Asus Z170 dedicated pump header

[MEC-777]

I said kill the motor, not the board.

 

Right, and I was asking how would powering it by molex instead of through a board header kill the pump/motor faster? It's pulling the same amps and getting the same 12v. 

 

And, like I said in an earlier post, these pumps are designed and supposed to run at 100% constant. 

[patrickjp93]

Right, and I was asking how would powering it by molex instead of through a board header kill the pump/motor faster? It's pulling the same amps and getting the same 12v. 

 

And, like I said in an earlier post, these pumps are designed and supposed to run at 100% constant. 

I meant in terms of running full time at full speed, which, no they are not. Not to mention there's usually no benefit to running above half speed unless you are after the last 1*C.

[MEC-777]

I meant in terms of running full time at full speed, which, no they are not. Not to mention there's usually no benefit to running above half speed unless you are after the last 1*C.

 

I've never seen or heard of any documentation about AIO pumps that say to run them at half speed or that they will die quicker if run at full speed constant. 

 

And yeah, it will make a difference in temps. The faster the water flows through the block, the more heat it can pull away. Otherwise the water will reach a hotter equilibrium resulting in a hotter running CPU. 

 

Some AIOs do use variable speed pumps that are controlled by their own software, but I can guarantee you, when the CPU is under higher loads, the pump gets cranked up. 

 

Really, the pumps in these AIOs aren't doing a whole lot of work. There's really no reason to believe they are strained at all running 100% constant. 

[patrickjp93]

I've never seen or heard of any documentation about AIO pumps that say to run them at half speed or that they will die quicker if run at full speed constant. 

 

And yeah, it will make a difference in temps. The faster the water flows through the block, the more heat it can pull away. Otherwise the water will reach a hotter equilibrium resulting in a hotter running CPU. 

 

Some AIOs do use variable speed pumps that are controlled by their own software, but I can guarantee you, when the CPU is under higher loads, the pump gets cranked up. 

 

Really, the pumps in these AIOs aren't doing a whole lot of work. There's really no reason to believe they are strained at all running 100% constant. 

A pump motor is built much the same way a case fan motor is. Running full tilt increases your in-motor heat, causes more rapid breakdown of your lubricants, and causes faster aging by the previous and by wear and tear. This is simply entropy at its most basic level. All processes move higher forms of energy (light, electrical, mechanical) to lower forms (sound, heat), and even if you try to do conversion purely between useful forms, some is lost to heat and can never be recovered in a useful form. Heat causes lubricants to break down faster in addition to the more violent frictional forces which spread it thinner and thinner, some of it slowly vaporizing and escaping even sealed bearings.

 

That difference has been measured, and if you take an MP655 (iirc) pump and run it between setting 1 and setting 5, even at high overclocks the difference in temperature is about 1*C, and that's because you have other limiting factors in the first place, not the least of which being 3-4 layers of solid or fluid material in-between the source (CPU die) and the water running through your block. That's why significant breakthroughs in CPU and GPU water block performance have not been made in years. There are already 2 additional layers in the way (unless you mount nakedly).

 

And the faster the fluid flows, the less heat the radiator can pull away before it's returned to the system. Not to mention the faster the pump runs, the more heat it adds to the system. You really can't win this one. Even at 400 RPM if the head pressure is sufficient, water does not spend enough time in the block to reach a higher equilibrium than it you were running at 1500RPM. You'll have the same turbulence and rising and falling of the water molecules as some heat and cool within the block as they travel in the microchannels.

 

Compared to the fans driving the air through your system, those pumps do an order of magnitude or more additional work. Water has a huge mass compared to air. Moving that mass over a distance is work. It's much harder to push water than it is to push air. That also shows up in the extra amperage needed to run those motors.

[AHaskin14]

Pretty sure my msi z87 and gigabyte z97 both have that

[MEC-777]

A pump motor is built much the same way a case fan motor is. Running full tilt increases your in-motor heat, causes more rapid breakdown of your lubricants, and causes faster aging by the previous and by wear and tear. This is simply entropy at its most basic level. All processes move higher forms of energy (light, electrical, mechanical) to lower forms (sound, heat), and even if you try to do conversion purely between useful forms, some is lost to heat and can never be recovered in a useful form. Heat causes lubricants to break down faster in addition to the more violent frictional forces which spread it thinner and thinner, some of it slowly vaporizing and escaping even sealed bearings.

 

That difference has been measured, and if you take an MP655 (iirc) pump and run it between setting 1 and setting 5, even at high overclocks the difference in temperature is about 1*C, and that's because you have other limiting factors in the first place, not the least of which being 3-4 layers of solid or fluid material in-between the source (CPU die) and the water running through your block. That's why significant breakthroughs in CPU and GPU water block performance have not been made in years. There are already 2 additional layers in the way (unless you mount nakedly).

 

And the faster the fluid flows, the less heat the radiator can pull away before it's returned to the system. Not to mention the faster the pump runs, the more heat it adds to the system. You really can't win this one. Even at 400 RPM if the head pressure is sufficient, water does not spend enough time in the block to reach a higher equilibrium than it you were running at 1500RPM. You'll have the same turbulence and rising and falling of the water molecules as some heat and cool within the block as they travel in the microchannels.

 

Compared to the fans driving the air through your system, those pumps do an order of magnitude or more additional work. Water has a huge mass compared to air. Moving that mass over a distance is work. It's much harder to push water than it is to push air. That also shows up in the extra amperage needed to run those motors.

 

I understand all that. I studied thermodynamics, physics and all that fun stuff in college. It was really not necessary to type all that out, sorry man.

 

There's nothing to "win" here. These pumps are so small and their loads are so light, the difference in wear and tear between running at 50% or 100% is not so significant that it will die WAY faster/sooner. Most will still out-live the systems they're installed in by a long shot.

 

Yes, water has more mass than air and it also has more momentum than air. Once the water is moving, it doesn't need as much amperage to maintain the flow. The extra work being done is only on initial start-up, to get the water moving. 

[patrickjp93]

I understand all that. I studied thermodynamics, physics and all that fun stuff in college. It was really not necessary to type all that out, sorry man.

 

There's nothing to "win" here. These pumps are so small and their loads are so light, the difference in wear and tear between running at 50% or 100% is not so significant that it will die WAY faster/sooner. Most will still out-live the systems their installed in by a long shot.

 

Yes, water has more mass than air and it also has more momentum than air. Once the water is moving, it doesn't need as much amperage to maintain the flow. The extra work being done is only on initial start-up, to get the water moving. 

Consider how rare you are as someone who took physics at all. Less than 4% of U.S. citizens ever take a single physics class in their lives. Now a much tinier fraction delves into thermodynamics.

 

Did you forget the gravitational forces, friction of the tubes, and the pushback from microchannels in your blocks? That's why overclockers really look into head pressure.

[MEC-777]

Consider how rare you are as someone who took physics at all. Less than 4% of U.S. citizens ever take a single physics class in their lives. Now a much tinier fraction delves into thermodynamics.

 

Did you forget the gravitational forces, friction of the tubes, and the pushback from microchannels in your blocks? That's why overclockers really look into head pressure.

Yeah, it's quite sad how few enter these important fields of study. Thermodynamics and fluid dynamics (forgot to mention that one too) I found quite interesting. 

 

Gravity doesn't play too much of a roll in AIOs because in a sense the whole unit is one reservoir and you have the same amount of water flowing down as well as flowing up, at the same time within the same system - so it pretty much balances out. Yep, there friction in the hoses and bends and restrictions through the rad. But the next question to ask is; what is the flow rate? I suspect the flow rate, even at full pump rpm, is still low enough that the given restrictions do not put significant strain/resistance on the pump/motor.

 

After just doing a quick google image search of impellers/pumps from some of the more popular AIO models, I can't see them being able to produce super-high flow rates, or flow rate that would be significantly bottlenecked/restricted by the rad. Thus, I'd still say it's better just to run them flat out, if anything, to maintain a decent flow rate. 

[nunya bus]

could be a rename and rip off. 
but if its got a bit better power delivery. as fans draw under 1 amp where as Im not really comfy putting a pump and 2 fans from my H80i GT run off the 1 header. 
So if this is giving out say 5 amps max Id feel much better about that. Currently 
my UD5H has CPU and CPU option headers. I dont run my AIO off them. I have a molex to PWM converter, and run them off my PSU. 
benefit of this is. if the pump blows the PSU will trip its protection circuits. 
on the mobo if your Pump dies. there are no safeguards. you will probably blow the header and the board! So there is still that! So I guess I wouldnt run them anyway. as they dont have surge protection anyway.... Just forget I even said anything LOL 

[Takata]

Maybe linus can test this pump header on his whole-room-water-cooling pumps.

[Ohlyver]

For people that saying this is a rip off:

1. Pumps actually suck up a lot more watts than fans. Normally normal fan headers aren't suitable for that kind of constant amp draw.

2. They spin slower than fans. A lot slower. So when you plug a pump into fan header, it tries to run it at full speed all the time (or at least on PWM Asus boards).

There is a reason why most pumps come with a Molex connector (or atleast on the side). So can you run your pump off a fan header? Yes. Should you do it? I don't think so.

I am entirely with you with the fact a pump consume more current than fans.

If anyone want to discuss why I will be glad to write down the reason why

So I come to the conclusion that the power is less limited to this header compared to standard fan connections.

[Jakusonfire]

A pump motor is built much the same way a case fan motor is. Running full tilt increases your in-motor heat, causes more rapid breakdown of your lubricants, and causes faster aging by the previous and by wear and tear. This is simply entropy at its most basic level. All processes move higher forms of energy (light, electrical, mechanical) to lower forms (sound, heat), and even if you try to do conversion purely between useful forms, some is lost to heat and can never be recovered in a useful form. Heat causes lubricants to break down faster in addition to the more violent frictional forces which spread it thinner and thinner, some of it slowly vaporizing and escaping even sealed bearings.

 

That difference has been measured, and if you take an MP655 (iirc) pump and run it between setting 1 and setting 5, even at high overclocks the difference in temperature is about 1*C, and that's because you have other limiting factors in the first place, not the least of which being 3-4 layers of solid or fluid material in-between the source (CPU die) and the water running through your block. That's why significant breakthroughs in CPU and GPU water block performance have not been made in years. There are already 2 additional layers in the way (unless you mount nakedly).

 

And the faster the fluid flows, the less heat the radiator can pull away before it's returned to the system. Not to mention the faster the pump runs, the more heat it adds to the system. You really can't win this one. Even at 400 RPM if the head pressure is sufficient, water does not spend enough time in the block to reach a higher equilibrium than it you were running at 1500RPM. You'll have the same turbulence and rising and falling of the water molecules as some heat and cool within the block as they travel in the microchannels.

 

Compared to the fans driving the air through your system, those pumps do an order of magnitude or more additional work. Water has a huge mass compared to air. Moving that mass over a distance is work. It's much harder to push water than it is to push air. That also shows up in the extra amperage needed to run those motors.

 

 

 

 

Well, no faster coolant flow is always better. A radiator just like anything else gets rid of more heat with more coolant flow.  Slowing coolant flow allows the coolant to change temperature by a greater degree and its average temperature drops. A lower average temp means less wattage dissipated. Somewhat counter-intuitively we want the rad to be as hot as possible. The reverse is true of the water blocks. We want their average temp to be as low as possible. By slowing the coolant flow the coolant increases in temp more and the average temp of the water block is increased.

[Master Disaster]

They are supposed to run at full speed. The quicker the water moves through the block, the more heat it can pull away. The instructions for my H60 said to make sure the pump is running off a constant 12v source. It's designed to run that way. If you slow the pump down, you're not getting the optimal cooling from your AIO.

Nope. Running a pump at full speed will do nothing for temperatures, all it will do is kill the pump bearings off very quickly.

Heat travels through water molecules and is able to spread through the water even without it moving. Having a water current helps this by keeping the water agitated but the current itself has very little impact on how fast heat is drawn away. The pumps main function is to push the water against gravity in the uphill sections so it can get the radiator quicker as the rad removes the heat from the water. This can be proven by turning your pump down as low as it will go, you'll notice the temps hardly increase at all. And if you then turn the pump up slowly you'll notice that at a certain point increasing the speed has zero affect on temperature (in my loop this point is 20% pump speed).

You could theoretically run a water loop with no pump at all if you could have no uphill sections (which obviously is not possible) but I assure you that pump speed has almost zero effect of heat transfer from the components to the water.

[MegaDave91]

 

Well, no faster coolant flow is always better. A radiator just like anything else gets rid of more heat with more coolant flow.  Slowing coolant flow allows the coolant to change temperature by a greater degree and its average temperature drops. A lower average temp means less wattage dissipated. Somewhat counter-intuitively we want the rad to be as hot as possible. The reverse is true of the water blocks. We want their average temp to be as low as possible. By slowing the coolant flow the coolant increases in temp more and the average temp of the water block is increased.

 

Did you copy and paste this? 

[Razarza]

Did you copy and paste this? 

Looks that way considering the formatting.

[MEC-777]

Nope. Running a pump at full speed will do nothing for temperatures, all it will do is kill the pump bearings off very quickly.

Heat travels through water molecules and is able to spread through the water even without it moving. Having a water current helps this by keeping the water agitated but the current itself has very little impact on how fast heat is drawn away. The pumps main function is to push the water against gravity in the uphill sections so it can get the radiator quicker as the rad removes the heat from the water. This can be proven by turning your pump down as low as it will go, you'll notice the temps hardly increase at all. And if you then turn the pump up slowly you'll notice that at a certain point increasing the speed has zero affect on temperature (in my loop this point is 20% pump speed).

You could theoretically run a water loop with no pump at all if you could have no uphill sections (which obviously is not possible) but I assure you that pump speed has almost zero effect of heat transfer from the components to the water.

 

I'm sorry but you're wrong. I've already gone over why in my previous posts in this thread.

 

The purpose of the pump is not to push the water against gravity. How is that possible? Think about it. An AIO is a closed system. It has the same amount of water traveling "down" and traveling "up". There is no need to overcome gravity. The main function of the pump is to move the water through the system. That's it. 

 

I see you have a D5 pump in your system, which means you have a custom loop, no? That pump has WAY more overhead capability than your average puny AIO pump. It cannot be compared in the same way. Of course it will still provide good cooling performance at 20% pump speed. It's not a tiny little impeller mounted on a CPU block.  

[Misanthrope]

Consider how rare you are as someone who took physics at all. Less than 4% of U.S. citizens ever take a single physics class in their lives. Now a much tinier fraction delves into thermodynamics.

 

Did you forget the gravitational forces, friction of the tubes, and the pushback from microchannels in your blocks? That's why overclockers really look into head pressure.

 

What? Even on public high schools basic Newtonian physic courses are mandatory here in Mexico, so it's algebra and calculus. How can it not be mandatory?

[SurvivorNVL]

What? Even on public high schools basic Newtonian physic courses are mandatory here in Mexico, so it's algebra and calculus. How can it not be mandatory?

Yeah.. welcome to America.

[Misanthrope]

Yeah.. welcome to America.

 

No wonder why Kent Hovind and his buddies can actually make a living preaching creationism. 

[MEC-777]

FYI, this thread is about 6 months old.

[Curufinwe_wins]

So I didnt bother reading the previous pages of comments, but I would like to point out a few things...

Generally speaking since all the flows here are turbulent the heat transfer to the block is not likely to be significantly affected by increased pump speed. There is however a non-negligible increase in radiator efficiency (energy dissipation) for the simple and undeniable reason that increased flow rate means the delta T across the radiator is much smaller and the fluid is at a higher temperature at the exit (as anyone with rudimentary thermo knowledge can tell you, the higher the temperature difference between the source and sink the higher the efficiency).

There are however some negatives associated with just throwing more linear velocity at a cooling loop (esp since we are so far into the turbulent regime). For starters, while the bulk will undoubtedly be at a near constant velocity, increasing that velocity can make the fluid profile less conducive to heat transfer (not a great example, but easy enough to understand is if you had sufficiently fast fluid that the small air content became trapped on the walls of the tubing/channels causing an insulating effect relative to a direct fluid interface).

Indeed if you read papers on similar topics (while I CBA atm to repost it, there is one such ACTUAL PEER-REVIEWED SCIENTIFIC study linked on the watercooling for starters thread by me about nano fluid vs DI water under specific conditions with various flow rates) you find there is an absolute maximum in heat transfer optimization along the flow rate curve.

Now it should be fairly obvious to you that engineering safety margins exist for these pumps. If NZXT says for example they will guarantee the pump for 5 years rated at continuous full power operation, then on average you should expect it to last much more than that. Indeed, good Astech based pumps these days are perfectly capable of running far beyond what most people will actually use them for and other sources of failure/disuse are much more likely (for example block/piping leakage esp after transportation). So really stop worrying about it.

It's like people who worry about killing their cpu by running them with 100 mV overvolting... Like really?

[Curufinwe_wins]

FYI, this thread is about 6 months old.

God dammit... This is what happens when I dont read all the comments...

[Master Disaster]

I'm sorry but you're wrong. I've already gone over why in my previous posts in this thread.

 

The purpose of the pump is not to push the water against gravity. How is that possible? Think about it. An AIO is a closed system. It has the same amount of water traveling "down" and traveling "up". There is no need to overcome gravity. The main function of the pump is to move the water through the system. That's it. 

 

I see you have a D5 pump in your system, which means you have a custom loop, no? That pump has WAY more overhead capability than your average puny AIO pump. It cannot be compared in the same way. Of course it will still provide good cooling performance at 20% pump speed. It's not a tiny little impeller mounted on a CPU block.

And the simple fact is, anyone with a smaller AIO pump won't be using a dedicated header to control it.

[patrickjp93]

What? Even on public high schools basic Newtonian physic courses are mandatory here in Mexico, so it's algebra and calculus. How can it not be mandatory?

 

Because freedom. I myself chose to be the exception instead of the rule and take AP Physics C, covering both classical mechanics and electricity & magnetism.

[MEC-777]

And the simple fact is, anyone with a smaller AIO pump won't be using a dedicated header to control it.

 

And that's why I said in my first post in this thread, just use a molex adapter for the pump.

 

It doesn't really matter how you power the AIO pump. This is just simply an additional mobo header for those who choose to run their pumps off the mobo - leaving more headers free for fans.