D5 Not Moving Water

[Magniphacent]

Hi everyone!

 

Built my new system and made sure everything was working fine with stock fan units.  Took the next step and pulled everything apart and installed Bitspower water blocks to the CPU and GPU.  Using Monsoon MMRS reservoir and stand-alone pump enclosure.  Installed flex tubing as hard line was too much for this first-timer and gave up on it.  Put DI water in it for the first pass to check for leaks... the water went through the pump and into all tubing that the water level (head) allowed (no leaks, so far - yay!).  Used a spare 500W PSU and turned on the pump.  Nothing.  Long story short, I had to apply 5V to the PWM connector to get the pump to run.

 

Now the pump makes noise (whines), but does not move the water.  I know that it is full as the inlet and outlet ports are full of water, thus it is primed.

 

Any ideas?

[Lipe123]

I'm not sure what you mean with 5V to pwm?

 

My D5 is older and just has a 12v molex connector, are you saying they run directly from pwm headers now? 

Did you try 12V instead of 5? 

[For Science!]

Pictures needed

[Magniphacent]

1 hour ago, Lipe123 said:

I'm not sure what you mean with 5V to pwm?

 

My D5 is older and just has a 12v molex connector, are you saying they run directly from pwm headers now? 

Did you try 12V instead of 5? 

My D5 has two wire sets coming out.  One red/black to the Molex power and another with two wires for the PWM header to the motherboard.  I had to put 5V to the PWM wire to get it to run.

[Magniphacent]

2 hours ago, BlueScope819 said:

Bad pump. Return it, and get a new one.

I was really worried about that.  It is a brand new pump.  I hate the idea of ripping it all apart again to get it out and wait another week or so to get a new one, but it is what it is, I guess.  Thanks for the suggestion.

[Lipe123]

makes no sense that it ran but is bad, it sounds like the pwm wires control the speed and the molex gives the power, set the pwm fan header to 100% in bios and see?

[For Science!]

35 minutes ago, Magniphacent said:

My D5 has two wire sets coming out.  One red/black to the Molex power and another with two wires for the PWM header to the motherboard.  I had to put 5V to the PWM wire to get it to run.

No power is provided through the rpm sense or pwm wire. You MUST power through the molex power.

[SolarNova]

11 hours ago, Magniphacent said:

My D5 has two wire sets coming out.  One red/black to the Molex power and another with two wires for the PWM header to the motherboard.  I had to put 5V to the PWM wire to get it to run.

Probably not the bets move move but still.

 

Make sure the PWN isnt plugged in, just plug in the molex , it should run full speed. Molex can sometimes be a bit of a hassle, make sure they are connecting fully. Assuming u get proper contact it should run, (also ofc make sure u have the spare PSU 24pin 'jumped' so u actually get power) if it doesnt then u've found a rare thing, a DOA D5 pump

 

 

[-rascal-]

On 8/13/2020 at 8:17 PM, For Science! said:

No power is provided through the rpm sense or pwm wire. You MUST power through the molex power.

 

@Magniphacent ^ This.

 

The 2-wire molex is what powers the pump.

The fan header is just to control the pump speed (otherwise 100% full speed all the time).

[Alphabear1978]

As SolarNova mentioned, make sure the PWN connector is not plugged in.  Both my pumps have issues starting the flow when the connector is already on the motherboard, even when that motherboard does not has any power.  Once disconnected, it runs at 100% and I don't have any issues anymore.

[Magniphacent]

14 hours ago, Alphabear1978 said:

As SolarNova mentioned, make sure the PWN connector is not plugged in.  Both my pumps have issues starting the flow when the connector is already on the motherboard, even when that motherboard does not has any power.  Once disconnected, it runs at 100% and I don't have any issues anymore.

Thanks for responding, everyone.  So, here is where I am at....  

 

Took the advice that my EK-D5 pump may be bad.  Thought that this made sense because I installed the ring incorrectly the first time and the case flooded with water.  Even though it seemed to be running at 100% when powered on, figured that I may still have blown something.  Ordered a PromoChill D5 (that is why I have been quiet for a week - waiting to get it and put it in).

 

Hooked up to external power (spare 500W PSU with Molex) and had PWM connector on Motherboard.  Ran noisy at first, but then quieted down nicely and it sounded like it was shuffling water around - the other one never did that.  But, it still was not moving the water in the loop.  I disconnected the PWM connector and put 5V on the control line and tried it again without anything connected, as Alphabear1978 suggested.  Still no movement.  I was looking forward to it gobbling-up the water from the reservoir, like all the videos that I see on YouTube.  /sad

 

I have attached a pick of the loop.  You may notice that I have connectors that I can open on the front of the fittings on each side of the GPU water block... I opened one to see if there was a air pocket problem and the water moved a bit and splashed into the water block for a second, but settled back again and still did not flow.

 

Any new thoughts?  I am worried that I am looking at a $6,000 boat anchor.  I am thinking that I may have to pull it all apart, try to put the GPU back together again, and re-install a fan tower to the CPU.  I don't want to, but I cannot think of anything other than taking it to a computer shop and pay them a boat load of money to figure it out for me.  SIGH!

 

Thanks.

[Magniphacent]

As an addition, here is a video of the water (I have a cleaning additive it it for the first run - if I can get that far) on each side of the GPU water block.  The pump is running and then, when I turn it off, you can see the pressure release.  This tells me that it is trying, but cannot push it for some reason.  I have checked already pre-cleaned my rads and have inspected the lines to make sure that there are no kinks.

[-rascal-]

@Magniphacent did you jump the spare 500W PSU?

Just connecting the molex to the PSU, and flipping the switch on the back of the PSU won't fire up the PSU.

You need to short the Green PS-ON pin with a black ground pin.

 

EKwb, EVGA, and even certain high-end motherboards will come with a "bridging plug" ... this like.

https://www.ekwb.com/shop/ek-atx-bridging-plug-24-pin

 

You probably also have air / air pockets trapped in your loop.

May need to tilt / shake your radiator / waterblocks to get them moving.

 

EDIT:

I'm just also looking at the photo of your loop.

How is it all hooked up?

How is the liquid suppose to flow (e.g. direction) ?

 

EDIT 2:

Is this correct? Or am I seeing it wrong?

 

[Lipe123]

1 hour ago, Magniphacent said:

As an addition, here is a video of the water (I have a cleaning additive it it for the first run - if I can get that far) on each side of the GPU water block.  The pump is running and then, when I turn it off, you can see the pressure release.  This tells me that it is trying, but cannot push it for some reason.  I have checked already pre-cleaned my rads and have inspected the lines to make sure that there are no kinks.

IMG_3185.MOV

From this video it looks like a blockage... hard to say though. Do you have a high point that's open to air so the air can be pushed out as you fill?

[Alphabear1978]

have you tried any elimination,  Try with res/pump to bucket with bucket slightly higher then combo, then combo back to the top if the reservoir, then add each component.  I know it is a lot of work but at least you might see when and where the problem start to occur.  Also try to bypass some areas.  Let say your CPU block is keeping you from going further, try the mobo first.  You also don’t need to close the loop, just use distilled water and a bucket or water jug to pickup the water again

[Magniphacent]

13 hours ago, -rascal- said:

EDIT:

I'm just also looking at the photo of your loop.

How is it all hooked up?

How is the liquid suppose to flow (e.g. direction) ?

 

EDIT 2:

Is this correct? Or am I seeing it wrong?

Thanks, again for the advice, everyone.

 

-rascal-   Good job on following the loop.  I realized that I should have done the same thing when I posted the picture.  You got it correct.  The new PSU didn't come with a jumper, but I grabbed one from my installed PSU.  I also have a meter on the power to make sure that I am getting correct voltages.

 

Lipe123   I am keeping the fill port open to bleed out any air.  But also notice that, when I pour in water, I have to open a blank fitting on the top of the reservoir to get it to fill or it just backs up in the fill funnel.

 

Alphabear1978   This is a great idea.  I can disconnect the loop at different locations and drain to a bucket.  That may be my next step.  I am glad that I bought an extra 10' roll of tubing as I have to cut the line to get the fittings off (they are not quick-disconnects and have to twist off, which 'twirls' the tube).  I do have one reservation, tho... if the water does not flow when I open the port at the GPU block (which is the first device), I am not going to get too far.  Guess I will start with the output of the pump.

 

I will let you all know how it goes.  Thanks, again!

[Arphis]

@Magniphacent Before you take your whole loop a part, have you actually tried to just blow through it? I would disconnect the pump and the reservoir then grab some papertowel as a mouth piece and try to blow air from the top tube into the system. Careful for residential water.

 

If you cant get air to passthrough there will be a blockage.

 

Also did you inspect the flow indicator for possible dirt/pieces of tubing that could jam it?

[MaratM]

Looks like your pump is side mounted, the inlet is in the midle and the outlet is on the side, and it sounds like it got some air trapped at the impeller. You are using soft tubes, thus unscrew the pump from the mounting bracket and twist it in a way that the inlet port is up. This way all the air inside the impellar will be forced to move out from the pump and that might solve you problem.

here is a picture of my d5 and it was a pain to bleed (even to start the flow)

[Magniphacent]

47 minutes ago, MaratM said:

Looks like your pump is side mounted, the inlet is in the midle and the outlet is on the side, and it sounds like it got some air trapped at the impeller. You are using soft tubes, thus unscrew the pump from the mounting bracket and twist it in a way that the inlet port is up. This way all the air inside the impellar will be forced to move out from the pump and that might solve you problem.

here is a picture of my d5 and it was a pain to bleed (even to start the flow)

Yes, it is side mounted.  The inlet is horizonal and the outlet is also horizonal from the side.  You just cannot see it as the GPU is in the way.  The trapped air makes sense as it makes a screaming sound when I first power on the pump and then settles down to a barely noticeable hum after about 10-15 seconds.  I was thinking of taking the bottom section off the Monsoon MMRS reservoir and installing the pump vertically in that section (will just need a few more parts and try to engineer a new location for the drain valve).  In the mean time, I will try to find a location to mount the pump with the inlet up - if nothing else but to test it.  Thanks.

[Magniphacent]

2 hours ago, Arphis said:

@Magniphacent Before you take your whole loop a part, have you actually tried to just blow through it? I would disconnect the pump and the reservoir then grab some papertowel as a mouth piece and try to blow air from the top tube into the system. Careful for residential water.

 

If you cant get air to passthrough there will be a blockage.

 

Also did you inspect the flow indicator for possible dirt/pieces of tubing that could jam it?

I will try this if a vertical mount to the pump does not work.  Thanks!

[-rascal-]

7 hours ago, Magniphacent said:

Thanks, again for the advice, everyone.

 

-rascal-   Good job on following the loop.  I realized that I should have done the same thing when I posted the picture.  You got it correct.  The new PSU didn't come with a jumper, but I grabbed one from my installed PSU.  I also have a meter on the power to make sure that I am getting correct voltages.

 

Lipe123   I am keeping the fill port open to bleed out any air.  But also notice that, when I pour in water, I have to open a blank fitting on the top of the reservoir to get it to fill or it just backs up in the fill funnel.

 

Alphabear1978   This is a great idea.  I can disconnect the loop at different locations and drain to a bucket.  That may be my next step.  I am glad that I bought an extra 10' roll of tubing as I have to cut the line to get the fittings off (they are not quick-disconnects and have to twist off, which 'twirls' the tube).  I do have one reservation, tho... if the water does not flow when I open the port at the GPU block (which is the first device), I am not going to get too far.  Guess I will start with the output of the pump.

 

I will let you all know how it goes.  Thanks, again!

 

2 hours ago, Magniphacent said:

I will try this if a vertical mount to the pump does not work.  Thanks!

 

If you need to open a fitting at the top, then that is another sign of air getting trapped in the loop / air not circulating out.

 

It makes a bit more sense now that you mention pump is side / horizontally mounted.

If your pump was to be vertically mounted, with the reservoir directly feeding it from above, it would be more typical ... like what you see in water-cooling guides on YouTube.

 

Let us know how it plays out.

[Magniphacent]

OK...  So, I disconnected the pump from the mounting and I tipped it inlet-up and it still would not push the water.  I then pulled it out (I felt like a doctor pulling the heart from a chest).  I brought the inlet below the bottom of the reservoir and turned it on.  Voila!  The water pulled from the tank and pushed through the system.  I continued to fill and it is now running without a problem - and NO LEAKS!.  With the tank being so big, I cannot mount the pump under it as I would have to hide it in the 'basement'.  I am going to let it run with the cleaner and then order a new set of Monsoon MMRS parts to mount the pump in the bottom section of the tank.

 

The important thing is that it is now working.  Thank you EVERYONE for not giving up on me!

[Lipe123]

aah your pump must have had a bubble inside it and just couldn't get the water going, it should be fine now that it's primed.

 

Glad you got some flow finally, much closer to a working system now

[MaratM]

once the pump is filled up with water you can mount it back there it was in the beginning 

[Magniphacent]

As a follow-up, I got the hardware I needed from Monsoon and mounted the pump at the bottom of the reservoir.  After building it and rebuilding it several times (long story that I REALLY don't want to get into), I got it all put together and filled.  Running without leaks first time.  Good flow and bubbles took care of themselves.  RTX208TI survived the transition to the water block.  The computer and I are VERY happy.

 

FYI, my CPU ran between 40-45 degrees with the stock fan stack that comes with the AMD CPU and it is now running at 31 degrees.  Running Uninine2 Superposition before gave me reading for 50 degrees on my CPU (now 38) and the GPU used to get to 60+ with stock fans and now maxed at 48 degrees

 

Good times!

 

Thanks, again!