second pump in loop?

[FoulOnWhite]

My loop has a single D5 in my ultitube 150 res/pump, but for some reason my flow rate as reported by the AQC high flow next is only 48.5L/H. Temps seem ok but from everything i have read since i started doing custom loop, the min flow is around 1G/M (225L/H), mine is much less than that.

I have a second D5 clone type pump that is pretty good i can add into the loop. Do i just add it into the loop wherever i can fit it in the case, or does it need to be in a specific place, ie before pump, after pump etc?

Here is a pic of my idea of placing the second pump-

 

[cvaldes]

This topic has been covered before, you should poke around for earlier discussions.

The main challenge is the water pressure variance between the two pumps and how that affects the loop (which appears to be of ordinary length based on this poorly lit photograph).

I doubt a second pump will improve coolant temperatures in your situation. It just adds expense, complexity and possibly more pump noise. The photo is lousy but it appears your coolant is less than 30C. That's perfectly fine. (Unfortunately you gave no indication whether there was any sort of GPU load when this photo was taken.)

My primary gaming rig features a Ryzen 5900X (Alphacool block + Primochill 240mm rad) and RTX 3080 Ti (Alphacool block + Alphacool 360mm rad) both cooled by a small Phobya pump in a single loop. I see no reason to add a second pump in my build.

If you want to knock down temps, you'd be better off routing the exhaust from the graphics card to a second radiator (240mm, 360mm) before sending it to the pump.

Anyhow, best of luck.

 

[FoulOnWhite]

This topic has been covered before, you should poke around for earlier discussions.

The main challenge is the water pressure variance between the two pumps and how that affects the loop (which appears to be of ordinary length based on this poorly lit photograph).

I doubt a second pump will improve coolant temperatures in your situation. It just adds expense, complexity and possibly more pump noise. The photo is lousy but it appears your coolant is less than 30C. That's perfectly fine. (Unfortunately you gave no indication whether there was any sort of GPU load when this photo was taken.)

My primary gaming rig features a Ryzen 5900X (Alphacool block + Primochill 240mm rad) and RTX 3080 Ti (Alphacool block + Alphacool 360mm rad) both cooled by a small Phobya pump in a single loop. I see no reason to add a second pump in my build.

If you want to knock down temps, you'd be better off routing the exhaust from the graphics card to a second radiator (240mm, 360mm) before sending it to the pump.

Anyhow, best of luck.

The second pump has a speed controller on the lead so i can pretty much match them(possibly) I edited the picture as it is massive, but coolant temp idle is 27.4.

I actually took out the second rad as my temps gaming are no more than 50c (i don't bench) My rad is a EK XE 360x60

 

[ShrimpBrime]

If suspecting a flow rate issue, why not just swap the pumps out?

Not enough loop there for 2 pumps really. A single rad shouldn't create too much restriction.

 

[FoulOnWhite]

If suspecting a flow rate issue, why not just swap the pumps out?

Not enough loop there for 2 pumps really. A single rad shouldn't create too much restriction.

Unless my D5 is faulty, but the RPM is seemingly ok, so why such low flow. Here is the full size pic-

 

[ShrimpBrime]

Unless my D5 is faulty, but the RPM is seemingly ok, so why such low flow. Here is the full size pic-

View attachment 323347

1500L an hour with a head of 3.7m (give or take)

I wonder if the pump just isn't getting the power it needs. It is pwm, can you change the header to DC or try a fan header instead?

I don't know the answer why.

You'd have to do some testing. But I wouldn't expect 1500L an hour. That's just an advertisement. Most pumps I've come across including a D5 isn't capable of its advertised specs unfortunately. It's probably the testing environment is different than ours, with different water blocks and rads, length tubes and so forth.

 

[Vya Domus]

From what I can see you have just 1 rad in the loop, I doubt this is a case of insufficient pressure, are you sure that thing even reports accurate flow rate ? I have both CPU and GPU like you but 4 rads in the loop with a VPP, don't know what the flow rate is but if mine seems fine I doubt this is a problem in your case.

If temperatures are fine I wouldn't worry, if you set the pump to full speed is there a significant drop in temperatures ? That's how you can tell if flow rate really is an issue.

 

[FoulOnWhite]

1500L an hour with a head of 3.7m (give or take)

I wonder if the pump just isn't getting the power it needs. It is pwm, can you change the header to DC or try a fan header instead?

I don't know the answer why.

You'd have to do some testing. But I wouldn't expect 1500L an hour. That's just an advertisement. Most pumps I've come across including a D5 isn't capable of its advertised specs unfortunately. It's probably the testing environment is different than ours, with different water blocks and rads, length tubes and so forth.

The pump is plugged directly into the PSU, speed connection on CPU header. Thing is i like the layout like this, I have spare new tube and fresh XSPC EC6 water to rejig it if i have to. What about taking the 2 90's out of the loop? Surely that can't be it can it.

From what I can see you have just 1 rad in the loop, I doubt this is a case of insufficient pressure, are you sure that thing even reports accurate flow rate ? I have both CPU and GPU like you but 4 rads in the loop with a VPP, don't know what the flow rate is but if mine seems fine I doubt this is a problem in your case.

The high flow next is very accurate yes.

https://www.igorslab.de/en/aqua-com...an-an-accurate-flow-meter-just-cant-talk-yet/

So if it says 48L/H i can be sure that is what it is. I only just stripped and cleaned the GPU block and the CPU on is relativley new, but i am still tempted to have a look in it to see if it has somehow become clogged. The only thing i can think of is the pump, but it's like £70 for a new D5.

 

[ShrimpBrime]

The pump is plugged directly into the PSU, speed connection on CPU header. Thing is i like the layout like this, I have spare new tube and fresh XSPC EC6 water to rejig it if i have to. What about taking the 2 90's out of the loop? Surely that can't be it can it.

Right, the LpH is based off pump revolutions, maybe the fan header is mistaken if in DC or PWM mode is what I'm saying.

How to test the reading is accurate?? That's what really needs to happen here honestly.

I haven't ran my D5 for a while, but I remember it being audible at full speed. If you hear it well enough, it's probably got really good flow.

 

[Vya Domus]

So if it says 48L/H i can be sure that is what it is. I only just stripped and cleaned the GPU block and the CPU on is relativley new, but i am still tempted to have a look in it to see if it has somehow become clogged. The only thing i can think of is the pump, but it's like £70 for a new D5.

It looks like your are using clear liquid, I don't know what it could become clogged with if the water looks clean. Do the test with the pump running at full speed, if you don't get a really big drop in temperatures like more than a couple of degrees then the flow rate is OK.

If one of the blocks really was clogged you'd see terrible temperatures not just a drop in flow rate.

 

[Calenhad]

No point daisy-chaining pumps. You will never match their speed 100%, so one pump will either push or pull the other pump. There are mounts for having two pumps in parallel though...

I have a similar setup. EK D5+res combo, EK fullcover GPU block, EK CPU+VRM "fullcover" block, EK crossflow 420mm radiator, and an Alphacool flowrate/temp sensor. With the pump at ~4000rpm that sensor is reporting ~200L/H. Iirc that is at 80% PWM duty cycle.

So, what RPM is your pump operating at?

And if temps are fine, nevermind the reported flow rate imho. Don't fix it if it ain't broke.

 

[FoulOnWhite]

No point daisy-chaining pumps. You will never match their speed 100%, so one pump will either push or pull the other pump. There are mounts for having two pumps in parallel though...

I have a similar setup. EK D5+res combo, EK fullcover GPU block, EK CPU+VRM "fullcover" block, EK crossflow 420mm radiator, and an Alphacool flowrate/temp sensor. With the pump at ~4000rpm that sensor is reporting ~200L/H. Iirc that is at 80% PWM duty cycle.

So, what RPM is your pump operating at?

And if temps are fine, nevermind the reported flow rate imho. Don't fix it if it ain't broke.

Pump is reporting 4736RPM so something is deffo squiffy.

 

[wNotyarD]

Pump is reporting 4736RPM so something is deffo squiffy.

What would its RPM normally be?

 

[ThrashZone]

Hi,
I've used two d5 pwm pumps in the past they do work pretty good especially with dual or more rads
Most do them series believe that is one after another I did them in parallel sort of half way in the system priority is having a d5 going into the cpu block
Best to always use identical pumps no matter which way you do series is the easiest and ek does sell dual pump setups
Weakest pump will be an issue and vario d5's I would not do it at all only pwm and connected to the same pwm pump port.
Otherwise don't connect them to the mobo and they will run full tilt I also did this hehe
Here's one octopus build my other looked identical.

 

[maxfly]

My guess is your getting a false/inaccurate reading. As far as running dual pumps without using a dualpump head, it wont hurt anything other than slightly lowered efficiency. Ive run single d5s with single ddcs and dual ddcs with single d5s. Never had a problem. Equilibrium will be reached and maintained in a pressurized loop. Martin covered this years ago. And pump placement doesn't matter at all with dual pumps. Put them wherever you have the space. Just don't run em dry.

 

[FoulOnWhite]

First i am going to try my other pump to see if the D5 is an issue. The d5 is running full speed at 47xx so there is certainly something not right. My loop cannot be thatconstrictive as i have seen loops with more 90's in them and more components and they have much better flow. I will figure out what the problem is. One thing i am sure of is the reading on the high flow next, if you check out the link, it is a very accurate meter, and i have no reason to believe mine is mis-reporting.

If there is still not much flow, i will get a dual pump unit, and try that.

Right, the LpH is based off pump revolutions, maybe the fan header is mistaken if in DC or PWM mode is what I'm saying.

How to test the reading is accurate?? That's what really needs to happen here honestly.

I haven't ran my D5 for a while, but I remember it being audible at full speed. If you hear it well enough, it's probably got really good flow.

The Lph is not based off pump revolution, the high flow next measures flow rate based on the fluid flowing through it. I guess pump speed does make a difference, but my pump is runnig 100% as i have set it in the bios. Pump is direct connected to the PSU, speed header goes onto the CPU fan header, then i set it at 100% using Qfan in the bios, so the reported speed must be correct.

 

[gurusmi]

The Lph is not based off pump revolution, the high flow next measures flow rate based on the fluid flowing through it. I guess pump speed does make a difference, but my pump is runnig 100% as i have set it in the bios. Pump is direct connected to the PSU, speed header goes onto the CPU fan header, then i set it at 100% using Qfan in the bios, so the reported speed must be correct.

Why not just connecting to the pump header? There is always 100% PWM.

 

[maxfly]

speed header goes onto the CPU fan header, then i set it at 100% using Qfan in the bios, so the reported speed must be correct.

Pump speed headers are notoriously inaccurate unfortunately, (for both d5 and ddcs). I'm assuming it's the blue wired variety. If I use a pump with one, I don't use them, tbh I usually cut them off. I try to stick with pwm enabled pumps if at all possible. Not saying you should rush out and buy a new pump as there's likely nothing wrong with your pump at all other than the typically janky header. Your temps look fine, I wouldn't worry about it.

 

[xtreemchaos]

you dont need another pump with the size of your loop mate unless yours becomes faulty. i run 2x 480 rads off a single ek D5 and its more than enough and there is a prob with too much water flow ie not getting enough time for the water to pick up the heat while traveling through the blocks which is a thing believe it or not.

 

[FoulOnWhite]

CPU was clogged, anyway the rubber seal in this new alphacool core 1 that was £120 is a bloody awful fit. Not even sure if I will be able to get it back together properly

I think I might be buying another block.

I'll be really annoyed if this leaks.

 

[ThrashZone]

Hi,
Yeah the faster the fluid goes through radiators they actually are less efficient at removing the heat from the fluid unless you increase the fan speed and of course you add noise
Plus d5's are water cooled so you actually add heat to the fluid having two pumps.

But the dual pump combo is the way to go you can put any d5's in them but as I said it is best to get same brand/ type for best results.
Save the extra for another build.

EK-XTOP Revo Dual D5 PWM Serial - (incl. 2x pump)

EK-XTOP Revo Dual D5 PWM Serial is a fourth generation high-performance aftermarket DUAL pump top (volute) for D5 type water pumps, pre-installed with two EK-D5 PWM water pumps.

www.ekwb.com

 

[xtreemchaos]

it looks like the jet plate director has deformed over time, id email AC to see if they would replace bud.

 

[ThrashZone]

CPU was clogged, anyway the rubber seal in this new alphacool core 1 that was £120 is a bloody awful fit. Not even sure if I will be able to get it back together properly
View attachment 323420

I think I might be buying another block.

Yeah that happens if fluid isn't clean I've been there lol
Use coffee filters to filter your fluid before using it.

Optimus cpu blocks are pretty damn good for US.

Amazon.com

 

[kapone32]

I have a system with 2 rads and 2 pumps. One of the pumps sits at the bottom of the case and the other one is on the side mounted rad. Doing that lowered my GPU(s) temp to idle at 25 C and not go above 49 C when working. One of the things I made sure was that they were both the same pump but I also made sure one res was much larger than the other.

 

[FoulOnWhite]

Don't really understand it, I use a fresh bottle of xspc ec6 every time. So God knows how it got clogged, which was some black stuff btw, But anyway the rubber seal is rubbish on something so expensive, it has stretched? So it does not even fit properly now over the hot plate. I'm a bit nervous about putting it back in there now.

Might be time to put my air coolers back on till I get another block, which will be another HK as I had no problems with that even stripping it a few times to clean it.