Let's talk Flow Sensors...

[TahoeDust]

I am looking for a flow sensor yo use on the flow senso header on a Asus Maximus XI Hero.  All the other fittings on the build are Bitspower, so that was my initial thought, but when looking at them I see a lot of complaints about restriction and/or reliability.  The more I looked the more I saw the same complaints about other brands/models.  What are you guys using and how do you like them?  How long have you had them?

[For Science!]

11 minutes ago, TahoeDust said:

-

During my early days of watercooling, I also had a Koolance Flow meter which was hookedup with a multiplier that then fed into the motherboard, which allowed me to measure the ml/min flow of my system as an rpm readout. My intentions were to use this as the CPU_FAN input signal, so that if the flow rate became too low for whatever reason, this would throw an CPU_FAN error. My justification at the time was that this is a more real-readout of the system working than for example pump rpm, which is not necessarily an indicator of flow.

 

A few problems arised however, my computer sometimes POSTed far faster than that of the flow of the system, and that occasionally I would get CPU_FAN errors because the system flow wasn't up to speed when the first CPU_FAN check kicked in. Secondly, the flow meter wasn't silent, there was this faint rotating sound that was audible enough in my quiet system, and so it had to go.

 

Now, I just rely on my pump rpm as "good enough" to feed into my CPU_FAN header.

 

[TahoeDust]

4 minutes ago, For Science! said:

During my early days of watercooling, I also had a Koolance Flow meter which was hookedup with a multiplier that then fed into the motherboard, which allowed me to measure the ml/min flow of my system as an rpm readout. My intentions were to use this as the CPU_FAN input signal, so that if the flow rate became too low for whatever reason, this would throw an CPU_FAN error. My justification at the time was that this is a more real-readout of the system working than for example pump rpm, which is not necessarily an indicator of flow.

 

A few problems arised however, my computer sometimes POSTed far faster than that of the flow of the system, and that occasionally I would get CPU_FAN errors because the system flow wasn't up to speed when the first CPU_FAN check kicked in. Secondly, the flow meter wasn't silent, there was this faint rotating sound that was audible enough in my quiet system, and so it had to go.

 

Now, I just rely on my pump rpm as "good enough" to feed into my CPU_FAN header.

Spoiler

 

 

 

Thanks for your time and your post.  So, you use your pump on the CPU_FAN header?  I was planning on using my pump on the W_PUMP header and running my CPU_FAN header to a PWM splitter that would feed my 7 fans on the loop.

 

System looks awesome BTW.

[For Science!]

1 minute ago, TahoeDust said:

-

Thanks I've done a few rearragements since and I think the flow meter looked way too bulky in hindsight though

 

Since the CPU_FAN header is the one that usually sends out warning signals when it fails, I prefer to have the most critical component tied to it, which in watercooling is undoubtedly the pump. In my case the pump receives seperate molex power and so the pump only returns the rpm signal and uses the PWM signal from the motherboard (so no concerns of power draw). This is also why I recommend AIOs to be plugged into the CPU_FAN and not the AIO_PUMP header as the AIO_PUMP header I believe has no consequence if left unplugged.

 

In my system below, the rear intake fan is powered by the HAMP_FAN header, purely because it is closeby, and the front 3 fans are PWM split of CHA_1 and the top 2 fans are PWM split off CHA_2. The headers have no problems with these numbers of fans and so I am using a simple unpowered splitter.

 

Another advantage for not using the CPU_FAN header to control the fans is that by default at least for ASUS boards, the CPU_FAN header is forcibly tied to the CPU temperature. This is not ideal since you would want fan speeds to react to fluid temperatures. I use a 2-pin temperature sensor probe that is connected to T_SENSOR1 and this is used as the temperature source for CHA_1, CHA_2, and HAMP_FAN. This means that fans react very gently to the gradual increase of fluid temperature as the system goes under load instead of whirring up and down with CPU or GPU load. This is also why I use the CPU_FAN with the pump since I set the pump at a constant speed regardless of the CPU temperature, and so this is a kill two birds with one stone scenario.

 

 

[TahoeDust]

10 minutes ago, For Science! said:

Thanks I've done a few rearragements since and I think the flow meter looked way too bulky in hindsight though

 

Since the CPU_FAN header is the one that usually sends out warning signals when it fails, I prefer to have the most critical component tied to it, which in watercooling is undoubtedly the pump. In my case the pump receives seperate molex power and so the pump only returns the rpm signal and uses the PWM signal from the motherboard (so no concerns of power draw). This is also why I recommend AIOs to be plugged into the CPU_FAN and not the AIO_PUMP header as the AIO_PUMP header I believe has no consequence if left unplugged.

 

In my system below, the rear intake fan is powered by the HAMP_FAN header, purely because it is closeby, and the front 3 fans are PWM split of CHA_1 and the top 2 fans are PWM split off CHA_2. The headers have no problems with these numbers of fans and so I am using a simple unpowered splitter.

 

Another advantage for not using the CPU_FAN header to control the fans is that by default at least for ASUS boards, the CPU_FAN header is forcibly tied to the CPU temperature. This is not ideal since you would want fan speeds to react to fluid temperatures. I use a 2-pin temperature sensor probe that is connected to T_SENSOR1 and this is used as the temperature source for CHA_1, CHA_2, and HAMP_FAN. This means that fans react very gently to the gradual increase of fluid temperature as the system goes under load instead of whirring up and down with CPU or GPU load. This is also why I use the CPU_FAN with the pump since I set the pump at a constant speed regardless of the CPU temperature, and so this is a kill two birds with one stone scenario.

 

 

Makes sense.  What temp sensor do you use and where did you locate it?

[For Science!]

Just now, TahoeDust said:

Makes sense.  What temp sensor do you use and where did you locate it?

I use a Bitspower 2-pin sensor (something like the one linked, but white). It is plugged into the spare port on the EK reservoir using the included port extender (otherwise it leaks). Location shouldn't really matter since the temp of the fluid is pretty much equal throughout the system, and so it is just a convenient place with a spare port.

 

 https://www.amazon.com/Bitspower-Temperature-Sensor-Fitting-Matte/dp/B003U3S0UQ

[TahoeDust]

2 hours ago, For Science! said:

I use a Bitspower 2-pin sensor (something like the one linked, but white). It is plugged into the spare port on the EK reservoir using the included port extender (otherwise it leaks). Location shouldn't really matter since the temp of the fluid is pretty much equal throughout the system, and so it is just a convenient place with a spare port.

 

 https://www.amazon.com/Bitspower-Temperature-Sensor-Fitting-Matte/dp/B003U3S0UQ

Thanks.  I went ahead and scooped one up.  I also had previously bought this for a visual display, but adding the bitspower will let me shape my fan curve to the fluid temp.  Thanks.

[Mick Naughty]

Had one, though it would be a good idea to have it. Kind of a waste and just did it because it was free. 

 

Water temp is the only thing that mattered and I can’t base what a good flow rate is supposed to be based on its rpm. 

 

Think the standard flow indicators are as fancy as one should get for a standard windowed case. 

[Neo-revo]

Will be doing my flow speed theory test soon when i am bored one day, i got 6 cheap spinners from china just for visuals.  was going to test it right away initially but went deeper and did a hard line conversion. will just use soft line to test the theory since compression's didn't like the splitters i got  @For Science!

 

3 hours ago, TahoeDust said:

So, you use your pump on the CPU_FAN header?

for if the pump fails with out you noticing the system will think your cpu fan died, get scared, and let you know.

[TahoeDust]

2 minutes ago, Neo-revo said:

for if the pump fails with out you noticing the system will think your cpu fan died, get scared, and let you know.

I'm suprised that Asus does not make the same alarm conditions for the W_PUMP header as the CPU_FAN.

[Neo-revo]

1 minute ago, TahoeDust said:

I'm suprised that Asus does not make the same alarm conditions for the W_PUMP header as the CPU_FAN.

maybe they do, but most of use have been cooling since before they even thought of things like this,  my rpm are in line with my fan controller so i just disable cpu fan fail notice.

 

if one of the pump fails i get a fan alarm off my scythe something or rather

[TahoeDust]

Since I have some knowledgeable people in here, let me ask a newb question.  There is an area of my loop (the way I have it planned) where a fitting will be pointing downward, and will need to be holding some weight. 

 

In this picture, should I be concerned about the weight of the fittings circled in blue, separating the tube/fitting circled in yellow?  If so, is there an adhesive I could use to be certain this won't be a problem?

 

[Mick Naughty]

Wouldn't worry about it. As it sits, wouldn't be hard to support it by putting something under it. Though the other tube will also be supporting it.

[For Science!]

13 minutes ago, TahoeDust said:

 

I also don't think its an issue in terms of leaks. May be difficult to get it to look straight though, so I would put some support underneath it for aesthetic reasons.

[TahoeDust]

1 hour ago, Mick Naughty said:

Wouldn't worry about it. As it sits, wouldn't be hard to support it by putting something under it. Though the other tube will also be supporting it.

 

1 hour ago, For Science! said:

I also don't think its an issue in terms of leaks. May be difficult to get it to look straight though, so I would put some support underneath it for aesthetic reasons.

Thanks fellas.  I need to find something cool to support...lol.

[Neo-revo]

i was scared, but i  have torque.   until i figure some thing better, a sanded pvc, with a groove notch and elastic to hold in place (not slide easily)