Overclock.net banner

1 - 20 of 24 Posts

·
Registered
Joined
·
1 Posts
Discussion Starter #1 (Edited)
OK well I took my time, bought the best parts I thought I could ever afford, did what I thought was a considerable amount of research into the subject of watercooling.Ive spent just short of six months putting this together. I could have possibly done it quicker but it was much easier on my budget this way.
The goal of what I was trying to do was to build a system that runs on the new 299 chipset, my target being the i9 7900X. It didn't take long for me to find out about the thermal requirements of this chip, so custom water cooling it would be.
Knowing I would need considerable room in the case for tubing ( my last rig used an AIO Corsair cooler that was cramped even in a “Super Tower” case (Corsair Graphite 760 white)..I bought the Tower 900 by Thermaltake.Next came the board, not looking for just the most expensive but for what fit my needs and had the best build quality within my budget I opted for the EVGA X299 ftwk.I held over my 32 g of Corsair Dominator RAM @3000 MHz and added another 32g of the same.I also reused my Samsung 960 pro M.2 drive.Lastly I purchased a EVGA 850 GT PSU.
Now for the water cooling parts.
Having no experience with this I just read a lot of reviews and it seems to me that EK WB was about the most solid company I could buy from.I got a EK XRES 140 REVO D5 Pump/Res combo, an EK Coolstream 240 radiator and an EK Supremacy EVO Nickel CPU block.I also bought all EK 16mm compression fittings and EK HD PETG tubing.
So it turns out that the chip was the last thing I bought,but once I had it I finished my build and began to attempt to fill my loop.
This where the trouble began.While I dont have any custom water-cooling experience I DO have considerable Coral Reef aquarium building/husbandry experience so Im no stranger to pumps and tubing..however Im baffled by the performance of my supposed "High Performance" pump/res combo unit.Since there was NO instructions as to initial operation I read online that you are to connect the pumps power cord to a molex from the PSU while jumpering the 24 pin Motherboard power cable,ignoring the 4 pin PWM fan cable for the time being allowing the pump to perform at 100% in order to fill the loop and purge any air within.
This I did, but no matter what Ive tried the coolant will only rise about halfway up the first tube,no where near emptying the res as one would fear..
If youve read this much I thank you.Before you began to formulate an assistance post allow me to list the things that I have already tried to avoid me wasting anyone's time :/
-As I said the first try was Pump power to Molex with no PWM connection,24 pin cable jumpered
-Next I tried connecting the PWM cable to my sons (2 days ago my) pc at the CPU fan header ,making sure in the BIOS that it was set for PWM not DC.
-I tilted shook and twisted that huge case in every possible direction in an effort to purge all the air out after reading that then and only then would my pump work...
-I drained the loop removed the Pump/Res and connected pipes to the In and Out ports to see if it was indeed functional at all.It performed as expected,It had no trouble pushing water out of the res and into a bucket.
-I even made some hackish connector with a paper clip when I read you could connect the green wire from the PWM to the red (5v) of the molex to trick the pump into firing up...
-Lastly (and probably foolishly) I connected my former AIO cooler after reading that the 2066 architecture was in no way different from the 2011 and that the pump would physically fit over the CPU.However when I booted it and got to the BIOS the temps hit 100 C even after re-cleaning the old thermal paste and applying new...
Im at my wits end here.Besides Intel's choice of cheap paste to attach the IHS I realize Ive no one to blame but myself for this setback.But Ive gone this far , Ive got to see this through to the end.
If anyone thinks they may know a way to coerce my pump into performing at 100% (or close to it) Im all ears and would be eternally grateful, or if you see any other part of my setup that may be causing the problem feel free to tell me.At this point my options are few but expensive..and Id like to work with the equipment Ive already bought.
Thank you !
Video of situation https://drive.google.com/open?id=1mIVWRf9CIzyTHVU-wp0gwzatwuUfCbLX
 

·
Registered
Joined
·
371 Posts
what happens if you thread some tubing to the top of your res and provide a head of fluid as the pump is running
have you got a blockage somewhere in the build? a kink in tubing or something?
 

·
Registered
Joined
·
371 Posts
is your flow meter the correct way around and free turning?
 

·
Overclocker in training
Joined
·
14,091 Posts
Hi,
Weird I have almost the same pump-res combo filling is always a breeze
Have to think there is a pressure point in the loop somewhere flow meter stuck....
evga 850 well I do have a 850P2 series it was okay but not on my x299 rig I have a 1200P2 on it because the x299 is a juice sucker :)

The 850P2 only has one port on it for a molex is there anything else connected to it too or just the pump ?
I couldn't get the video to work no flash installed could you just extract one image of the entire case and loop order and upload it here

A 240se rad is not really going to do much for cooling the 7900x I have one
A delid is needed regardless to get rid of the Intel pigeon poop they used of course that whacks Intel 3 year warranty too.
Silicon lottery is who I used to delid my 7900x they give 1 year.
 

·
Newsfiend
Joined
·
2,423 Posts
Picture of the setup?

Something is very wrong for a d5 to not be able to push up a the first tube. Even at low speeds this should be doable no problem with a single block and rad.
 

·
Registered
Joined
·
1 Posts
Discussion Starter #6
well I have tried adding more fluid to the res while its running,that just filled it up .I dont believe there is a blockage or a pinch in the line ,Ive got very large pipes 16mm OD.But I will check for that when I get back home,Thank you!

The only picture I have just shows the first part of the loop.But here it is. https://drive.google.com/open?id=1HCr5pFgli-NTV0NuN7afkavJ-U57M8Kv

Yes I do need to up the PSU to about 1000w, but no nothing else was connected besides the 24 pin.

Im sorry but what do you guys mean by "flow meter"
 

·
Crayon Evangelist
Joined
·
7,609 Posts
took a S/S of the video so we could discuss.


I have never used this pump/res combo, so I dont know if I have allot to add. I am used to pump and res being different units, so I dont know what the guts of this thing looks like.


Is the impeller just exposed inside the res? I am used to 'in' lines feeding into the res, not directly into the pump. I would think that the cap should be on the port (bottom right) that has the 180 bend, and the inlet line should be at the top of the res...but again, I dont know the guts of this combo unit...it just looks like you have it connected at two out ports.



Does it work if you make a simple loop without any components?


Edit to add: Nevermind, it does look like you have the in/out correct
 

·
Registered
Joined
·
1,451 Posts
I'd suspect something is wrong wtih the pump. I've always had my res empty in seconds (larger res too) when I turn on my pump. THe easiest way to test the pump is to pull it out and build a simple loop outside the rig (don't even need blocks, just get tubing).
 

·
Registered
Joined
·
72 Posts
I would say bad pump. Looking at your video it appears either the pump is severely cavitating (not likely as reservoir on top of pump should give plenty of fluid directly to the pump intake) or the pump impeller is binding up and not spinning or the pump motor is not working correctly. The impeller on the pump rests on a ceramic bearing and is driven by magnetism to spin the impeller. The impeller is not directly connected to the motor of the pump with a shaft. It is actually part of the motor itself. The pump motor body is the stator and the impeller is the commentator. Disassemble the pump and remove the impeller and check to see if any debris is binding up the impeller on the pump body cavity. If not, you have a bad pump that is not driving the impeller as it should (creating enough magnetism to spin the impeller correctly). This would follow what you said that when you removed the pump from the loop to test, it flowed water as there would be no resistance for the pump to work against and would flow water through it although probably not flowing as much water as it should be (you would need a flow setup to measure and test). When the pump is in the loop, the pressure resistance of the water block, radiator and water lines would create enough resistance to keep the pump from pushing fluid through it.
 

·
Big Radiators!
Joined
·
825 Posts
I have noticed something in the video. The CPU waterblock's upper tube is partly filled but no fluid is filling the lower tube. The liquid should freely flow from the upper tube into the CPU waterblock and then down the lower tube. I suspect you have a blocked CPU waterblock or a blocked radiator.
Most likely the radiator is jammed. I suggest you disconnect the radiator and simply blow some air into it with your mouth, see if it works.
 

·
Registered
Joined
·
72 Posts
I have noticed something in the video. The CPU waterblock's upper tube is partly filled but no fluid is filling the lower tube. The liquid should freely flow from the upper tube into the CPU waterblock and then down the lower tube. I suspect you have a blocked CPU waterblock or a blocked radiator.
Most likely the radiator is jammed. I suggest you disconnect the radiator and simply blow some air into it with your mouth, see if it works.
This would happen if the pump is not properly working........or massively cavitating.
There is quite a bit of resistance in the water block from the nozzle plate in the block and if the pump is not spinning the impeller correctly, there would not be enough fluid pressure coming out of the pump to push the fluid with enough force to over come the resistance of the water block. Not seeing any fluid coming out the outlet port on the block is because of the nozzle plate restriction.
The radiator and water block he has are pressure tested for leaks at the factory so there should be no blockages in either, but worth checking out just in case as you never know. Stranger things have happened :).
If you look closely at the hard tube feeding into the water block, it is not perfectly level (slightly angled down from 90 degree bend at the top of it before bending another 90 degree into water block inlet). This would cause fluid to be trapped there.

Still think it is a bad pump......
 

·
Registered
Joined
·
72 Posts
I will also say your loop order is not the most efficient either. You have your radiator at the end of the loop feeding back into the res.
It should be the first from the pump outlet.
The pump motor will generate a little heat into the fluid so the radiator should be first from the pump outlet and before your water block to ensure you are getting the most cooled fluid first into the water block that generates the most heat in the system.
 

·
Crayon Evangelist
Joined
·
7,609 Posts
Just bend a quick loop from in to out and check that the pump is operating.



While you have it apart, try to suck air through the loop.
 

·
Registered
Joined
·
1 Posts
Discussion Starter #16
Thank you so much for all your time and thoughts.When I return home from work I plan to check for blockages and have a look at the pump impeller.I will update as to these findings.
Again,thank you all.
JP
 

·
Registered
Joined
·
65 Posts
Thank you so much for all your time and thoughts.When I return home from work I plan to check for blockages and have a look at the pump impeller.I will update as to these findings.
Again,thank you all.
JP
In your video it looks like you have the fill port on top of the res capped off with a plug. If that's accurate then fill your res again, like you had in the video, and take the plug off before turning the pump on. Might just be all the air in the system that had no where to go. That's the only insight I could offer based on the info. I hope it helps.
 

·
Happy Dad
Joined
·
346 Posts
Any updates?
 

·
Registered
Joined
·
830 Posts
I will also say your loop order is not the most efficient either. You have your radiator at the end of the loop feeding back into the res.
It should be the first from the pump outlet.
The pump motor will generate a little heat into the fluid so the radiator should be first from the pump outlet and before your water block to ensure you are getting the most cooled fluid first into the water block that generates the most heat in the system.

It doesn't matter. Loop temp will equalize.


In your video it looks like you have the fill port on top of the res capped off with a plug. If that's accurate then fill your res again, like you had in the video, and take the plug off before turning the pump on. Might just be all the air in the system that had no where to go. That's the only insight I could offer based on the info. I hope it helps.

This could be your issue. Its a lot easier to bleed the system when you have a port thats open, to let the air escape while you bleed.
 

·
Registered
Joined
·
492 Posts
This could be your issue. Its a lot easier to bleed the system when you have a port thats open, to let the air escape while you bleed.
My initial thought was it makes more sense to use one of the top ports as the inlet rather than the lower suction port and also leave a fill port open to let air out while filling. Never done this sort of setup though so cannot say for sure that would work.
 
1 - 20 of 24 Posts
Top