Overclock.net › Forums › Case Mods & Cases › Builds & Case Mods › Build Logs › [Build Log] - IRONBEAST - A fully water cooled, X99, Quad GPU, 900D build
New Posts  All Forums:Forum Nav:

[Build Log] - IRONBEAST - A fully water cooled, X99, Quad GPU, 900D build - Page 110

post #1091 of 1313
Quote:
Originally Posted by d0mini View Post

This project is just wonderful. That looks amazing, I love it, and it's just the back of the thing. Good luck with your leak testing. I recently completed my first custom loop; it was nothing compared to the scale of this towering beast, but it did leak, and I had to go through the the painful process of draining, drying, tightening and retesting.

It was painful, stressful and at times downright terrifying. If this leaks, I'm sure it will be hard on you, but from my meagre experience, just push through it. If it leaks, turn it off, clean up the worst of it, then take a break. Just do anything else for a bit, then come back to it.

I'm sure you'll be alright, you've already gone through one heck of a time getting it this far. This is a drop in the ocean in comparison to the amount of time and planning you've put into this already! biggrin.gif

This is why u air leak test first.

Nice progress on build thumb.gif
Edited by catbuster - 8/9/16 at 12:13am
    
CPUMotherboardGraphicsRAM
i7 4770k Evga z87 Stinger mITX Evga gtx 980 sc acx Crucial Ballstix Ellite 2x8gb 1866mhz 
Hard DriveOptical DriveCoolingOS
Samsung 850 pro 512gb Nice joke bruh Corsair h100i gtx Windows 10 
MonitorKeyboardPowerCase
Asus vs228 Poker 2 mx red SilverStone SX600-G Compact Splash #55 
MouseMouse Pad
Nixeus Revel black Steelseries Qck M 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 4770k Evga z87 Stinger mITX Evga gtx 980 sc acx Crucial Ballstix Ellite 2x8gb 1866mhz 
Hard DriveOptical DriveCoolingOS
Samsung 850 pro 512gb Nice joke bruh Corsair h100i gtx Windows 10 
MonitorKeyboardPowerCase
Asus vs228 Poker 2 mx red SilverStone SX600-G Compact Splash #55 
MouseMouse Pad
Nixeus Revel black Steelseries Qck M 
  hide details  
Reply
post #1092 of 1313
Just do the leak test using a psi gauge. Although that's so many fittings, I wish you luck. Really this should of been bp sponsored. Dat fitting cost.
post #1093 of 1313
drool.gif

This build... man oh man. Th.... never mind.
post #1094 of 1313
Quote:
Originally Posted by OCDesign View Post

(Or alternatively we can admire how lovely all my lighting stuff looks when it’s reflected in an enormous puddle…)
Looks amazing! Like mentioned above, you could easily do an air test to avoid filling the loop. Even if the loop is not leaking, I would imagine that draining it would be quite the effort... Another advantage is that you could get a higher system pressure, thus stress testing it.

Aquacomputer has a kit for doing this: http://shop.aquacomputer.de/product_info.php?language=en&products_id=3415

You might find a UK supplier for this aswell.
post #1095 of 1313
Thread Starter 
A small follow-on update to my previous post.

The temporary connections required are now in place and the loop is closed, pending an initial integrity and flow test of the rest of the pipework before the system hardware goes in.


For the purposes of this test, and in the absence of anything to fix them too, the motherboard and CPU waterblocks have been attached directly to the motherboard mounting plate, courtesy of some small tabs of self-adhesive Velcro (as with the lighting blocks). All three blocks are linked up with basic point-to-point connections using the Primochill tubing, though my obsessive-compulsive tendencies do bristle a bit at the fact I had to use fittings of two different colours…(Yes, I know. It’s a travesty).


That’s the repositioned Aqua Computer High-Flow I mentioned previously. Above it and to the right will be the PWM hub for the top fans (two of the plugs you can see, two will require extenders). Below it I intend to fit the internal USB hubs (for the Aquaero, Poweradjusts, etc.) and likely the holder for the I/O connections to the motherboard and the DisplayPort connections to the GPUs will go here too. Below that will be the Aquaero and the three Poweradjusts, plus the relevant power cables up from the PSU and down to the pumps. The Arduino microcontroller to run the lighting setup will also fit in here too.

(You may note some scratches in the paintwork there. This is because the 5.25” drive bay mounts for the Aquaero and Poweradjusts push-fit and don’t need to be screwed in. By sheer fluke that width is just tight enough that they will fit in and securely hold in place, very firmly, just on friction.)


This is where the Aquaero waterblock would be, but for the moment that has been bypassed completely - in order that I can keep the controller safely out of the way too. A set of my QD3s from the radiator testing are bridging the gap, with the addition of a couple of extra fittings necessary to ensure the tubes don’t kink.


For the EK GPU waterblocks there was a bit more of a problem, since they’re quite heavy and there is nothing in place yet that I can fit them to. Therefore they have had to go on the floor in front of the case, and for convenience they have been hooked up to their connection points with some more QD3s (so it can be easily removed if and when I need to move the case). The manifolds presented a bit of a problem, since some of the fittings that I originally used to make them (in order to show the idea to you lot – way, way, back) have now been used elsewhere in the build. However the actual splitter-integrator part of those assemblies I do still have ready and waiting, thus letting me set up the waterblocks in this somewhat unorthodox fashion. They’re supported by bits of the packing foam, some spare pieces of the sound deadening material, and two of the empty boxes that contained the Monsoon fittings. It was a complete pain to get that connected up with the parts I have available (not much finger room) but it should prove adequate enough to let me do this test. (That said, I would have liked to have been able to fit the four MPS flow sensors – to monitor the flow split and find out how well the manifold idea works – but I just haven’t got the parts available to set it up. I guess that test will have to wait.)


Thanks very much for the comments everybody. All your encouragement really does help so much in keeping me pressing forward with this insane scheme. And you needn’t worry, I haven’t forgotten the advice to air leak-test first. Amongst myriad other things there is already an air pump and pressure gauge on order - that exact kit in fact – which should be here sometime in a few days. Nonetheless I am still utterly convinced that the beast is going to end up sitting in the middle of its own beautiful reflecting pool the moment it gets so much as a drop of water in it…
post #1096 of 1313
....Holymoly. That was a fun read.

Consider me Subbed.
LunarTide
(23 items)
 
  
CPUMotherboardGraphicsGraphics
i7-5820K Asus Rampage V Edition 10  GeForce® GTX 1080 WINDFORCE OC GeForce® GTX 1080 WINDFORCE OC 
RAMHard DriveHard DriveHard Drive
Corsair DOMINATOR Platinum 16GB  Samsung 850 Pro Series 2.5" Solid State Drive Samsung 850 Pro Series 2.5" Solid State Drive Western Digital Black 3.5" 7200RPM Internal Har... 
CoolingCoolingCoolingCooling
EK-RES X3 ​​250 X2 EK-FC1080 GTX G1 - Nickel x2 EK-Supremacy EVO - Nickel EK-FC1080 GTX G1 - Acetal+Nickel x2 
CoolingCoolingCoolingCooling
Aqua Computer D5 pump pwm x2 Corsair Air Series SP120 Quiet Edition x8 Corsair Air Series AF120 Quiet Edition x2 Black Ice ​Nemesis Ra​diator GTS​ 480 XFlow 
CoolingMonitorKeyboardPower
Black Ice ​SR2 Xtreme​+ 480 MP M​ulti Port ​Rad... ASUS ROG Swift P278Q G-Sync Monitor Logitech Orion Spectrum EVGA 850W 80+ Gold Certified Fully-Modular ATX ... 
CaseMouseAudio
CaseLabs SMA8 Two-Tone (White & Black) Logitech Chaos Spectrum Sony Shake33 2200W 
  hide details  
Reply
LunarTide
(23 items)
 
  
CPUMotherboardGraphicsGraphics
i7-5820K Asus Rampage V Edition 10  GeForce® GTX 1080 WINDFORCE OC GeForce® GTX 1080 WINDFORCE OC 
RAMHard DriveHard DriveHard Drive
Corsair DOMINATOR Platinum 16GB  Samsung 850 Pro Series 2.5" Solid State Drive Samsung 850 Pro Series 2.5" Solid State Drive Western Digital Black 3.5" 7200RPM Internal Har... 
CoolingCoolingCoolingCooling
EK-RES X3 ​​250 X2 EK-FC1080 GTX G1 - Nickel x2 EK-Supremacy EVO - Nickel EK-FC1080 GTX G1 - Acetal+Nickel x2 
CoolingCoolingCoolingCooling
Aqua Computer D5 pump pwm x2 Corsair Air Series SP120 Quiet Edition x8 Corsair Air Series AF120 Quiet Edition x2 Black Ice ​Nemesis Ra​diator GTS​ 480 XFlow 
CoolingMonitorKeyboardPower
Black Ice ​SR2 Xtreme​+ 480 MP M​ulti Port ​Rad... ASUS ROG Swift P278Q G-Sync Monitor Logitech Orion Spectrum EVGA 850W 80+ Gold Certified Fully-Modular ATX ... 
CaseMouseAudio
CaseLabs SMA8 Two-Tone (White & Black) Logitech Chaos Spectrum Sony Shake33 2200W 
  hide details  
Reply
post #1097 of 1313
Thread Starter 
Tiny little update:

Having spent the last week sitting around, idly tinkering with stuff and waiting for my latest shipment of stuff to get here, it has now arrived and I can make some more progress.


On the right is the Bitspower Flow Indicator that I mistakenly ordered before, and on the left is the new replacement Bitspower Flow Sensor that I actually meant to get. The latter has a header on the back for an RPM signal, whilst the former is just a visual display. The two have now been switched over and the indicator will be utilised elsewhere later on.

(Such a silly mistake to have made, they are so obviously different from each other. I do hope you’ll overlook it.)


More importantly I now have the Aqua Computer pressure testing kit, so I can now air leak test the loop before trying with water. I hope to try and get that done sometime tomorrow and update you on the results - although I will be frankly astonished if there are no problems at all, so I’ve no idea when a first test with liquid may follow.

However you may also find it amusing to know that my prediction has already come true - and that last week Ironbeast did indeed end up sitting in the middle of an enormous puddle. Nothing to do with the loop integrity, and everything to do with the fact that the moron building it was too busy obsessing over missing O-rings and clumsily knocked a jug of water off the table... (I really am hopeless sometimes.) No harm done fortunately, beyond some slightly soggy bits of paperwork and a lot of time wasted finding all the little shards of glass spread over the floor. Fingers crossed the real water testing won’t now produce a similar result, though I’m not holding my breath...

Quote:
Originally Posted by LeviathanVI View Post

....Holymoly. That was a fun read.

Consider me Subbed.

Thanks very much, welcome to my little project!

I know my writing hasn’t really maintained the same standard of quality as it had at the start (there's just a lot less time to devote to drafting build log entries properly since we got to the stage of actually building this thing), but I do try my best to keep it a reasonably good read. Glad you're enjoying it.
post #1098 of 1313
I'm glad to see that you aren't one of those people that straight up puts water in the loop. I have seen many people destroy their systems by adding water and it then going everywhere. Good Job with the progress thumb.gif
post #1099 of 1313
Subbed!

I've been reading your log for the past months! It's stunning. Keep up the fantastic work!
post #1100 of 1313
Thread Starter 
Small update:

Ran into a couple of issues over the last few days. Firstly I have had to redo the pipework for the Bitspower Flow Sensor, going from one long section to two shorter ones.


The Flow Indicator fitted in the space very nicely, as you saw before, and the Flow Sensor also fitted ok with adiquate extra room to accomodate the header on the back. However after my last post I discovered that, once you plug the cable in, it protrudes from the back just a little bit too far, hits the pipe behind, and as a result the whole thing doesn’t quite sit properly anymore. Therefore it was necessary for that section to come out yet again, and for me to shift the unit further across (to the point where the transfer pipe behind angles downward) in order to give it enough room. I don’t have any spare extender fittings the right size, so it had to be done with two - annoyingly short - lengths of acrylic which were a distinct pain to cut and glue. (Because both of those gold collars have to slide along the tube before you glue the second cap on, it was getting really rather close to the minimum limit of what those fittings will give you space to do, especially in the case of the slightly smaller one on the right hand side.) A bit of a faff, but it’s all in place again now.

This brings me to the second issue, the air leak test. Suffice it to say that the system would not pressurise - at all - and I have spent the last few days trying to figure out why. Ultimately I tried unplugging the connections down into the subfloor section and testing that part of the loop in isolation. The pressure went up and then rapidly dropped back to zero, over the course of about 5 seconds. Unable to figure out what was wrong with it, and getting very frustrated, I switched tack and attempted to test the manifold/GPU block section in isolation – largely in a vain effort to try and cheer myself up by finding something that worked. Exactly the same problem. As you can probably imagine that started to make me rather suspicious. So I switched to testing just the basic Aquaero waterblock alone, and then just to using a capped-off section of tube. Again, same problem.

Unless I am being incredibly obtuse about something, I fear that pressure gauge isn’t working properly. Very occasionally the reading drops off much more slowly (without any change to what it’s connected to) taking about a 30 seconds to reach zero - but it always depressurises. By chance we had a bike tyre pump around which uses a compatible fitting; hooked it up to the gauge, and (very gently, because it’s a much bigger pump) pushed it up to the recommended pressure as per instructions. No luck. As such it appears the air pump is probably ok but that the gauge’s input valve isn’t sealing properly. I thought it might be a problem with the little pressure release button not returning fully, but it doesn’t seem to be. So it looks like it will be necessary for me to get a replacement before continuing with the test.

Yet more delays…

Quote:
Originally Posted by MinyMan644 View Post

I'm glad to see that you aren't one of those people that straight up puts water in the loop. I have seen many people destroy their systems by adding water and it then going everywhere. Good Job with the progress thumb.gif

Thanks. In truth I probably would have done, but fortunately there are lots of people here who give much more sound advice - the least I can do is listen to it!

Quote:
Originally Posted by Valthar View Post

Subbed!

I've been reading your log for the past months! It's stunning. Keep up the fantastic work!

Thanks very much, you’re too kind. I do get very frustrated when progress gets bogged down with delays like this, and it always cheers me up so much whenever I hear that people are still interested in my ludicrous project - even though it’s been slowly lumbering along at snail’s pace for all this time, and that the tech will probably be antique by the time it’s finished…
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Build Logs
Overclock.net › Forums › Case Mods & Cases › Builds & Case Mods › Build Logs › [Build Log] - IRONBEAST - A fully water cooled, X99, Quad GPU, 900D build