Gezzer

Members
  • Content Count

    9
  • Joined

  • Last visited

Community Reputation

3 Neutral

About Gezzer

  • Rank
    LED Head

Profile Information

  • Gender
    Not Telling

Recent Profile Visitors

115 profile views
  1. Okay, so after tourofrooms responded I decided to explain everything, once I had it all worked out in my head. But once I did, I realized that it would be one :) of a wall of text. Now I'm a layman, so it's most likely not 100% accurate, but I think the reasoning is sound, and anecdotally it seems to fix the problem. It has to do with things like flow rate and different areas of differing restriction, and how that relates to pressure zones, back pressure, and laminar flow. So I can walk anyone through it if they want me to. But in a nutshell, the kit's loop is very poor at bleeding air naturally. But with a small pump at say 1000 RPM the back pressure problem would be moot. With a much more powerful pump like it now has (3600 RPM) it has all sorts of pressure issues and this causes the pump to cavitate, which is very bad and will cause it to eventually fail. With a custom loop after your first fill of cooling fluid, you momentarily start just the pump to pull the fluid into the loop and force the air out. Then you pour more in, and repeat till the loop is full. I even like to pick the computer up and move it through different planes of rotation to help any stubborn air pockets work their way out. But the thing is you'll never get all the air. There's always nooks and crannies that will trap small pockets of air, but as long as they aren't able to work themselves into the pump it's fine. As well some of this air will slowly be picked up by the flow of the fluid and eventually reach the highest point in the loop (should be the fill port). That's why you need to check the fluid level every now and then. But just a second you say, isn't the kit a "sealed" unit? Well yes it is, but it's still very hard to get all the air out before they seal it. As well there's something else. The tubing allows for a small amount of evaporation. It's vinyl tubing which is not totally impermeable. Now the rate this happens at will increase with temperature, but it can still happen, just more slowly, at room temp. So if a unit has been on the shelf a long time.... Have a look at this picture: http://best-b2b.com/userimg/927/951-1/computer-cooling-radiator-998.jpg It's actually originated perfectly for my explanation on what I did. You'll notice that where the threaded holes are (that's where the tubes hook up) there's two big chambers, with one behind each port. This might be hard to visualize because, well the case I'm using is a bit different : http://www.newegg.ca/Product/Product.aspx?Item=N82E16811119265 ( yeah I was going for a manufacture theme, lol) If you look at the pictures you'll see that the CPU/Pump sits just a bit lower than the 120 fan mount location. So I've mounted the 120 rad with the two ports facing towards the card slots and on it's side. If you were looking at the case from the front the 120 rad would be with the ports both facing the left case side. I then mounted the pump so it's ports were as well facing towards the left side of the case as viewed from the front. Then took the outlet from the pump and plumbed it to the bottom port on the 120 rad. This means I have maybe 2-3 inches from the pump outlet to the bottom 120 port. I used friction fittings and it was tight to work, with a bit of an angle to deal with, but I got it. I'd post a pic, but my camera's battery is recharging at the moment. So what happens with my solution is any air in the pump flows easily into the chamber behind the 120's bottom port and gets trapped on that little piece dividing the two chambers. This is because the chamber is less restrictive than the pump and the tubing, so has less back pressure. Unlike how with the original 240 rad the pipes enter a tiny chamber that's all most right up against the rad's channels. This makes the original rad more restrictive than the pump and tubing, which creates a lot of back pressure and makes the air less likely to naturally flow out of the pump (trust me, I'm pretty sure that's the problem). With a low RPM pump there just wasn't a lot of pressure period. In fact this current version of the kit would work great with that sort of pump. High RPM/pressure? Totally different story. So for how my loop runs? I mounted the 240 rad that comes with the kit on the front in a push/pull (intake) with a 4 channel fan controller, oriented so the ports were on the right side if viewed from front of the case. I ran a tube from the bottom 240 rad port to the inlet port on the pump. Another tube from the outlet port on the pump to the bottom port on the 120 rad (single fan exhausting from case). A tube from the top port on the 120 rad to the closest port on the GPU. (GPU uses low clearance fittings to reduce chances on air getting trapped) And a tube from the other GPU port to the top port on the 240 rad. Lastly I put a fill tube on the fluid fill port of the 240 rad with a clear tube section so I have a visual on fluid level and it's easy to top up if needed. Now keep in mind my case is not a standard tower, so with one the loop might need a bit of tweaking. But I'm pretty certain that with my case this is the optimal setup. Wow. What a wall of text, eh? Just think if I went more in depth. lol As well a photo would really help, so when my camera's ready I'll add that if I can. P.S. Since I've solved the issues I have Speedfan set to spool up the pump when the CPU temp is over 35C (25% to 100%). I also have the exhaust fan hooked to the motherboard and set to spool up when the CPU is over 35C and/or the MB temp goes over 30C. This is to allow for less airflow over the MB due to using water instead of air cooling. As well my GPU will spool both up if it tops 50C (which it never does). The 4 fans on the 240 rad are on a controller. The thing is if you haven't solved the air problem, I wouldn't let it spool up. At least not till you're pretty sure all the air has bleed from the pump, if it even can. P.P.S As for power issues having it draw power directly from the PSU instead of a MB CPU header will in fact help, because a DC motor has current issues with a pump that's cavitating or seized (that's why the motor will eventually go "poof"). But it also might hide the air bleeding issue because the pump has an unlimited (so to speak) power draw from the PSU compared to being limited on the MB CPU header. So it simply would cause the failure at a later date, but it'll still eventually fail IMHO. There. Done. I promise.
  2. I've used it in two different systems. It's actually pretty nice once you correct for it's defects. The pump is actually fine, it's just way too powerful. When they designed this one, it was an older design and they simply put a more powerful pump in it but didn't take into consideration what that did to the rest of the system. There's very little area devoted to a reservoir, which is where the problem is. A water cooling loop's reservoir isn't just for holding extra cooling fluid, it also acts kind of like a shock absorber. This is compounded by the fact the lines into the radiator enter on the other end and not the reservoir end. With a much weaker pump this wouldn't of been a problem, but once they changed that.... So what happens is the pump at full speed pushes the fluid really hard which hits the rad which resists the pressure. Then the fluid in the small res empties too quickly into the pump and the pressure on the pump's inlet side drops. This is why so many people talk about hearing the weird noise. The noise cycles between louder and softer and sounds almost like the pump is growling. What's happening is as the pressure drops the pump's impeller (the thing that moves the fluid) speeds up due to the pressure drop, and the pump comes close to cavitating (running with no or frothy liquid in the pump). This is a problem because the liquid both lubricates and cools the pump, and IMHO is why they eventually burn out. There's two ways to fix this. The best way is to do what I did and add some sort of res into the loop. I was putting a GPU in the loop and once I realized what was happening I added a 120 rad with a proper res on it and the problem was solved. Something like this. http://www.frozencpu.com/products/4061/ex-rad-80/Black_Ice_GT_Stealth_120_Radiator_-_Black.html You'll notice how the lines go into the the res side, unlike the way the Glacer setup is. You can also just add a reservoir. Something like this. http://www.frozencpu.com/products/13183/ex-res-282/Alphacool_Cape_Corp_Coolplex_Pro_10_LT_-_100mm.html Not sure if it would be better on the inlet or outlet side of the pump. I'd say inlet, but since it wasn't my solution I'm not too sure. Lastly if you want to use it as is you need to crank the pump speed way down so it doesn't overwhelm the system. I used a program called speedfan before adding the 120 rad and I found keeping the fan at 25% speed with the program stopped the surging that the pump was experiencing. In fact I use speedfan with the current setup (i7-875K @3.9Ghz and a EVGA GTX 680 FTW in loop) and at 25% with my fans also at a reduced speed the system is dead silent, with the CPU temp never over 34C. It's set to spool up both the pump and fans if the CPU goes over 35C. Here's a link to the program: http://www.almico.com/speedfan.php It's really kind of sad how the product worked out. The idea was sound, but it wasn't really tested very well. With a different rad that actually has a larger res integrated into it, and the lines going into the proper side it would be great. An upgradeable AiO, who wouldn't like that?
  3. No PSU included. To be honest the majority of PSUs that come with cases are pretty much junk. As well because there can be so many different systems in this case, it'd be real hard for CM to add a PSU without the chance it'll be too low in wattage. Plus if they did supply a PSU it'd most likely double the price, maybe even more. Try this link to figure out the wattage you'll need. http://www.extreme.outervision.com/psucalculatorlite.jsp This site is a great resource for everything PSU related. http://www.jonnyguru.com/modules.php?name=NDReviews If he says it's a good PSU, it'll be a really good PSU. But you are pretty safe with either a Corsair or EVGA. Just don't go too low on the price a good 750 watt should set you back around 150 minimum, any cheaper and I'd be leary of it. Just remember your PSU is the foundation that your entire system is built on, always better to pick quality over price.
  4. @sg8berg1720 You know if your willing to pay shipping I could send you the two I'm not going to use. Just send me a PM with info if your interested.
  5. Well my idea is pretty much for easier cleaning but I also think it'll just look better that way. I'm going to have the 360 in the 915 with the fan controller and have quick disconnects on the hoses between the 915 and the 935, plus just one semi easy to reach molex connecting the controller to the PSU. So with three quick disconnects the 915 separates from the 935 so I can give it a real through cleaning. I'll even be able to still boot the computer while I'm cleaning if I want. l'll post a couple of pics if it turns out as good as I hope.
  6. Actually I do have to cut it because the fan controller needs to mount in the front screw holes of the 5 1/4 cage. I guess I could try jury rigging something with wire or strapping to hold it in place, but it would look like :). I've got a dremel for modding and know a place with a metal cutting band saw, so it's all good.
  7. Yeah but as I said in the post previous to that last one I'm putting a 360 rad in the 915R, so that's out too. it's all good though. It would of been nice if they had given a second mount option for the lower cage in the 935 maybe a couple of inches closer to the PSU, but it is what it is. Kind of like how there's no hard reset switch, no case I've used has ever been perfect. For example I'm going to have to cut about 60-70% off the back end of the 5 1/4 cage in the 915R so that I can mount the 360 rad and also use my fan controller in the 915R. No biggy was kind of hoping Coolermaster would save me some work, but if it's not meant to be....
  8. Well no answer to my previous question, but no problem. I've just ordered this: http://www.newegg.ca/Product/Product.aspx?Item=N82E16817997056 And I won't be using any of the drive cages. Kind of a waste of both the cages and one 5 1/4 slot, but it is what it is. BTW, who in the design team decided that there was no need for a hard reset switch on the case?
  9. Hi. I'm in the process of building my ultimate water-cooled computer and have chosen the stacker for my case. I'm going to be cooling a lot of over clocked gear and want to do a multi rad build. A 360 push pull in the 915 and maybe a 120 or 240 in the 935 on the front. My problem is any pictures I've seen are of a push pull 240 on the front of the 935 with no drive cages except for this picture. I gather this was done using the optional adapter that is mentioned on the product page right? If it was great, how do I get my hands on one? Here's a picture showing the complete setup.