|
Yeah I was going to ask what weird GPU that is. You don't have to make excuses for your bejeweled RAM, be true to yourself. e: and in being true to myself and my idiot brain I didn't order a CPU block with the big EKWB order. Yay! I ordered one from a Canadian retailer. 
VelociBacon fucked around with this message at 05:25 on Jan 25, 2022 |
#
?
Jan 24, 2022 23:32
|
|
|
|
| # ? May 11, 2024 13:45 |
|
|
saw the discussion way too late, a few questions: Why not a D5 PWM as pump? With a good BIOS it is awesome Why would you ever use rotatable fittings above a GPU or other valuable parts? Imo they are fine on the edges of the tower or in the lower third.
|
|
#
?
Jan 27, 2022 08:27
|
|
|
AutismVaccine posted:saw the discussion way too late, a few questions: I swapped from an EK DDC to an EK D5 pump, noise is more manageable with the D5 at full speed (I believe full speed isn't as fast as some other D5 though) than the DDC on PWM. DDC is just unpleasant sounding, even at same dB. You get insane pressure out of them though, good if you have a very restrictive loop.
|
|
#
?
Jan 27, 2022 10:43
|
|
|
movax posted:Oh no! I have to remove RAM / get the radiator out of the way! What do I do! Yeah, if you're using soft tubing and there isn't a length constraint stopping you, QDCs are simply a must have in my opinion. The fact that no one is going to see your Trident Z Royals bedazzled memory sticks is funny to me because it is the same with any RGB in my system. But it looks pretty slick in there. Anyway my ITX board only has one PCIe slot anyway so yeah it gets the 3080 
|
|
#
?
Jan 27, 2022 15:05
|
|
|
AutismVaccine posted:saw the discussion way too late, a few questions: I asked about it a bit earlier in the thread (I think); I do have an Aquacomputer OCTO so I could control it but it seemed like most people just didn't bother with / care about varying pump speed too much. The fittings... well, I thought I needed them, but I could try it again with STCs sticking straight out of it. Is that bend up out of the water block to hit the rad doable without an adapter? I guess I do have the length to burn on it. I do have more adapters below to get the pump lined up and the res, but those are at the bottom of the loop. Any thoughts on suspending a flow meter in a run of soft-tubing? It's either that or a rotary angled 45 holding it. Not sure I'll be able to get the QDC I want in there either, but I'm trying to get the entire front radiator + pump + res assembly to be one unit I can remove/install and easily split from the loop. movax fucked around with this message at 01:35 on Jan 28, 2022 |
|
#
?
Jan 28, 2022 01:30
|
|
|
movax posted:I asked about it a bit earlier in the thread (I think); I do have an Aquacomputer OCTO so I could control it but it seemed like most people just didn't bother with / care about varying pump speed too much. Is that bend up out of the water block to hit the rad doable without an adapter? Cut a piece of hose and test it out, imo it should be doable. Any thoughts on suspending a flow meter in a run of soft-tubing? Is it possible to just attach it directly to the pump/rad with a small double male adapter or a 90 degree no rotary piece? the entire front radiator + pump + res assembly to be one unit I can remove/install and easily split from the loop. But why? when the loop is running, it is running. Every unnecessary connector kills the flow rate, just like using 25mm rads with only one layer of tubing. Placing the outlet on the bottom and using one of the many holes from the top rad for letting air in.
|
|
#
?
Jan 28, 2022 12:59
|
|
|
I have a bunch of water cooling related stuff up for sale in sa mart if anyone is interested. I’d like to sell it as a lot, but I might be willing to split it up if you buy a decent portion of it.
|
|
#
?
Jan 28, 2022 18:08
|
|
|
AutismVaccine posted:Is that bend up out of the water block to hit the rad doable without an adapter? Yeah I’ll try it this weekend… there’s probably a ton of room for a bend in there because it’s literally empty space. quote:Any thoughts on suspending a flow meter in a run of soft-tubing? I have it directly attached right now with a double-male and a rotary 2x 45… I need it to swing out of the way of the GPU. I might try an offset rotary fitting. quote:the entire front radiator + pump + res assembly to be one unit I can remove/install and easily split from the loop. Serviceability mostly — I know this is mostly gonna stay together for the next few years but I know I’ll be tweaking poo poo and from my cursory research, flow rate drops with various adapters seem negligible… movax fucked around with this message at 20:00 on Jan 28, 2022 |
|
#
?
Jan 28, 2022 19:58
|
|
|
The main reason to minimize the number of fittings in a loop isn't because of the increase in flow restriction but because more fittings is an easy way to nickel and dime a loop to over $1000.
|
|
#
?
Jan 29, 2022 00:25
|
|
|
Indiana_Krom posted:The main reason to minimize the number of fittings in a loop isn't because of the increase in flow restriction but because more fittings is an easy way to nickel and dime a loop to over $1000. I think I’m under $300 in fittings… the QDCs were still worth every penny!
|
|
#
?
Jan 29, 2022 00:48
|
|
|
Hey so I posted awhile ago about my build, it's a 9900k and 3090, with two 420mm rads. I've really been waffling on orientation of the fans on the rads, which will be front and top. I've always had my rads exhausting to the outside of the case, but I worry if I do that with this setup I'll have a pretty serious negative pressure/dust issue in the case. Here's the mockup again but with the non-radiator fan locations marked in purple, and to clarify, the basement plate above where the two bottom fans are will be removed anyways for the radiator so there's nothing really blocking them: So if I set the rads up as exhausts, there are 3 140mm fan locations available for intakes. So I would have like... 6 exhaust fans and 3 intake fans. Even given that the exhaust fans are pushing through radiators, and won't need to be ran very fast due to the amount of radiator I'm running, and I could run the three intake fans pretty high speed without hearing them (Noctua), it just feels like I'll still be negative pressure. If I set it up so air is being pulled through the rads into the case, I only really have one good exhaust fan option, the rear top one. The two fans in the basement area will not really be getting a lot of hot air and probably wouldn't be great as exhausts. I also don't love the idea of keeping every single component in the case warmer than necessary, especially as I still have two HDDs in there. What would you do and why? e: All the fan locations are filtered, even the back 140mm fan has a filter.
|
|
#
?
Feb 2, 2022 02:13
|
|
|
VelociBacon posted:Hey so I posted awhile ago about my build, it's a 9900k and 3090, with two 420mm rads. I've really been waffling on orientation of the fans on the rads, which will be front and top. I've always had my rads exhausting to the outside of the case, but I worry if I do that with this setup I'll have a pretty serious negative pressure/dust issue in the case. Here's the mockup again but with the non-radiator fan locations marked in purple, and to clarify, the basement plate above where the two bottom fans are will be removed anyways for the radiator so there's nothing really blocking them: Personally, I would go front intake, top exhaust on the radiators, then just gave the rear 140 on intake. The bottom two slots will just mess with the flow, I feel. You'll also run into space issues with the max size rads and a decent PSU trying to fit bottom fans. I have 360 front intake, 280 top and 120 rear exhaust rads and I have to run the front fans a slightly higher curve to have positive pressure (testing just by feeling airflow over little gaps)
|
|
#
?
Feb 2, 2022 02:56
|
|
|
BurritoJustice posted:Personally, I would go front intake, top exhaust on the radiators, then just gave the rear 140 on intake. The bottom two slots will just mess with the flow, I feel. You'll also run into space issues with the max size rads and a decent PSU trying to fit bottom fans. I thought about this but thinking about the hot air from the front rad going right up into the top rad makes me feel bad + gross. I agree that the bottom fan location kinda sucks. You might find it interesting, I use the barometer in my S9+ to check things like the pressure level inside negative/positive pressure room in the hospital. It's a sensitive enough ... sensor that just raising the phone from the floor to the level of a desk changes the pressure. I put the phone inside my case and looked at it through the panel and basically I'm dumb as hell because the PC case has all sorts of areas where air flows in and out to reach pressure equilibrium. So I guess if I was really curious I'd light a shishkabob stick and blow it out and watch what that smoke does when I move it around the exterior of the case. e: I should also add that the case can take bigger rads, I'm using 420mm rads but it can actually take 480mm rads so there's a fair bit of room to move things around.
|
|
#
?
Feb 2, 2022 03:38
|
|
|
Maybe 2022 is they year I'll get a Mo-Ra 420 with 4 x 200mm noctua fans  Currently I have 2 x 360 rads in the case and one external 420 rad and water temp doesn't go past 37-38 when gaming so I don't need to upgrade.
|
|
#
?
Feb 2, 2022 12:01
|
|
|
Sorbus posted:Maybe 2022 is they year I'll get a Mo-Ra 420 with 4 x 200mm noctua fans I’m gonna do a 360 with 4 Silverstone 180 mm Air PENETRATORS. I’m gonna see if I can get an external power supply to power everything separate from the PC. Old Xbox 360 power supplies put out 12v at decent amps.
|
|
#
?
Feb 3, 2022 02:13
|
|
|
VelociBacon posted:e: I should also add that the case can take bigger rads, I'm using 420mm rads but it can actually take 480mm rads so there's a fair bit of room to move things around. If you try lighting an incense stick over the top of your case, you’ll see that spare 60mm at the top rear will function as a passive exhaust. I have tried and observed this on my system (2 bottom intake, 2 front intake pulling through GPU AIO rad, 2 top intake, Noctua DH-15 pulling all that air to the rear of my case). Because my top fans are 2*140 I have a gap at the rear so I cranked my rear 140 exhaust 10-20% and the top rear passively exhausts the rest. The difference with your setup is there is no extra airflow in the centre of your case as mine is with a CPU air cooler. You could slightly increase the front intake RPM to mitigate this, however.
|
|
#
?
Feb 4, 2022 00:17
|
|
|
adeadcrab posted:I would probably go intake on all rad fans and bottom two fans, exhaust on rear. Thanks. Is there no concern about the ambient temp in the case being higher? For my HDDs I guess.
|
|
#
?
Feb 4, 2022 19:38
|
|
|
VelociBacon posted:Thanks. Is there no concern about the ambient temp in the case being higher? For my HDDs I guess. Should be fine as long as HDD temps don't exceed 45c, and unless the rads are constantly cooling down CPU/GPU in the thermal throttling range, any airflow is welcome.
|
|
#
?
Feb 5, 2022 01:32
|
|
|
I think my pump might be air locked, in the test setup I have to check things out prior to install. Simple test setup, res -> flow meter -> inlet on Heatkiller D5 Top -> outlet -> rad -> return to reservoir. Shouldn’t cause too many weird things to happen, right? I could see a pump -> res only loop being kinda weird, but the radiator should provide some restriction, right?  I’m pretty sure what I’m hearing is air, or cavitation — plus, when disconnecting the return hose from the res, I get no water moving out of it unless I place it below the pump (i.e., doesn’t seem to be pumping… but is drawing current). Flipping the setup end over end to try and get any bubbles seemed to help a bit (in the sense fluid level on res would drop), but… shouldn’t a D5 be able to easily pump through any air bubbles? Especially if I have the outlet dumping into the sink? At that point its just a water pump flushing through the radiator and into the drain. I found a few threads that have me checking to see if I read the Heatkiller manual backwards on what IN/OUT are, but I’m pretty sure I got it right. Even with those 90s, the pump should always be wet / primed. movax fucked around with this message at 04:32 on Feb 6, 2022 |
|
#
?
Feb 6, 2022 04:28
|
|
|
I’d just run the pump and the res to make sure the pump is functioning properly. A full rpm d5 should have no problem pushing air out of that radiator. I see that it has the manual vario knob, is it turned up? Maybe post a better photo of the pump top to make sure you don’t have double outlets or something?
|
|
#
?
Feb 6, 2022 06:08
|
|
|
Is the pump actually running? And does your flow sensor pick anything up? Also, echoing Canna, could use some better pictures including the radiator hookups and the pump hookups. Looking at the two you posted, I can't see the second pump connection or any of the radiator hookups. Unless the bottom of picture two is it.
|
|
#
?
Feb 6, 2022 09:29
|
|
|
Sure thing! For reference it is the HEATKILLER D5-Top, which as 4 ports (IN1, IN2, OUT1, OUT2) — they recommend IN1/OUT1 for ‘best performance’, but any combination of in/out is valid.  This is what I came up with to get my pump attached to the rad in my case. That’s the outlet at the top of this picture (OUT1 from the D5) attached to the radiator (GTS 360), via a rotary M-M extender. The bottom of the pic is an offset rotary hitting a rotary 90. This then nominally in the loop would go to my CPU block; in this case its just looping back to the res. I hope I don’t have insane vibration issues doing this, but we’ll find out! I’m mounting the radiator to the case through some gaskets / rubber as well. ‘  This is the inlet side (IN1) of the pump. Rotary 90 on the input. All Tygon A-60-G tubing; I used the McMaster EPDM stuff at first / checking lengths, and then cut it to size.  Side view of the res (EK XRES 110) — nominally the port on the left is the inlet / return, the port on the right I am using as the outlet. It hits a rotary 90, the flow meter, and then the pump inlet — unless those adapters are really really hosed up, I believe this should still keep the pump nicely flooded in water. I didn’t have the flow meter / powered yesterday, but yeah, I’m thinking I might just get the OCTO down here, power it up and see if I have a spare USB harness I can use to convert it to a proper plug. In my head, it seems simple — it’s a not pressurized system / boiling point doesn’t matter (unlike a car), so if I run this with the reservoir lid off, the pump should suck in a ton of water, push air out of the rad, into the res where it escapes, while I keep topping off water. E: a very  question… I thought Imgur did sound? I have a few videos of it going / gurgling, but is there somewhere besides YouTube I should upload them…
movax fucked around with this message at 18:54 on Feb 6, 2022 |
|
#
?
Feb 6, 2022 18:35
|
|
|
The pump top is installed in the correct orientation? I've not worked with D5's, but it's a stab in the dark because yeah I can't think of anything goofy here. I don't see why there wouldn't be enough head pressure to do exactly what you're saying. Maybe tilt it around to encourage the air to move out of the radiator? Again, shouldn't be a problem at all--the pump should have more than enough power to force air bubbles out into the res over a reasonable timeline, and there should be no situation where the pump itself is dry. One of the big advantages of the reservoir on a custom loop over the AIO is that you're not trying to trap the air in the radiator. It all naturally moves to the res over time.
|
|
#
?
Feb 6, 2022 18:59
|
|
|
Warmachine posted:The pump top is installed in the correct orientation? I've not worked with D5's, but it's a stab in the dark because yeah I can't think of anything goofy here. I don't see why there wouldn't be enough head pressure to do exactly what you're saying. Maybe tilt it around to encourage the air to move out of the radiator? Again, shouldn't be a problem at all--the pump should have more than enough power to force air bubbles out into the res over a reasonable timeline, and there should be no situation where the pump itself is dry. One of the big advantages of the reservoir on a custom loop over the AIO is that you're not trying to trap the air in the radiator. It all naturally moves to the res over time. Hmmmm — I didn’t think that mattered, since the pump was spherical. (You’re referring to the clocking of the pump itself relative to the pump top, right?) I definitely did rotate it around to optimize where the speed control was, so it wasn’t buried behind that adapter. If you’re referring to the inlets/outlets though, I’ve been staring at this for literally months as I planned the loop slowly in my head… please sanity check and make sure I’m not a dumbass:  I’m like 99.9999999% sure I’ve done that right, just rotate that drawing 90 degrees to the right and that should be it. The IN1 even ends up above the OUT1, so I think that is sane. To my knowledge, I’ve never run this pump dry — the first time I turned it on was yesterday, and I had the res filled up feeding it. My expectation when I disconnected the return line from the res is that it should be shooting a jet of water out of that hose as it drew from the rad. Instead, I get bubbling / gurgling sounds from the pump that lead me to think it is airlocked, “cavitating” (I know its not technically cavitation / people mis-use that all the time). Not sure if audio made it into any of these: https://m.imgur.com/a/zn3wn9y E: Yeah, I don’t see anything specifically called out here, unless there’s an implication somewhere of the text being aligned, but again… spherical pump right? 
movax fucked around with this message at 19:12 on Feb 6, 2022 |
|
#
?
Feb 6, 2022 19:10
|
|
|
Quick noob question, do these 90 degree fittings need these fittings attached to connect to tubes?
|
|
#
?
Feb 6, 2022 19:31
|
|
|
I’d take it apart and run a simple run from pump to res and back just to see that it’s working. Is the pump speed turned up? How about a quick photo of the male male to the radiator? Is that a qdc on the backside?
|
|
#
?
Feb 6, 2022 19:45
|
|
|
The Atomic Man-Boy posted:Quick noob question, do these 90 degree fittings need these fittings attached to connect to tubes? Yes. Those fittings are way overpriced on amazon.
|
|
#
?
Feb 6, 2022 19:52
|
|
|
Have you tried blowing through your loop to make sure it isn't blocked?
|
|
#
?
Feb 6, 2022 20:43
|
|
|
Indiana_Krom posted:Have you tried blowing through your loop to make sure it isn't blocked? This is a good idea, and also if you're worried about air in the pump, you could theoretically just siphon water into the pump. I've not worked with D5's, so I'd assume you're right about orientation. My experience with pumping water out of the loop is that you should get a steady, low pressure flow. Canna Happy posted:I’d take it apart and run a simple run from pump to res and back just to see that it’s working. Is the pump speed turned up? How about a quick photo of the male male to the radiator? Is that a qdc on the backside? This as well. Kinda standard troubleshooting methodology to try and eliminate what it might be. Try it without the flow sensor too.
|
|
#
?
Feb 6, 2022 23:34
|
|
|
I’m refilling an older (non-asetek) AIO. Is there anything cheap that I can use as a biocide? It will all be in the black tubing so there’s really no point to fancy colors/etc, I just need it to not breed mold. Would dumping in a portion of isopropyl alcohol work, or something like that?
|
|
#
?
Feb 8, 2022 01:41
|
|
|
Paul MaudDib posted:I’m refilling an older (non-asetek) AIO. Is there anything cheap that I can use as a biocide? It will all be in the black tubing so there’s really no point to fancy colors/etc, I just need it to not breed mold. Alcohol will probably dissolve the plastic parts in the impeller, I would not recommend it. IIRC a lot of AIOs are filled with a mix of ethylene glycol (a common automotive antifreeze).
|
|
#
?
Feb 8, 2022 01:50
|
|
|
Indiana_Krom posted:Alcohol will probably dissolve the plastic parts in the impeller, I would not recommend it. IIRC a lot of AIOs are filled with a mix of ethylene glycol (a common automotive antifreeze). I ran a ~75/25 water antifreeze mix back in 2003 in my first loop. It also slows/stops mixed metals from corrosion and has additives to keep gaskets and seals in good shape. I’d almost say it’s a better option than most of the junk that they sell for loops these days.
|
|
#
?
Feb 8, 2022 06:34
|
|
|
The issue with modern (many decades, not a few years) antifreeze is that it is designed for an aluminum loop and not a copper one. The protection additives are designed with that in mind. For cost reasons, hardly any automotive applications use copper. That works out great for AIOs for the same reason. It's been awhile since I researched, but as I recall it isn't bad to use antifreeze per se, but it may have less longevity than would be expected in a car. Some of the protective qualities still apply in a copper loop. I would trust something like the prestone universal antifreeze, I think it's called Total Protection now, used to be something like Worldwide, to behave. I think all of the vehicle type specific stuff ("Asian", "European" etc) is not a good choice for water cooling, they are compromises that don't need to be made any longer, or fixes to things that don't apply such as iron in the loop, and have drawbacks like being more acidic. I'd agree that it is probably a better choice than a lot of the watercooling specific stuff. taqueso fucked around with this message at 07:23 on Feb 8, 2022 |
|
#
?
Feb 8, 2022 07:17
|
|
|
https://i.imgur.com/9gzLVNx.mp4 Hope the link works, I'm on my phone in bed exhausted after building the custom loop for the last 4-5 hours. (The PC isn't turned on during that video, just powering the fan hub with a spare PSU, I guess the fan hub feeds back into the mobo somewhere, it's a fractal thing) I took my time and had some dinner and cat petting in there but it definitely took longer than I expected, but super fun. Took at least an hour just removing the existing stuff and swapping stock fans back into the AIOs, putting the old GPU into a box (2080 to hybrid --> 3090), trying different rad/pump configurations, learning how to use the fittings, etc. Total loop volume is around 1.5L with the two extra thicc 420 rads and 250 res. Still have to wire the pump and fans properly now that I've more or less bled the system, and make the fan curves. And add a thermocouple to monitor ambient temps since those rads are both intakes. Oh yeah gotta add an exhaust fan too actually. Anyone have a decent fan curve based on coolant temperature that they want to share? I wonder how quiet I can run it with this much radiator.
|
|
#
?
Feb 9, 2022 09:44
|
|
|
Indiana_Krom posted:Have you tried blowing through your loop to make sure it isn't blocked? Warmachine posted:This is a good idea, and also if you're worried about air in the pump, you could theoretically just siphon water into the pump. So I'll be heading out for work soon, which will put this project even further behind, but I drained the loop and checked a few bits in isolation: * The pump in isolation, when just feeding it tap water (I'm going to do a DI rinse, don't worry...) and power up, shoots a healthy stream of water from its outlet, so it seems to be working! * A thought occurred to me where the flow meter might be too restrictive on the inlet of the pump... most builds I think go res ---> pump with the lowest impedance possible. But, looking at the inside of it, it doesn't seem very restrictive... I need to upload my videos with sound, because the gurgling / bubbling sounds will probably make the issue screamingly obvious.
|
|
#
?
Feb 10, 2022 19:28
|
|
|
I’d honestly look at the qdc before anything else if the pump is working.
|
|
#
?
Feb 10, 2022 19:55
|
|
|
Canna Happy posted:I’d honestly look at the qdc before anything else if the pump is working. Hmm -- I think I've mated/de-mated them all enough, but the loop doesn't hit a QDC until the output of the radiator, and I've tried removing that QDC to see how much water comes flowing out. Next step is testing the rad in isolation and seeing if there's some kinda blockage in there (and hooking up the flow meter).
|
|
#
?
Feb 10, 2022 20:38
|
|
|
Are rotatable fittings more prone to leaking? I read where they’re good for bottom 1/3 of case but I was thinking of having 45’s come off my cpu block and gpu and figured get them rotatable while I was at it.
|
|
#
?
Feb 11, 2022 01:57
|
|
|
Coredump posted:Are rotatable fittings more prone to leaking?
|
|
#
?
Feb 11, 2022 02:24
|
|
|
|
| # ? May 11, 2024 13:45 |
|
|
The 45 and 90 deg fittings I got from EK... it's really wild how much movement they have in terms of tilting side to side. No leaks though.
|
|
#
?
Feb 11, 2022 03:07
|
|