Question Pump orientation and more ?

[Seek54]

Hello, I'm currently upgrading my system in increments and have have run into a few problems and questions.

PC pic:

https://imgur.com/a/2mLGoHY

View: https://imgur.com/a/2mLGoHY


So I recently bought a new case MB520 mesh and a Aio cooler ML240 Atmos. Did quite a bit of reading but due to conflicting info couldn't be certain.

Is my current cpu cooler (pump) orientation fine? Read that it should point downwards or to the side, but logically air rises up so the tube exits should be at the top? And the "Atmos" lettering is horizontal, if I mounted it sideways the letters would also be sideways? But the tubes feel as if they are bending in this orientation, turning it counterclockwise would make them less strained. I guess the top is removable so the lettering could also be turned? Temps while gaming rarely rise over mid-50 deg C, and no bad noises.

Another problem is these case fans are 3 pin, and my mobo has only 2 headers, a pwm controlled cpu fan and 3 pin cha_fan. So I have 2 headers and 5 fans. The AlO had a splitter so the rad fans are in the pwm, everything OK there and the pump is in the 3pin cha fan running at 100%. Dont have any more headers for the case fans. I thought of adding a splitter or hub/controller but the fans ran at 100% and very loudly when connected to the cha_fan header and it can't control voltages as far as I found. Also the cpu_fan header has max of 1A, and the cooler fans use 0.3A each, case fans are 0.16A each. What are my options?

Another thing, what could I do about the gpu cable to make it less ugly?

PC Specs
Mobo: Asus P8H77-M
CPU: I7-3770k
GPU: RTX 3060
Cooler: MasterLiquid ML240 Atmos
PSU: EVGA B5 750w Bronze (I believe)
Case: MasterBox 520 Mesh

Thank you.

 

[punkncat]

That orientation should be fine as you mention yourself, bubbles will rise.
Consider picking up a fan hub. Arctic makes an inexpensive one:

https://www.amazon.com/dp/B0887VG14J?ref_=ppx_hzsearch_conn_dt_b_fed_asin_title_1&th=1

The fan in the first position controls speed for the others. Simple and reliable.

 

[Seek54]

That orientation should be fine as you mention yourself, bubbles will rise.
Consider picking up a fan hub. Arctic makes an inexpensive one:

https://www.amazon.com/dp/B0887VG14J?ref_=ppx_hzsearch_conn_dt_b_fed_asin_title_1&th=1

The fan in the first position controls speed for the others. Simple and reliable.

I thought about the hub, but my case fans will run at 100 percent at max speed then right? Since they are 3 pin fans. And where should I connect the fan hub to? Currently no headers are free, if I use a splitter with a pump and connect to cha fan which is 3 pin they will run at 100%, if split to cpu fan, since the fans are not pwm they will still run at 100%?

 

[punkncat]

I thought about the hub, but my case fans will run at 100 percent at max speed then right? Since they are 3 pin fans. And where should I connect the fan hub to? Currently no headers are free, if I use a splitter with a pump and connect to cha fan which is 3 pin they will run at 100%, if split to cpu fan, since the fans are not pwm they will still run at 100%?

Run the pump off of the CPU fan header at 80-100% (check your AIO instructions about that). Run the hub off chassis fan header and in BIOS set to follow CPU temp. If your mobo has it, set "smart fan" to on for that header.

 

[Seek54]

Run the pump off of the CPU fan header at 80-100% (check your AIO instructions about that). Run the hub off chassis fan header and in BIOS set to follow CPU temp. If your mobo has it, set "smart fan" to on for that header.

As far as I found the mobo's cha fan doesn't support control, here's what's written in the manual

https://imgur.com/a/LTCZkSV

View: https://imgur.com/a/LTCZkSV


The is no control, at least that I found, to control the cha fan header, it always works at full capacity unless I missed something.

The only way I found was to buy an expensive hub that you can control the voltage with a dial or one which has a pwm to dc converter, or just run them at max speed, but that is too loud if watching videos or just browsing stuff, managing files etc when not gaming with headphones.

Are there any other options?

 

[punkncat]

You may still have some control by way of "DC" but easiest way to find out would be as you have tried.

 

[Paperdoc]

Within the limits of what you have there is an easy way to get this all to work. It makes use of quirks of the differences in design of 4- and 3-pin fans. The key point there is that if you plug a 3-pin fan into a mobo 4-pin header that is set to use the new PWM mode of control signals, that fan will always run full speed. (That is what you have experienced with those 3-pin case fans and a Hub.) Next, note that the PUMP in that AIO system is designed and wired exactly like a 3-pin fan, so the same behaviour applies. Further, like most AIO systems, the design INTENDS that the pump runs full speed always, and all adjustment of CPU cooling is done only by changing the speed of the rad fans.

To do this you will need two 4-pin Splitters with three output "arms" each, like this example

https://www.amazon.com/Splitter-Header-Converter-Computer-Cooling/dp/B0D53G6133/ref=sr_1_9?crid=Q7ZKEOTAM0GX&dib=eyJ2IjoiMSJ9.KAgiCJfcF_3LSdySoPQylGldhJg0861AlFIMmIeKCCA7zhcPXaFWnO_9DwXfjik8MG7SHr7qvJxw-kAnPqc4PqJY-yTJBH2rwfURBo3il9fjpvSe11t_vQix4NvK-TorUcwdRMQ2gcCgiJWSjwtvvnHwfLqk7ind0be20o1r6XPEYlA24vYw5zgRNLxhjF1hKDxGY3ejNKSego3HZ52FJdiPnRzJ4wYGUwnsG-Vba_0.IWDQKLwFXuvZXVtxH8d2ei_VZXp-Rp9ArNFNvp3h8qE&dib_tag=se&keywords=4-pin+splitter+pwm&qid=1744832932&sprefix=4-pin+splitter,aps,123&sr=8-9#customerReviews

You will use one of them to connect both rad fans and the pump of the AIO all together to the CPU_FAN 4-pin header in a particular way. This leaves the 3-pin CHA_FAN header free for the 3-pin case fans. You can use the second Splitter for that.

A little more background info. Any fan header does four jobs:
a. Provide power to run the fan
b. Control the fan speed according to temp sensors
c. Count the pulses of the fan's speed signal coming back on Pin #3 to display speed, but it does NOT use that for speed control.
d. Monitor that speed for NO speed, indicating fan FAILURE. If that happens it tries to re-start it and, assuming that doe not work, it pops a warning onto your screen so you know you have a problem. On most mobos the CPU_FAN header does more that just warn you. If the fan fails it may shut down your system completely without even waiting for the temp sensor inside the CPU chip to show a high temp. This is to prevent rapid overheating from no cooling that might do permanent damage to your CPU.

Regarding fan speed control, an older 3-pin fan's speed is controlled by sending to it a Voltage supply on Pin #2 that varies, from 12 VDC for full speed down to 5 VDC for minimum speed without stalling. For new 4-pin fans that Voltage on Pin #2 is always 12 VDC, and the header also sends out on Pin #4 the new PWM signal. This fan design includes inside it a small chip that uses that PWM signal to modify current flow from the fixed 12 VDC supply line though the windings to achieve speed control. If you plug a 3-pin fan into output from that 4-pin header, it gets full 12 VDC from Pin #2 always and cannot use the PWM signal from Pin #4 to make any change, so it runs full speed.

So now to the details of connecting all of the AIO system to the CPU_FAN header with a Splitter. NOTE an important thing about a Spltter - it has ONLY one input from the host header and some output arms to the fans. It does NOT have an extra connection to the PSU directly - devices with that extra connection are HUBS, which is what you already have tried to use. Now note also this: among the output arms of the Spplitter only ONE has all 4 pins in it, and the others are missing Pin #3. On that one output Pin #3 can connect to the speed signal of its fan and send that back to the host header. And that is becasue that header can only count the pulses coming to it from ONE fan. The fact that the other Splitter outputs are missing Pin #3 means that those extra speed signals will go nowhere and never be "seen". (On the Splitter web page ignore user complaints about Pin #3 missing - those people do not understand it is supposed to be that way!) So, plug the Splitter female connector into the CPU_FAN header. In an AIO system it is important that FAILURE monitoring be done for the PUMP because no pumping means NO cooling of the CPU. (Failure of one or even both rad fans will reduce cooling and the slow temp rise inside the CPU will be detected by another system. But from time to time YOU should just look at those fans to verify they are working.) So you plug the PUMP 3-pin connector into the only Splitter output with all 4 pins to guarantee that the PUMP speed signal gets to that header to be monitored. Then you plug the rad fan cables into the other Splitter outputs.

When connected this way here's what happens. The PUMP gets full 12 VDC power at all times and does NOT get a PWM control signal it could not use anyway, so it always runs full speed as intended. Further, the pump's speed IS reported to the header so it can be montored for failure. Meanwhile the rad fans get their normal full 12 VDC power supply AND their normal PWM speed control signal, so their speeds ARE under control by that header. However, the header cannot monitor the fans' speeds for failure. You need to do that on occasion.

By the way, you CAN use a 4-pin Splitter for your 3-pin case fans from the 3-pin CHA_FAN header. The Splitter simply will not make any connections to Pin #4 that deos not exist at either end.

 

[Seek54]

Within the limits of what you have there is an easy way to get this all to work. It makes use of quirks of the differences in design of 4- and 3-pin fans. The key point there is that if you plug a 3-pin fan into a mobo 4-pin header that is set to use the new PWM mode of control signals, that fan will always run full speed. (That is what you have experienced with those 3-pin case fans and a Hub.) Next, note that the PUMP in that AIO system is designed and wired exactly like a 3-pin fan, so the same behaviour applies. Further, like most AIO systems, the design INTENDS that the pump runs full speed always, and all adjustment of CPU cooling is done only by changing the speed of the rad fans.

To do this you will need two 4-pin Splitters with three output "arms" each, like this example

https://www.amazon.com/Splitter-Header-Converter-Computer-Cooling/dp/B0D53G6133/ref=sr_1_9?crid=Q7ZKEOTAM0GX&dib=eyJ2IjoiMSJ9.KAgiCJfcF_3LSdySoPQylGldhJg0861AlFIMmIeKCCA7zhcPXaFWnO_9DwXfjik8MG7SHr7qvJxw-kAnPqc4PqJY-yTJBH2rwfURBo3il9fjpvSe11t_vQix4NvK-TorUcwdRMQ2gcCgiJWSjwtvvnHwfLqk7ind0be20o1r6XPEYlA24vYw5zgRNLxhjF1hKDxGY3ejNKSego3HZ52FJdiPnRzJ4wYGUwnsG-Vba_0.IWDQKLwFXuvZXVtxH8d2ei_VZXp-Rp9ArNFNvp3h8qE&dib_tag=se&keywords=4-pin+splitter+pwm&qid=1744832932&sprefix=4-pin+splitter,aps,123&sr=8-9#customerReviews

You will use one of them to connect both rad fans and the pump of the AIO all together to the CPU_FAN 4-pin header in a particular way. This leaves the 3-pin CHA_FAN header free for the 3-pin case fans. You can use the second Splitter for that.

A little more background info. Any fan header does four jobs:
a. Provide power to run the fan
b. Control the fan speed according to temp sensors
c. Count the pulses of the fan's speed signal coming back on Pin #3 to display speed, but it does NOT use that for speed control.
d. Monitor that speed for NO speed, indicating fan FAILURE. If that happens it tries to re-start it and, assuming that doe not work, it pops a warning onto your screen so you know you have a problem. On most mobos the CPU_FAN header does more that just warn you. If the fan fails it may shut down your system completely without even waiting for the temp sensor inside the CPU chip to show a high temp. This is to prevent rapid overheating from no cooling that might do permanent damage to your CPU.

Regarding fan speed control, an older 3-pin fan's speed is controlled by sending to it a Voltage supply on Pin #2 that varies, from 12 VDC for full speed down to 5 VDC for minimum speed without stalling. For new 4-pin fans that Voltage on Pin #2 is always 12 VDC, and the header also sends out on Pin #4 the new PWM signal. This fan design includes inside it a small chip that uses that PWM signal to modify current flow from the fixed 12 VDC supply line though the windings to achieve speed control. If you plug a 3-pin fan into output from that 4-pin header, it gets full 12 VDC from Pin #2 always and cannot use the PWM signal from Pin #4 to make any change, so it runs full speed.

So now to the details of connecting all of the AIO system to the CPU_FAN header with a Splitter. NOTE an important thing about a Spltter - it has ONLY one input from the host header and some output arms to the fans. It does NOT have an extra connection to the PSU directly - devices with that extra connection are HUBS, which is what you already have tried to use. Now note also this: among the output arms of the Spplitter only ONE has all 4 pins in it, and the others are missing Pin #3. On that one output Pin #3 can connect to the speed signal of its fan and send that back to the host header. And that is becasue that header can only count the pulses coming to it from ONE fan. The fact that the other Splitter outputs are missing Pin #3 means that those extra speed signals will go nowhere and never be "seen". (On the Splitter web page ignore user complaints about Pin #3 missing - those people do not understand it is supposed to be that way!) So, plug the Splitter female connector into the CPU_FAN header. In an AIO system it is important that FAILURE monitoring be done for the PUMP because no pumping means NO cooling of the CPU. (Failure of one or even both rad fans will reduce cooling and the slow temp rise inside the CPU will be detected by another system. But from time to time YOU should just look at those fans to verify they are working.) So you plug the PUMP 3-pin connector into the only Splitter output with all 4 pins to guarantee that the PUMP speed signal gets to that header to be monitored. Then you plug the rad fan cables into the other Splitter outputs.

When connected this way here's what happens. The PUMP gets full 12 VDC power at all times and does NOT get a PWM control signal it could not use anyway, so it always runs full speed as intended. Further, the pump's speed IS reported to the header so it can be montored for failure. Meanwhile the rad fans get their normal full 12 VDC power supply AND their normal PWM speed control signal, so their speeds ARE under control by that header. However, the header cannot monitor the fans' speeds for failure. You need to do that on occasion.

By the way, you CAN use a 4-pin Splitter for your 3-pin case fans from the 3-pin CHA_FAN header. The Splitter simply will not make any connections to Pin #4 that deos not exist at either end.

Thanks, that's a lot of info. Confirms and expands on what I have gathered so far.
But a bit more info and questions about the current setup.

The pump has a 4 pin connector, I have it plugged into the 3 pin cha_fan header. Hwinfo shows it is running at 10k rpm, but as I've read it is counted incorrectly by around 3 times more.
Now if I will connect it like you said to the cpu _fan, will it not also control the pump speed as well, since it is a 4 pin pump?

Another issue (unless I misunderstood something) is that if I connect the splitter to the cha_fan, and then to my fans, they will still run at maximum speed since they are 3 pin and will get 12V? Theoretically I could use a 3 pin splitter, not that it makes sense to use it, but the result would be the same?
I looked around in the bios and couldn't find anything that would be like a pwm or dc control of the case fans (cha_fan header), only for the cpu_fan. How could I control the speed of the fans so that they wouldn't be as loud? It doesn't need to follow cpu temps, but just not run at maximum speed?

Lastly, what about the header amperage ratings? As far as I read in the manual, it states that cpu_fan has a maximum of 1A,12 W. Doesn't mention anything, unless I missed it, about the cha_fan, but probably 1A as well. Since the 3 front fans are 0.16A, that makes it around 0.48 amps, that is fine. But what about the cpu header? If it is 1 amp, the pump consumes around 4w, according to the product page, so around 0.33A and the rad fans are 0.3A each, that adds up to 0.93A, but I've read that the real starting amperage may be higher than the stated nominal amperage. What would happen if it were to go above 1A? In cars, if the current exceeds the rated current of the fuse, the fuse pops, so in a computer would that destroy my motherboard's cpu_fan header?

Sorry for so many questions, my old case was a prebuild from around 2013 and had an air cooler and no case fans. I've upgraded the gpu and psu since then and added some ssds, but all this is new to me 😀. Thank you again.

 

[JayGau]

As you said, the air is going up, so it's gonna go into your radiator, not the pump, no matter the tubes's orientation (it's why the rule number one with AIOs is too always have the top of the radiator higher than the pump).

And every picture and video of this cooler I can find show the tubes on the side (left or right). It probably doesn't make much difference for the pump, but if you have to "bend" the tubes when they are at the top it might be better to have them on the side to release the tension.

But something is puzzling me. You said you are upgrading your system in increments, but in general, people with a so outdated system don't have the CPU cooler as their highest upgrade priority since it's not a part that brings any direct performance improvement. Why upgrading the cooler first?

 

[Seek54]

As you said, the air is going up, so it's gonna go into your radiator, not the pump, no matter the tubes's orientation (it's why the rule number one with AIOs is too always have the top of the radiator higher than the pump).

And every picture and video of this cooler I can find show the tubes on the side (left or right). It probably doesn't make much difference for the pump, but if you have to "bend" the tubes when they are at the top it might be better to have them on the side to release the tension.

But something is puzzling me. You said you are upgrading your system in increments, but in general, people with a so outdated system don't have the CPU cooler as their highest upgrade priority since it's not a part that brings any direct performance improvement. Why upgrading the cooler first?

Well the thing is that I'm unsure about how to upgrade this pc, I mean I haven't thought about the specific components I would like to have in it since I don't have a high amount of spendable money to buy combo like cpu, motherboard and ram, so I buy parts that I can spend money on without taking a hit in the moment, and the parts have to be able to go towards the next build.

Originally the pc had an i5 3350p, and a gtx 650ti boost or smth like that.

Then the gpu died at the highest price points, barrowed a used gpu from a friend for a bit then bought the rtx.

Noticed that rtx is getting very bottlenecked by the cpu, and couldn't spend like 700 or so euros at that moment so I found a used 3770k and swapped it in. But the cooler was very weak, or my case's airflow, that it kept heating up to like 80-90 deg, and I had to add a case fan from an even older pc (think 2005-2010) it had a direct connection to molex, not even a fan connector.

So the fan was very loud and annoying, so I bought this case and a week later the cpu cooler, which I think I got for decent prices, the case was 75 eur, the cooler around 62 eur.

That's pretty much the story and I haven't upgraded yet because it's not logical to buy a processor first then mobo, because one of the parts will probably be outdated in half a year or a year, I prefer Intel, probably because that's all I've had, but I've heard the 13/14th gen had overheating issues and dying chips, so that kind of scared me off as well.

I have already bought Kingston fury ddr5 ram 2x16 gb 6000MT/s so that kind of limits my options too.

I thought about using my mom's pc's mobo with a 7400 but it's weaker than the 3770k, and upgrading it to 7700k is not ideal because it has 4gb ddr4, and mine has 12gb ddr3, so I would need to buy ram again, and the ddr5 would be left unused.

So I like to buy parts that I can actually use and would likely needed to buy anyway for the futures "upgraded pc".

I plan to upgrade to something like i7-xx700k and use my ddr5 ram, this psu, cooler and case, along with the storage and gpu.

 

[Paperdoc]

Replying to your questions for me.

1. Almost all AIO systems are designed to have the pump run at full speed all the time, so they are NOT designed to include the most recent technology (PWM) for optimum control of speed. Their speed is NOT controlled. There ARE some systems that allow the user (via manual software tools) to fix pump speed at some value less than full, but those are uncommon. The fact that the connector on the end of the pump's cable has 4 holes in it does NOT make that unit a PWM-style device - it just eliminates worry for users who cannot figure out what to do with a 3-hole connector. So I assume your pump, like most others, is designed like a 3-pin fan. You MAY be able to verify this yourself just by looking closely at the WIRES in hat cable from pump to header. If it is as I suspect, there will be only THREE wires in that cable, and at the female connector on the end there will be NO wire to the fourth hole - that is, the one hole outside of the two ridges down one side. That Pin #4 is the one that carries the mobo PWM signal to the device. If there is no connection there, then the pump cannot receive the PWM signal which I THINK it cannot use, anyway.

2. I doubt very much that the pump speed is 10KRPM, as you suspect. The issue there may be the took you are using, HWInfo. Some of those third-party tools get a few things wrong. You have two other options. The easiest is to look in BIOS Setup at the header the pump is connected to and read that speed. Alternatively, I believe that AIO system also comes with a free software utility to observe and configure it, and that will show the the pump speed. BUT that latter tool may well ASSUME that the pump IS connected to the CPU_FAN header, so you might need to set it up that way as I suggested.

3. For using the 4-pin Spliter for the case fans from the 3-pin mobo CHA_FAN header, the label of "4-pin" for the Splittter has no impact. The key points are the number of pins on the mobo header, the number of pins on the fans, and the type of signals sent out by the header. For any header and / or fan to use the new PWM control system the connections MUST be 4-pin. A 3-pin header can ONLY send out signals in the older Voltage Control Mode system, and a 3-pin fan can ONLY use that signal type. Such a mobo HEADER will ALWAYS send out a supply VOLTAGE from Pin #2 that VARIES to alter the motor speed. It will NOT send out only a fixed 12 VDC that would cause a fan to run full speed.

Just three notes on this. First, yes, you could get a 3-pin Splitter for these fans, but they are hard to find today. There is NO neeed for this. When you use a 4-pin Splitter the wire for the 4th pin simply makes NO connection to anything because the mobo header and the fan have NO 4th pin or hole. Second, IF your CPU_FAN header's BIOS Setup configuration screen has a choice available for using Voltage Mode (aka DC Mode) or PWM Mode or Automatic Mode, make sure you set that to PWM Mode for the system I recommended. The mobo manual does not tell us those choices are available, but implies that the CPU_FAN header is always using PWM Mode so I did not give you this info. With the connections I recommended it is important that this header DOES use PWM Mode. Lastly, FYI, the of the speed of the rad fans controlled by the CPU_FAN header always is based on the temperature sensor built into the CPU chip itself. But for the CHA_FAN header intended for case ventilation fans, the temp sensor used normally is a different one on the mobo. SOME mobo headers like this offer you a choice of which sensor to use. For a board with only one CHA_FAN header I would not expect you to have a choice on this item, but maybe you do.

4. You are concerned about amps load on the CPU_FAN header. First, the 1.0 A max current rating on that header already includes some allowance for a very brief start-up current higher than that. There is no fuse involved. The limit is imposed only by the ability of the components of the header power output circuit - mainly, their max heat generation in normal operation. A very brief excess current merely causes the component internal temperature to rise for a moment and then drop back down as ths surge disappears. A sustained overcurrent condition, however, would cause that component to overheat and be damaged. Secondly the specs for that AIO say the pump max current is 0.32 A as you say, and each of two fans has a max SAFETY current of 0.30 A, but a max normal current draw of 0.20 A. The "Safety" current spec refers to the fact that each fan has its own internal limiter circuit to ensure that it never exceeds 0.30 A under even abnormal conditions, But with all normal conditions the current (except briefly for start-up) will not exceed 0.20 A. So the max total AIO system current draw will be 0.72 A.

 

[Seek54]

Oh, if the 3 pin header is controlled by temperature that kind of changes things then. I assumed it was constant 12V because on the header pin layout I posted from the manual it's written as +12V.

I called around and no shop near me has a 4 pin or other splitter, only hubs locally, I'll try others tomorrow.

Oh, and would it be very stupid to do something like this? https://electronics.stackexchange.com/questions/259378/undervolting-pc-fan-from-12v-to-7v


Could you explain what this circuit board in my case is?

https://imgur.com/a/RqE5Trd

View: https://imgur.com/a/RqE5Trd


Is it a splitter or hub? It has a 3pin fan exit, 3pin rgb exit as well as sata exit. If sata is connected it runs the argb lights, but not fans. If only the 3 pin connected then it runs the fans but no lights.

EDIT: I checked the pump connector, it has 4 pins and 4 wires:

https://imgur.com/a/ji1eub4

View: https://imgur.com/a/ji1eub4

 

[Paperdoc]

1. The manual for your mobo does not actually detail whether the 3-pin CHA_FAN header does control fan speed using the older Voltage Control Mode. I assumed it does because I have never seen such a header that does NOT. For that type of header the normal functions of pins are:
Pin 1 - Ground - Black wire to fan
Pin 2 - +DCV power supply varies from 12 VDC down to 5 VDC - Red wire to fan
Pin 3 - Fan Speed signal (pulse train) sent back to header - Yellow wire to fan.

A 4-pin header is different in 2 details.

2. Sellers OFTEN mix up the labels Splitter and Hub, so when they tell you on the phone that they have no SPLITTERS, they may NOT know what they are saying. Many use the label "Hub" for BOTH types of devices. YOU need to see the device and look for whether or not it has a third connection type that gets power from the PSU directly - that's a Hub. If it has ONLY one input and several outputs and nothing else, it is a Splitter.

3. That page you linked to I do NOT recommend doing that for most people. That is a "work-around" for people who have NO way to control fan speeds from a mobo header and can connect the fan only to a power supply output directly. It's a way to custom-wire the fan to the 12VDC and 5VDC lines from a Molex 4-pin PSU output so the feed to the fan is 7 VDC.

4. That circuit board shown in your black-and-white photo is a Hub - well, actually, 2 Hubs. In one board they have built a fan MOTOR Hub with only 3-pin fan output connectors (four of these), an ARGB LIGHTING Hub with wider outputs (five of these marked LED) and still only 3 pins (1 of 4 is missing), and two connection cables to other places. At the top of the photo is input from a SATA power output from the PSU. At the bottom left the two-pin connector marked SWITCH has a cable marked RESET plugged in. That cable comes from the case Reset button on its front. I has been re-purposed here so it is NOT a Reset button any more. When you push that repeatedly, it cycles the LIGHT display pattern through several choices. I note TWO things about this that would be a problem for you regarding fan SPEED control. First, it has only 3-pin fan outputs, so presumably it cannot do any PWM control function. But more than that, it has NO connection to input a speed control signal from any mobo fan header, so it cannot do ANY fan speed control. It appears in that photo that you have 3 fan MOTOR cables plugged into the Hub's fan outputs. Thus I assume that all three of those fans always run full speed. IF you want control of those threee fans' speeds you will need to unplug the fan motor cables from this Hub and get a Splitter to connect all three to the CHA_FAN header as I suggested. Leave the three fans' LIGHTING cables on that Hub so you can power and control those lights using the Reset button.

5. Look more closely at the colour photo of the connector from the AIO pump unit. Yes, it has 4 holes. But from what I see, it has THREE wires entering the connector body from the cable cover. And I'm pretty sure those go the the three RIGHT-most holes, and there is NO wire going to the hole on the far left. If that is right, it IS a "3-pin fan connector" in function despite the 4th hole.

 

[Paperdoc]

Update to my post above, Item 4 on the circuit board, regarding the labels.

Really, that is not a typical dual HUB. Its MAIN function is for the ARGB lights in the fans. And for that, it is a CONTROLLER. A lighting HUB merely distributes power from the PSU and control signals from some other source (for example, an ARGB header on a mobo) to a group of lighting units. A Controller actually creates the display control signals by itself and sends those out to its lighting units along with power for them. The latter is what this board does. It allows you, the user, to select from several pre-programmed display patterns via the Reset button on your case front.

A Hub for fan motors gets the PWM speed control signal from a mobo fan header and power for its fans from the PSU directly and sends those out to all its fans.This board does NOT get any control signal from anywhere and does not appear to have any way to input a speed setting from another source. It does get power for its fans from the PSU and send that out, but nothing more.

 

[Seek54]

I may have worded my reply incorrectly.

What I meant with the "it has a 3 pin fan exit" is that the circuit board has soldered connections (I think, didn't remove to check) at the top where you mentioned the sata cables etc.

Some of those wires is a 3 pin connector which goes to the motherboard. If that 3 pin connector is not plugged into the motherboard, none of the fans work even if they are connected to the circuit board controller.

If the 3pin connector is connected to the motherboard, the fans spin and I can see the fan speed, which is around 1250 rpm, in software like hwinfo, if it is disconnected then it shows no fans or speed.

What I am unsure of is maybe this controller (or hub, splitter whatever) uses the 3 pin for signal or something? Like a hub but not a splitter for 3 pin fans? And maybe if my motherboard had the dc fan controls it would send the same voltage to the fans but using the energy from sata? Or maybe it's just a basic splitter.

In the picture it's bit difficult to see, my phone was dying and I couldn't use the flashlight below 15 percent, but I'm 99% sure there are actually 4 cables, and 4 pins in the slots. I bent the cable at the connector and there were 4 cables, but I will recheck tomorrow, maybe due to the lighting I saw it wrong.

If it really is 4 wires, then, as I've read in other forums or reddit, not sure, it is possible to control the pump speed, as some say they have the pump running at 70% or so, because the thermals don't change much and they want less noise. If I were to use a fan splitter and connect my pump to cpu fan header, where the aio fans are connected right now (they use a 4 pin 1 to 2 splitter supplied with the aio) and which are running at around 20-30%, then the pump would also be around the same percentage?

I think for the mean time I will get a 4 pin splitter, buy a pwm fan and connect it to the cpu fan splitter along with the aio fans, which should all run at the same speed?
I need a fan because sometimes the gpu fans turn on even while not gaming, as the gpu temp goes over 40C, but can't stand running the 3 pin fans like vacuum cleaner in another room type of sound. Then while upgrading make sure to buy a motherboard that supports pwm/dc.

What other ways are there to control the voltage coming from a 3 pin splitter or hub excluding the motherboard and the molex workaround? I have read something about a inline voltage (lowerer?) or regulator, but the links were dead that got me nowhere.

 

[Paperdoc]

Yeah, it did appear to me that the connections at the top of that board are soldered. What I did NOT expect is that three of those wires go "to the mobo" and are obviously used to provide power and control the the three case vent fans plugged into the board's FAN headers. That is a board design I have never seen before. From what you describe, the board's fan MOTOR system apparently IS a basic SPLITTER, and not a Hub. The fans work only when the cable to the board is plugged into the mobo CHA_FAN 3-pin header, so those fans get NO power from any other source. Further, the CHA_FAN header cannot send out PWM Mode signals because it has no Pin #4. But I DO expect that those fans connected to that board have their speeds controlled by the CHA_FAN header, just as they would be if you were using s separate Splitter to make that connection. Is that correct?

Can you tell where the other wires from that board top go - to what mobo headers, or perhaps to a SATA power output connector from the PSU?

For the connector on the end of the cable from the pump unit, you are probably right. Re-examining the photo maybe I do see 4 wires. Certainly the AIO system web page specs say the pump has a 4-pin connector (well, we know that!). Those specs do NOT tell us whether the pump speed is always full speed or not - no speed spec at all. So, both possibilities exist. The pump MAY be wired as a new 4-pin fan with speed controlled by the PWM signal on Pin #4 of the header - that would be unusual, but possible. OR the pump may be wired just like a 3-pin fan as most are. IF that is the case, then as long as the pump is plugged into a header (like your CPU_FAN header) that is sending out signals in the 4-pin PWM Mode, it will always run full speed. For your particular situation - you have only ONE header (the CPU_FAN header) to use for the AIO system - you still should use a simple 4-pin SPLITTER to connect the pump AND both rad fans to that header. Again, ensure that the PUMP cable is plugged into the only Splitter output with all 4 pins. The pump will operate as the designers intended with its speed shown and monitored for failure. The rad fan speeds will be adjusted by the header, although they will not be shown anywhere.

Generally pump speeds are NOT intended to be less that full speed, but there ARE ways to do that using configuration options if the pump is the only device on one header. And there are a few AIO systems designed to alter pump speeds, but almost all of them do that with their own propreitary software tool. FEW people find that pump noise is troublesome, but some do. More often IF the pump makes noticeable noise that is because of a different problem - air bubbles trapped in the pump.

FYI, here is why it is NOT normal to have BOTH pump speed and rad fan speeds controlled in the same way. Those two devices have different response times - that is, the time from changing the speed to actually seeing a change in temperture is different. So suppose BOTH are being altered in the same way (from the same signal). Now the temperture at the sensor rises, so both speed up to remove more heat. The result will be that the measured temp drops. But it will drop quickly because of ONE of those changes and get really close to what is needed. Then a little later it will drop even more and get too cool. So now the system will slow BOTH down and try to fix that. Again, the system will adjust then overshoot. Repeat, repeat, etc. The solution is to have only ONE item (the rad fans) altered automatically while the other is kept constant - usually, the pump at max speed. In systems that allow it, the fan speed may be less than max but still fixed. This allows good cooling with slightly higher fan speeds, but it can limit the max amount of cooling available in high workload situations. In your case you can understand that if the rad fans AND the pump BOTH were running at 20% to 30%, cooling would be VERY limited and most likely BOTH would speed up to compensate, leading to that cycle of pump and fans "chasing" each other.

If you were to buy an extra 4-pin PWM fan and, using a 4-output Splitter connect that plus the AIO pump and 2 rad fans all to the CPU_FAN header, then yes, that extra fan would run under the same control as the rad fans. Because the extra one is not identical to the rad fans its speed might be somewhat different, but that does not matter. The actual speed is not important.HOWEVER, you must bear in mind the current draw calculation in Item 4 of my earlier post of April 17. That showed AIO system load to be 0.72 A max. So the added fan cannot have a max load over 0.25 A.

However, I am curious why you would do this that way. First of all, do you NOT have the three 3-pin case vent fans running under speed control through that odd board? Do they really run full speed all the time, or did that happen only when you connected them somewhere else? My thinking here is that you COULD connect a new 3-pin fan (pus one existing fan) to one output port of that same board using a simple Splitter. Although a 4-pin fan also could be used there (because, confusingly, the 3-pin Voltage Control Mode signal system can control the speed of a 4-pin fan), a 3-pin unit is likely a little cheaper. And I expect that board (and certainly the CHA_FAN mobo header supplying it) can handle four simple 3-pin fans - the existing 3 total only 0.48 A. By the way, the fact that the GPU board fan DOES come on and off at idle is NOT a problem, nor a symptom of inadequate cooling.

If you do replace your mobo, create a new thread for advice on how to use its features to best advantage. If you like, left-click on one of my message icons and send me a Direct Message to call attention to that new thread.

On your last item I can see two options. IF the fan power source you will use DOES change its fan speed by varying the Voltage output (the normal 3-pin Voltage Control Mode system), then some way to reduce that voltage in-line would give you both ALL speeds reduced and speeds varied by the header supplying that power. BUT that might NOT work as you expect. Remember that the header sends out whatever fan signal is required to get the actual TEMPERATURE at the relevant sensor to its target. So if you deliberately slow the fans down the temp will rise, and the header will demand more fan speed to offset what your Voltage Reducer device did! The net result would be only that you limit the MAX speed (and cooling) available.

On the other hand, if your issue is that the existing power supply system to your 3-pin fans can NOT alter the fan speeds and they always run full speed, then what you really want is a fan power supply system with MANUAL speed control. This is one example of that.

https://www.amazon.com/Cooling-YACSEJAO-Channels-1Knob-Controller（1Knob/dp/B0B2ZJK282/ref=sr_1_8?crid=32U0BJKBTCLLM&dib=eyJ2IjoiMSJ9.lTUimjtsmeNK-JIQF1CCjJ7AggDTXeO4dIgo2U4EP424nENS3wuBrW2qZpYSvYeKKLXAhy-TaqHe9hg78MtrvJA-aH5L5K_y9Xz51piQq_8UDJYIPt6V3ddIbhk9gdrj7DdteuS4AqVHqm7CGtTXC3AvcnKCmJ3t2D2WYyCDNGzYW33L0Bgw0gBRTrhWKityzWr-YOoSumlIPhsNZb3LfkdZJUxhG2neDhcPVdCpTTw.QyDy-4X2klqIsCHF-3d0vq8ky3ipwW2ABMyVrhn5WEM&dib_tag=se&keywords=computer+fan+speed+control+3-pin&qid=1745077977&sprefix=computer+fan+speed+control+3-pin,aps,126&sr=8-8

Its main plate mounts on the back of your case in an unused PCIe card slot. It gets all power from the 12 VDC lines of a 4-pin Molex power output from the PSU. It has three output connectors for fans that WILL accept either 3- or 4-pin fan connectors. What it really does is send to all its fans ONLY a VOLTAGE power supply that can be altered by its control knob. It does not try to use PWM speed technology nor can its send any fan speed signal anywhere. In your case, you would connect all your case vent fans to this controller and NOT use the odd board's fan motor outputs.

 

[Seek54]

Yeah, it did appear to me that the connections at the top of that board are soldered. What I did NOT expect is that three of those wires go "to the mobo" and are obviously used to provide power and control the the three case vent fans plugged into the board's FAN headers. That is a board design I have never seen before. From what you describe, the board's fan MOTOR system apparently IS a basic SPLITTER, and not a Hub. The fans work only when the cable to the board is plugged into the mobo CHA_FAN 3-pin header, so those fans get NO power from any other source. Further, the CHA_FAN header cannot send out PWM Mode signals because it has no Pin #4. But I DO expect that those fans connected to that board have their speeds controlled by the CHA_FAN header, just as they would be if you were using s separate Splitter to make that connection. Is that correct?

Can you tell where the other wires from that board top go - to what mobo headers, or perhaps to a SATA power output connector from the PSU?

For the connector on the end of the cable from the pump unit, you are probably right. Re-examining the photo maybe I do see 4 wires. Certainly the AIO system web page specs say the pump has a 4-pin connector (well, we know that!). Those specs do NOT tell us whether the pump speed is always full speed or not - no speed spec at all. So, both possibilities exist. The pump MAY be wired as a new 4-pin fan with speed controlled by the PWM signal on Pin #4 of the header - that would be unusual, but possible. OR the pump may be wired just like a 3-pin fan as most are. IF that is the case, then as long as the pump is plugged into a header (like your CPU_FAN header) that is sending out signals in the 4-pin PWM Mode, it will always run full speed. For your particular situation - you have only ONE header (the CPU_FAN header) to use for the AIO system - you still should use a simple 4-pin SPLITTER to connect the pump AND both rad fans to that header. Again, ensure that the PUMP cable is plugged into the only Splitter output with all 4 pins. The pump will operate as the designers intended with its speed shown and monitored for failure. The rad fan speeds will be adjusted by the header, although they will not be shown anywhere.

Generally pump speeds are NOT intended to be less that full speed, but there ARE ways to do that using configuration options if the pump is the only device on one header. And there are a few AIO systems designed to alter pump speeds, but almost all of them do that with their own propreitary software tool. FEW people find that pump noise is troublesome, but some do. More often IF the pump makes noticeable noise that is because of a different problem - air bubbles trapped in the pump.

FYI, here is why it is NOT normal to have BOTH pump speed and rad fan speeds controlled in the same way. Those two devices have different response times - that is, the time from changing the speed to actually seeing a change in temperture is different. So suppose BOTH are being altered in the same way (from the same signal). Now the temperture at the sensor rises, so both speed up to remove more heat. The result will be that the measured temp drops. But it will drop quickly because of ONE of those changes and get really close to what is needed. Then a little later it will drop even more and get too cool. So now the system will slow BOTH down and try to fix that. Again, the system will adjust then overshoot. Repeat, repeat, etc. The solution is to have only ONE item (the rad fans) altered automatically while the other is kept constant - usually, the pump at max speed. In systems that allow it, the fan speed may be less than max but still fixed. This allows good cooling with slightly higher fan speeds, but it can limit the max amount of cooling available in high workload situations. In your case you can understand that if the rad fans AND the pump BOTH were running at 20% to 30%, cooling would be VERY limited and most likely BOTH would speed up to compensate, leading to that cycle of pump and fans "chasing" each other.

If you were to buy an extra 4-pin PWM fan and, using a 4-output Splitter connect that plus the AIO pump and 2 rad fans all to the CPU_FAN header, then yes, that extra fan would run under the same control as the rad fans. Because the extra one is not identical to the rad fans its speed might be somewhat different, but that does not matter. The actual speed is not important.HOWEVER, you must bear in mind the current draw calculation in Item 4 of my earlier post of April 17. That showed AIO system load to be 0.72 A max. So the added fan cannot have a max load over 0.25 A.

However, I am curious why you would do this that way. First of all, do you NOT have the three 3-pin case vent fans running under speed control through that odd board? Do they really run full speed all the time, or did that happen only when you connected them somewhere else? My thinking here is that you COULD connect a new 3-pin fan (pus one existing fan) to one output port of that same board using a simple Splitter. Although a 4-pin fan also could be used there (because, confusingly, the 3-pin Voltage Control Mode signal system can control the speed of a 4-pin fan), a 3-pin unit is likely a little cheaper. And I expect that board (and certainly the CHA_FAN mobo header supplying it) can handle four simple 3-pin fans - the existing 3 total only 0.48 A. By the way, the fact that the GPU board fan DOES come on and off at idle is NOT a problem, nor a symptom of inadequate cooling.

If you do replace your mobo, create a new thread for advice on how to use its features to best advantage. If you like, left-click on one of my message icons and send me a Direct Message to call attention to that new thread.

On your last item I can see two options. IF the fan power source you will use DOES change its fan speed by varying the Voltage output (the normal 3-pin Voltage Control Mode system), then some way to reduce that voltage in-line would give you both ALL speeds reduced and speeds varied by the header supplying that power. BUT that might NOT work as you expect. Remember that the header sends out whatever fan signal is required to get the actual TEMPERATURE at the relevant sensor to its target. So if you deliberately slow the fans down the temp will rise, and the header will demand more fan speed to offset what your Voltage Reducer device did! The net result would be only that you limit the MAX speed (and cooling) available.

On the other hand, if your issue is that the existing power supply system to your 3-pin fans can NOT alter the fan speeds and they always run full speed, then what you really want is a fan power supply system with MANUAL speed control. This is one example of that.

https://www.amazon.com/Cooling-YACSEJAO-Channels-1Knob-Controller（1Knob/dp/B0B2ZJK282/ref=sr_1_8?crid=32U0BJKBTCLLM&dib=eyJ2IjoiMSJ9.lTUimjtsmeNK-JIQF1CCjJ7AggDTXeO4dIgo2U4EP424nENS3wuBrW2qZpYSvYeKKLXAhy-TaqHe9hg78MtrvJA-aH5L5K_y9Xz51piQq_8UDJYIPt6V3ddIbhk9gdrj7DdteuS4AqVHqm7CGtTXC3AvcnKCmJ3t2D2WYyCDNGzYW33L0Bgw0gBRTrhWKityzWr-YOoSumlIPhsNZb3LfkdZJUxhG2neDhcPVdCpTTw.QyDy-4X2klqIsCHF-3d0vq8ky3ipwW2ABMyVrhn5WEM&dib_tag=se&keywords=computer+fan+speed+control+3-pin&qid=1745077977&sprefix=computer+fan+speed+control+3-pin,aps,126&sr=8-8

Its main plate mounts on the back of your case in an unused PCIe card slot. It gets all power from the 12 VDC lines of a 4-pin Molex power output from the PSU. It has three output connectors for fans that WILL accept either 3- or 4-pin fan connectors. What it really does is send to all its fans ONLY a VOLTAGE power supply that can be altered by its control knob. It does not try to use PWM speed technology nor can its send any fan speed signal anywhere. In your case, you would connect all your case vent fans to this controller and NOT use the odd board's fan motor outputs.

The circuit board's exiting connections are argb connector, sata power for argb I believe, because only with it connected, the fans do not run, the 3 pin fan connector to the motherboard and that's it.

With the fan(s) connected to the circuit board, their speed is shown which is around 1250, it jumps around 1200-1300, but it doesn't seem to be related to temperature. I believe they are running at their max speed.

Interestingly if only the aio pump is connected to the 3 pin fan connection, the pump's speed is not shown, but that's probably because it is a 4 pin connector.

As far as I know and found with my current setup it's impossible to control the 3 pin fans unless I buy your linked item, but the shipping price is the same as the controllers, a bit higher actually, so I tried to search for the item locally, but can't find anything similar in a reasonable price, what I found was over 50 euros, ant it was a completely different design.

Does the controller with the knob have a certain name so I can look for it?

Tried to play Cs2 yesterday, the cpu temps are good maxing out at 55, but without any other fans in the case the gpu temp climbed up to 72 deg C, when I turned on the 2 front fans, the gpu temp max was 65, but I would like not dismantling the pc every time if I want it quiet to watch a movie or video, or want more cooling while gaming, as having the fans at their max speed is pretty annoying, although the temps are good then.

My original idea was to get a splitter into the cpu fan header along with the aio fans, and buy another pwm controlled fan to put in the back, leaving the aio pump in the chs fan or the circuit board fan connector, as in any case it will be running at 100 %, so the temps should be fine when doing basic things and be quiet.

Another option would be to remove 1 3pin front fan and put it in the back for exhaust, even running at 100% it will probably be quite a bit quieter than in the front, but the better option would be to somehow control the 3 fans in the front, as I'd like to use them not only for their lights.

 

[Paperdoc]

First, note that the NUMBER of pins (holes on the female connectors) of any fan connector does NOT tell you what it all does. What is important is which functions are associated with exactly which pins (holes), and whether the cable has wires to those pins. It is possible rarely that the device is wired in some odd manner. For FANS in particular, this is how the wires and connectors work.
Pin 1 - Ground - Black wire
Pin 2 - +DC power supply - Red wire
Pin 3 - Fan speed signal - pulse train sent by fan to mobo
Pin 4 - IF it exists -PWM speed control signal sent to motor
In the cables to new 4-pin fans, the colour codes of the wires are NOT as above.
For older 3-pin fans the Voltage on Pin #2 varies from 12 VDC for max speed down to about 5 VDC for min speed without stalling. This is the ONLY way to control the speed of such a fan. For the newer design of 4-pin fans the Voltage on Pin #2 is always 12 VDC, and the extra Pin #4 carries the PWM signal to the fan. Inside that fan there is a small chip that uses this signal to modify the flow of current from Pin #2 through the motor windings to change fan speed. Note the whether or not you can "see" a fan speed on its header is determined by whether or not there IS a connection via Pin# 3 to carry that signal. The NUMBER pf pins in the connectors is not the key factor. Likewise, whether or not a header can control the speed of a fan does NOT depend on the pin count. Speed control depends on which type of control signals the header sends out on its pins, and whether those signals match what the fan can use.

Fan male headers and cable female connectors are designed so that they can fit together only one way, and either connector CAN plug into either header type.
On most modern mobos (not yours, OP) only 4-pin male headers are used, BUT the options in BIOS Setup allow the user to specify whether the header sends out signals in the older Voltage Control Mode or the new PWM Mode. Note that a 3-pin header can NOT send out PWM Mode signals because there is no Pin #4. If you connect a 3-pin fan to a 4-pin header set to use the PWM Mode of signals, that fan will receive from Pin #2 a fixed 12 VDC power supply and will not receive the PWM signal from Pin #4, but it could not use that signal, anyway, because it has no special chip. So that fan always runs full speed. If you connect a 4-pin fan to a 3-pin header that fan never gets a PWM signal from the non-existent Pin #4 so its chip cannot alter current flow. However, that fan DOES receive a varying Voltage from Pin #2, so its speed IS controlled just as if it were a 3-pin fan. In this way it is easy with modern mobos to use either type of fan with any header as long as you set the header signal type to match the fan connected. But regarding speed control, of course the header can NOT control a fan speed at all if it does NOT use one of those two signal systems. For example, the header may send out only a fixed 12 VDC supply on Pin #2 at all times and NO PWM signal (or maybe not even have a 4th pin), and that can only let its fan run full speed always.

Just to be clear, OP, on YOUR mobo it appears that the CPU_FAN header is 4-pin and always uses the PWM signal mode. The only other header is the 3-pin CHA_FAN one. I'm thinking now that I have made several ASSUMPTIONS about this mobo that apply to almost all modern mobos, but not to yours. I assumed the 3-pin CHA_FAN does control the speed of its fan by varying the Voltage it supplies to the fan from Pin #2. Your posts say that, in fact, any 3-pin fan connected to that little board always runs full speed no matter what the workload is. That would indicate that the board is NOT making any attempt to control its 3-pin fans. That MAY mean that it actually does NOT get its fans' power from the CHA_FAN header and simply supplies all its fan motors with a fixed 12 VDC supply from the PSU. Alternatively, the board MAY be set to relay power from the CHA_FAN header, BUT that header on your mobo does NOT vary the power on Pin #$2 to control fan speed and only sends out a fixed 12 VDC supply. The mobo manual kind of hints that the CHA_FAN header does NOT control its fan speed.

YOU may be able to figure this out. The answer would help us devise a plan forward. Run this experiment. Disconnect the cable from that little board to the CHA_FAN header. Move the cable for ONE of the three case vent fans from that board to the CHA_FAN header. Start up and IMMEDIATELY watch that fan. The NORMAL start-up sequence for a COLD system is that it will start at full speed amd, after 5-10 sec, slow down to what is needed. Then it slowly speeds up as heat develops. Observe the speed readings for that header. Now try to create a heavy work load so that the system heats up. Does that fan speed up? IF the fan starts at full speed and stays that way always, then the HEADER is the issue because it is NOT able to vary its output Voltage on Pin #2 and control fan speed. The test in the next paragraph is not needed then. But IF the fan starts fast, slows down, then speeds up when the system gets hotter, then the header IS performing normally.

IF that test says the header can control speed, now re-connect as before: all 3-pin case fans to outputs on the board, and cable from that board to CHA_FAN header. Let the system cool down thoroughly. Same routine as above: WATCH one or more of those fans carefully immediately after start-up to see if they start fast and then slow down. Run a heavier workload and see if the all change speed. IF the test above said the header CAN change fan speed but they do NOT change their speeds in this second test and rull full speed always, then the board is the problem. By the way, I see you are RIGHT that a speed of 1200 RPM is the fan max spec.

Those two tests will tell us why the case fans do not appear to change speeds, and how to fix that.

Regarding the speed READINGS at the CHA_FAN header, I am intrigued. First, look closely at the cable from that little board that goes to the CHA_FAN header. It has 3 holes in it. BUT, how many WIRES are connected to the connector? If it has THREE wires it can send all signals between board and header. But if it has only ONE wire to Pin #3, then it can ONLY send a fan speed signal back to the header and can NOT convey the VOLTAGE supply from the header to the board.

Further, I am puzzled by another item. You report that, when the board is connected to the CHA_FAN header and all case fans to the board, that header shows a speed of about 1200 RPM, which is reasonable for a fan. PLUS that says the header CAN display the speed of its device. But IF you connect to that header instead the cable from the AIO system PUMP, you get NO speed there ?? That would indicate that the pump does NOT send out any speed signal. That would be NOT normal, but certainly is possible. Is this all correct?

 

[Seek54]

Sorry for the delayed response, had some things come up.

What I said about the aio speed not being showed was due to the fact that the aio pump was not connected to fan1 header of that hub board. When connected to fan1 it showed up.

The aio pump connector actually has 4 pins.

https://imgur.com/a/HID4eqa

View: https://imgur.com/a/HID4eqa


The hub board's connector to the mobo has 3 pins and 3 wires.

Any fan connected to the 3 pin cha_fan header speeds up and then runs at constant speed.

For now I have bought a fan splitter, and connected an exhaust fan p12 arctic pwm to the back of the case. Since it is controlled the same as the aio fans, it is almost non audible. Currently the loudest part is the hdd, which after 12 years or so is probably at end of its life. Should probably make a backup of it as it sometimes makes a few scratching like noises which last like half a second a few times.

Probably should make a different thread, but not much is certain yet. Would it be logical to buy a cpu, and the motherboard later?
Since I currently have already bought 32gb ddr5 for almost a year, I thought I could buy a cpu like intel ultra i7 265k, although mostly everyone says go amd, but the prices are very high actually. Did not search for many options, but the better ones start at like 450. And the motherboards are expensive as well. If I were to buy the 265k and then after a bit maybe motherboard prices for the z890 would drop? And then I could buy the motherboard? Or is that unreasonable.
I also did not fully understand the amd lineup with the x, x3d, and the generation overall.

 

[Paperdoc]

I am surprised that the 3-pin CHA_FAN header can NOT control speed of its fans, but that IS what the manual indicates. Not a common design these days, but some older boards were made that way. So you should have the case vent fans (except the new rear one) connected via a Splitter to that CHA_FAN header. That means they always will run full speed, but if the noise of that is acceptable then you can tolerate that.

"Loudest part is the HDD"? Unusual, but can happen, obviously. Some older drives were noisier than current models, but not really noisy. A backup is a REALLY good idea, and keep re-making that so it is kept up to date. The "scratching noises" may be this: if the HDD experiences an error in reading or writing, its normal action is to retract the head arm to its rest position at the edge of the disks, then send it back to the track it was trying to read and re-attempt the read. If that works, no problem. If it gets another error it will try the same process again. The noise you hear is just the head arm moving. There is a limit to how many time it will re-try before telling the mobo that it has an error. All of that is because one common reason for a read error is that the head arm did not position the head over the track properly, and re-doing it from a known starting position may fix that error. But other causes (substantial wear of the arm mechanism, or a real magnetic flaw in the track it is reading) cannot be fixed with this remedy. If you are experiencing such re-read actions frequently that is a sign of a failing drive unit and replacement is a good plan.

For your other queries I am not good a making recommendations for those items. I do think a new thread to deal with those is the better path.