the wire in the connector or a loose pin. Also try this. With your system off, disconnect that connector, then plug it back in. Do this several times carefully so you do not disconnect something else in that area. Finish with it re-connected then try turning on power. Sometimes this process can "scrub" small bits of dirt or oxidation on the connector and pins to make a good contact again.

Note the hint by Seba above to test the header itself with another fan. If none of that works, then there must be a broken connection somewhere inside the pump unit, and there really is no way to disassemble that to repair. That points to replacement.

Your system does NOT need that speed signal for proper control of cooling. The speeds of the pump and VRM fan are controlled by the header in this system (that is unusual for AIO's) but that does not require knowing what the speed is. HOWEVER, the CPU_FAN header has an important second function. It does monitor that speed signal for NO signal (which is what you day IS happening now!) and that is interpreted to mean FAILURE of the pump. On most mobos that would pop a prominent warning on your screen AND, in a short time, may shut down your system entirely even though the temp sensor inside the CPU chip does not report high temps. This is to protect the CPU from very rapid overheating with no cooling. So I am surprised that your system is running at all! MAYBE you have turned OFF the speed display and monitoring feature of your CPU_PUMP header, OR it was re-set that way during your updating. Look closely in the options for the header in BIOS Setup for some option like NOT using the default automatic controls or not using the header at all. Ensure it IS set to exert normal control of its "fan" or pump.

 

[BoomerD]

Check in BIOS (UEFI) if that fan header is set in PWM mode (and not in DC mode). That change might have happened if you performed a Clear CMOS at one point or reset BIOS to default (depending on Default settings for that motherboard).

Also test with a known good PWM fan in that header.

OK, Swapped the pump and CPU fan connectors. Pump still works, fans still work...but the pump shows zero RPMs.
(more info below) (note that the Pump 1 connector is connected to two 140mm fans, NOT the pump. That is connected to CPU 1)

There are only a few things that can cause this. The pump and its VRM fan are working and you report it appears that cooling is correct, so the only issue is the pump speed signal.That signal, generated by the pump motor, is fed back to the mobo CPU_FAN header on Pin #3 of the header. On the header and fan female connector the "tongue" beside the header is next to Pins 1 to 3, and Pin 4 is beyond the tongue width. For some reason the signal is NOT getting to your header.

Inspect the female connector and header to look for any indication that there is a poor connection of the wire in the connector or a loose pin. Also try this. With your system off, disconnect that connector, then plug it back in. Do this several times carefully so you do not disconnect something else in that area. Finish with it re-connected then try turning on power. Sometimes this process can "scrub" small bits of dirt or oxidation on the connector and pins to make a good contact again.

Note the hint by Seba above to test the header itself with another fan. If none of that works, then there must be a broken connection somewhere inside the pump unit, and there really is no way to disassemble that to repair. That points to replacement.

Your system does NOT need that speed signal for proper control of cooling. The speeds of the pump and VRM fan are controlled by the header in this system (that is unusual for AIO's) but that does not require knowing what the speed is. HOWEVER, the CPU_FAN header has an important second function. It does monitor that speed signal for NO signal (which is what you day IS happening now!) and that is interpreted to mean FAILURE of the pump. On most mobos that would pop a prominent warning on your screen AND, in a short time, may shut down your system entirely even though the temp sensor inside the CPU chip does not report high temps. This is to protect the CPU from very rapid overheating with no cooling. So I am surprised that your system is running at all! MAYBE you have turned OFF the speed display and monitoring feature of your CPU_PUMP header, OR it was re-set that way during your updating. Look closely in the options for the header in BIOS Setup for some option like NOT using the default automatic controls or not using the header at all. Ensure it IS set to exert normal control of its "fan" or pump.

So, as mentioned above, I swapped the CPU fans and the pump connectors. No difference. Pump still shows zero RPMs, fans work fine.
Reconnected things to their proper headers, verified that all were set to PWM and Smart Fan, turned everything to FULL SPEED. Still no pump speed showing.



I'm scratching my head...gotta be something in the pump. The wiring and connector look good, as mentioned, temps are fine and the VRM fan works.
I've contacted Arctic Cooling's tech support (again) to see if they can help.

 

[Paperdoc]

If the pump appears to work AND you do get good cooling, the most likely failure is simply in the speed signal transmission wiring. The pump is not delivering a signal to the header. Arctic Cooling's Tech support might have a solution. But what you really are NOT getting right now is the ability of the CPU_FAN header to detect when the pump actually fails. That's a very useful protection when it happens, but many pump failures are do not happen that way and are detected from some other failure symptoms. Unless Arctic can give you a fix, you may have to decide whether to run without that failure detection system working or to replace it.

 

[BoomerD]

If the pump appears to work AND you do get good cooling, the most likely failure is simply in the speed signal transmission wiring. The pump is not delivering a signal to the header. Arctic Cooling's Tech support might have a solution. But what you really are NOT getting right now is the ability of the CPU_FAN header to detect when the pump actually fails. That's a very useful protection when it happens, but many pump failures are do not happen that way and are detected from some other failure symptoms. Unless Arctic can give you a fix, you may have to decide whether to run without that failure detection system working or to replace it.

Seems like that might be a warranty issue...

 

[Seba]

Are you sure it used to report the pump speed?

After some searches, it looks like on this AIO cooler (or on some variants of it) the pump speed is not reported (but supposedly it can still be controlled).

 

[BoomerD]

Are you sure it used to report the pump speed?

After some searches, it looks like on this AIO cooler (or on some variants of it) the pump speed is not reported (but supposedly it can still be controlled).

Turns out…no. Per Arctic Cooling, the pump DOESN’T report speed… at all.

The Liquid Freezer II wasn't designed to show the RPM values of the pump, as we believe the pump should always be running at 100%. Due to this, the pump doesn't have a signal pin to send and receive RPM values, which makes the RPM value zero.

The signal that was being registered before you removed the fans was the fans themselves since the fans are the only ones with signal pins.

The pump should be running at 100% as long as it is plugged in CPU, AIO, or PUMP header.

 

[Paperdoc]

OK, with more hunting I see what is going on. The Arctic Freezer II system uses ONE cable to connect all of its components to one header. Normally that is the CPU_FAN header. The manuals for that system say that THREE items are powered and controlled that way, all by using the PWM speed signal from the CPU_FAN header.Those items are: the pump, the small VRM cooling fan mounted in the pump body, and the rad fans, and the speeds of ALL of them ARE changed by the PWM signal. UNlike most AIO systems this one was NOT designed to have the pump run full speed all the time.

Now, in any system where more than one item is fed from a header, that header is only provided with the speed signal of ONE item because that's all it can deal with. In this case apparently the designers wired it so that the speed of ONE of the RAD FANS is what is sent back to the CPU_FAN header.

SOME people - and apparently OP, but he / she did not say this - do NOT connect the RAD FAN cable to the 4-pin output socket on the pump body so that it is powered and controlled as designed. Instead they connect the rad fans to some different mobo header. When that is done, there is NO speed signal connected back to the CPU_FAN header by the cable from the pump. So, there is no pump speed available for display, ever.

This kind of wiring design makes it impossible for anything to monitor the speed signal from the PUMP for possible failure. That's a weakness, although many would consider this missing protection not a big problem.

 

[BoomerD]

OK, with more hunting I see what is going on. The Arctic Freezer II system uses ONE cable to connect all of its components to one header. Normally that is the CPU_FAN header. The manuals for that system say that THREE items are powered and controlled that way, all by using the PWM speed signal from the CPU_FAN header.Those items are: the pump, the small VRM cooling fan mounted in the pump body, and the rad fans, and the speeds of ALL of them ARE changed by the PWM signal. UNlike most AIO systems this one was NOT designed to have the pump run full speed all the time.

Now, in any system where more than one item is fed from a header, that header is only provided with the speed signal of ONE item because that's all it can deal with. In this case apparently the designers wired it so that the speed of ONE of the RAD FANS is what is sent back to the CPU_FAN header.

SOME people - and apparently OP, but he / she did not say this - do NOT connect the RAD FAN cable to the 4-pin output socket on the pump body so that it is powered and controlled as designed. Instead they connect the rad fans to some different mobo header. When that is done, there is NO speed signal connected back to the CPU_FAN header by the cable from the pump. So, there is no pump speed available for display, ever.

This kind of wiring design makes it impossible for anything to monitor the speed signal from the PUMP for possible failure. That's a weakness, although many would consider this missing protection not a big problem.

Pretty much this. I connected the pump to the PUMP_FAN header, the two radiator fans to the CPU_ FAN header. Since the pump wiring included the PWM wire, I “Ass- U-Me”d that the pump had a tachometer signal that reported its speed separately to the motherboard. I agree it’s kind of a poor design…but as long as it works…then it’s not a big deal. I was surprised by AC’s tech support’s response that they expect the pump to run at 100%, since the spec sheet says 800-2000 RPMs, controlled by PWM.

https://www.arctic.de/media/92/24/d4/1690272957/Spec_Sheet_Liquid_Freezer_II_280_EN_update.pdf

 

[BoomerD]

So I questioned Arctic Cooling’s tech support…

By the way, if the pump is supposed to run at 100% as you say, why does the spec sheet say "800-2000 RPMs controlled by PWM?"
I think that's what threw me on this. Maybe the pump doesn't have a "tachometer" that reports its speed to the motherboard like a fan does. That seems...odd to me.

Their reply:

Excuse me for not clarifying, that is correct, the pump cannot show RPM values but that does not mean you cannot adjust the pump's RPM with RPM controls.

 

[Paperdoc]

Just an FYI. We talk all the time about fan SPEED control. But the truth is that all fan headers concentrate on the TEMPERATURE as measured by a relevant sensor. It changes the "speed signal" it sends out to its "fan" (or other device) according to that temperature and its "fan curve". The actual signal is really a "% of full speed" signal, and it does NOT care what the real speed is. So it never USES that speed info for its control purposes, because it cares about temperature. Thus the LACK of actual fan speed or motor speed info is not an issue for control of cooling.

The speed signal the header receives from its "fan" IS used, though, to monitor that device for possible FAILURE. If that occurs (no fan speed signal) it takes action by alerting you on screen and possibly taking more active measures. It is this function that is not possible when the speed of the pump in an AIO system is NOT reported by sending its signal back to a header.