I've been industriously working on my watercooling rebuild as my old case was getting a bit warm. I've got: Corsair 700D case Corsair HX1000 Asus RIIE 6 GB of Corsair XMS ram I7 970 2xCrucial M225 HD's for my OS (Raid 0) 2xVelociraptor 300Gb HD's for my DB work (Raid 0) (which in turn gets periodically backed up to my server) 2xGTX 480s with EK waterblocks X58 "Northbridge" water block Swiftech Apogee XT Koolance RP-452X2 Dual 5.25" Reservoir Single (Dual Loop) Phobya 3.120 (double thickness) rad Swiftech 2.120 rad Phobya 1.120 (double thickness) rad and about a dozen Elbow fittings to make things nice and tidy. My first assembly job didn;t quite go to plan last night. The route I'd planned was: D5 (in res) -> 3.120 -> CPU -> X58 NB sink -> 2.120 -> GTX480(1) -> GTX 480(2) -> 1.120 -> res I cranked it up to leak test last night and....... well nearly nothing to be honest! it leaked a little on a couple of the fittings, which I then tightened up, but my main observation is that I could barely make out any flow. I do have probably a dozen or more elbow fittings in the loop which I didn't think would have a SIGNIFICANT effect on the flow rate - this is looking to be an error in my judgement. Should I just remove all the L shaped elbows and retry, or am I trying to pump too much with one pump? I have a second D5 which I have ready set up, so I could split the loop, but I wanted to use this to pump the contents of the Res through a bespoke cooler system I was going to work on. Any suggestions please food folks? I'm a bit gutted to be honest... Many thanks in advance Tatty
Swiftech XT waterblock frustrated at this seemingly low flow rate, I dismantled my loop from the end, and checked the components in doings so - My apogee XT seems to be blocked up with a bit of "gunk". I have now cleaned and rebuilt loop and "blow" through using my mouth - seems that this block is quite constrictive! Certainly WAY more so than my GTZ which I upgraded from a little while back. Turned the pumps on and I can hear the pump pumping, but there is no froth. I'm thinking I might go back to the GTZ. Can anyone confirm that the flow rate through an XT SHOULD be this low, compared to a GTZ?
You might still have a 'leak'? I had this problem with my watercooled system, one of the fittings wasn't brilliant, it didn't actually leak but seem to hinder the flow rate (air pockets were constantly circulating)... I sorted out the fitting and all was well.
I've not powered up the system yet. Unwilling to do so while it's in it's current state. Maybe tonight when I've got some evidence of flow out and returning, rather than just the sound of the pump to go on. The flow rate with the XT in the loop seems to be the issue. I'm going to measure the rate with it in, and then with it out, then with the GTZ in to see what the difference is. Just comparing the "resistance" when you blow (using your mouth) through the XT vs the GTZ, the GTZ appears to be massively more "flow-some", so I don't believe the problem is fittings related. Going to switch them out tonight and compare with a small plexi res in the loop so I can SEE if there's any flow.
I have a XT and my flow rate is still high. I just a Cyclone res and it flies round the thing. You said it had gunk it is, were you using colours fluid? Try popping off the little top overs (the one that allows you to use extra large compression fittings) and make sure there no gunk there. Is your pump turned up full?
XT flow rate Hi Pete - thanks for feeding back. Everything that I've seen seems to point to this being the XT, but to be honest I've never used most of the kit that's in the rig. I'm using Feser 1 non conductive (red) liquid at the moment. Brand new. Waterblock was used in another rig where I didn;t notice this phenomenom before - I'm migrating from a HAF 932 to a corsair 700D, and picking the best bits to go into the Corsair. I've run white wine vinegar through the old loop to clean it out, and after leaving that for 24 hours, then running neat Distilled water to clean it through, changing after a few hours. The "gunk" in the cpu block was directly under the removable inlet plate that you speak of, on top of the pin matrix. I've scrubbed it with a toothbrush (hers ) to remove and it looks clean to the eye now. I've then re-assembled (the washers wouldn;t go back easily) and then blown through and compared how that feels vs the Swiftech GTZ, and it's massively more restrictive. I also forgot to put back in the washer that sits on top of the matrix, which forces flow through the matrix, so I guess that it's more "flow-some" now as a result of my shoddy re-assembly skills. When I've plumbed the XT unit in, it's not leaked overnight, so it looks like I'm dry, but I can;t see any of the air-bubbles that I would expect to see coming back. In fact there's no visual sign of flow anywhere. The pumps nicely quietly whirring away, and it sounds like it has fluid in there. The pipes out of the res/pump feel like they have fluid moving, but pinching the tubing doesn;t increase with whiney tone that I would expect. The pump is off e-bay, so is an unknown. I dismantled it and it looks to be in excellent condition, and has been cleaned. I think tonight I'm going to take the res/pump out of the loop and make sure that it is pumping through a dummy loop. If that doesn;t show a fault, then I'll switch out the CPU sink for my older GTZ and see if that makes any difference to the flow. Current loop order is: Koolance Res (450X2) -> Pump (D5) -> Phobya 3.120 -> Swiftech XT -> EK X58 Block -> Black Ice 1.120 -> res I can blow through the connections that affix to the back of the res, and this feels ok blowing from the res up to the XT, and likewise from the XT back to the Res, so I don;t have a pinched or restricted pipe. Maybe I'm just panicking over nothing, but in my prior system it ran: XSPC Twin bay res -> Pump (D5) -> Black Ice 1.120 -> Swiftech XT -> Generic 1.140 rad -> GTX 480 EKWB -> GTX 480 EKWB -> Swiftech 2.120 -> Res and I could see flow returning to the res, and when you pinch the tubing you could clearly hear this stress the pump. Could it be the pin matrix is completely bunged up in the CPU block, or a defective pump? Pump is on 5 and you can hear it running (just). Any suggestions?
Switched out the XT block for the GTZ block flow rate was notably higher. Bubbles flowing through back to the Res, so problem solved. Going to put in my second loop tonight and over the weekend (need to butcher my case, which I'm putting off doing) Sounds like my XT is blocked. Anyone got any ideas how to clean the micro-channels of the block? Also, any idea where I can get replacement washers for the block?
Contact Swiftech direct, you may get the bits for free http://www.swiftnets.com/ A toothbrush is good for gentle cleaning. Lemon juice and salt is good for cleaning more ingrained deposits. However check the bits every 30 minutes until clean as it can be quite aggressive over longer periods.
You might be getting a low flow rate because you are running everything through the one loop. You have 2 reservoirs, 3 rads, 2 GPU blocks, a restrictive CPU block and a NB block as well. The D5 is a good pump but you can only push so much so fast through that labyrinth of a loop! Try adding a second pump somewhere on the loop so the first one isn't working quite so hard or a second seperate loop and pump incorporating just the GTX480 blocks. The elbow fittings should not matter much to your flow rate, I would be more concerned about the pumping power. Do you have room for another pump? There are also dual pump tops you can add that fit directly into the front bays of the case. EK has one and I think Bitspower as well.
