Koolance 452X2 Res Questions

[GMdoubleG]

Hello all,

I just installed a new Koolance 452X2 res into my custom mountain mods case and it seems to be acting weird, at least to my knowledge. I have the D5 vario pumps in series with the "U" connector fitting that allows one pump to flow to the other. The pumps seem to get into sync and sound nearly silent (level 3 vario), then a few seconds later they seem to get out of sync and become very loud/whiny.

First question, should the pumps be on the same setting?

Secondly, would making one faster than the other pump cause problems? Maybe early wear that causes it to break down?

Third, if running one pump faster than the other is ok, should the "forward" pump in the loop be the faster one so it is pulling more from the slower pump?

Fourth, for those people with pumps in series right now, do you experience the same thing? I realize that the noise level will be louder with two pumps, but this is much louder then when I had them running in dual loops next to each other.

Thanks for the help.

[jkresh]

I have two pumps, both are at the same volume and I am not experiencing the issue that you mentioned (though at the beginning when the loop still had air in it that was not uncommon). It should not matter if the pumps are on the same setting (but I do not see why you would have them on different settings).

[Martinm210]

Try either matching the RPM or making them very different. Sound like you have a slight harmonic between the two noise sources. I have seen this both with fans and pumps. You get it when both soirces have very similar noise rythms but just slightly different.

[GMdoubleG]

So I realized that a major portion of the noise was air bubbles still going through the system, even after 30hrs of running the system. I opened the top screw cap and filled it with a little water, and then a little more. No improvement at any point. It appeared that the Res would never fully fill up and the pump was taking water from the top port in the Res which caused air to go through the system. After filling the top fill ports to the point that the water almost overflowed, I put the caps back on. After starting the system, water started coming out of the left pump and out of the inlet area to that pump.

Now I guess I need to tear down the Res and redo all pumps and fittings.

Any ideas what is going on? Could I flip the Res so the inlet ports to the pump going to the system will be on the bottom?

[Martinm210]

If you are not using the new pipe bleed kit you will need to fill that right reservoir completely full by tipping the case to the left while bleeding and topping. Check with Koolance, bit you might still get a bleed pipe kit for free. I have a sample I have yet to test, but it was suppose to improve the right res high port location.

I would suggest the tip left and top off method first. I found you need to tip quite a ways over and cap the res before shutting the pump off. The left side doesn' t have that because both ports are down low.

[GMdoubleG]

Ok, I'll try that after taking apart the Res and figuring out why I got a leak. Also, I changed the plug to the short stopper to make both Res' act as one. Think I should go back to the long stopper so the Res' stay separate?

[Martinm210]

Yeah, you need them separate if you want full series pump power. If you bridge the two reservoirs in series, the first pump's pressure differential gain is lost since the in and out are shared. Then you basically only have the second pump doing work. Shared is only good for one pump or two pumps where you have separate loops for each pump but want to share radiator capacity between the two.

[GMdoubleG]

Ok, glad I asked. Could that possibly be a cause to the unusually loud noise when running both pumps?

I just completed the RMA process with Koolance to get the bleed pipe kit. The bleed pipe kit is a permanent installation for the Res, how does this effect the performance of the Res? If it restricts the performance a lot, I might just try to stay with the tip and fill method.

[GMdoubleG]

Martin: I followed your video on Youtube and it helped wonders.

http://www.youtube.com/watch?v=j7g3R3rm9Jo

Switching the inlet/outlet plate to the bottom and the closed plate for series to the top was a night and day difference for bleeding out the system. Also, the noise problems was fixed completely when I put the long barb back in to keep the two Res' separate and allowed the pumps to work fully in series. I think the reason for so much noise in the other setup was because the pumps weren't working together and were "fighting" each other.

Edit: After 1hr of running the system to fully bleed it. The left pump started leaking. This is the 4th time so far. I've reseated the pump, reseated the O-ring, switched sides, etc. Anybody have any ideas? I sent in for an RMA request to see where Koolance wants to go with it.

[NaeKuh]

you didnt get one of these guys did you?

http://www.koolance.com/water-cooling/product_info.php?product_id=1181

^ is a fix to bay res's.

Getting one of those fixes the bleeding issues.

[GMdoubleG]

I purchased the Res from a secondary supplier so my Rev. 1.1 didn't come with the bleed pipe and I didn't realize that there was one before installing. I already filled out the RMA info and they are sending the bleed pipe for free. Does anyone know how it changes the performance of the Res system?

[Mobsters]

I tried the bleed pipe and while it helps in some degree, i still had to use T-line to get major air out and flow going on.
I think if they would have separated the in and out holes in upper reservoir and made that tiny hole that is suppose to push air out to reservoid much larger it would work better as air would be forced in chamber while water gets sucked to pump in lower portion in chamber. Now air just goes back to system. I am using the block that comes with u-pipe kit on lower reservoir.

117680

[Phatboy69]

I tried the bleed pipe and while it helps in some degree, i still had to use T-line to get major air out and flow going on.
I think if they would have separated the in and out holes in upper reservoir and made that tiny hole that is suppose to push air out to reservoid much larger it would work better as air would be forced in chamber while water gets sucked to pump in lower portion in chamber. Now air just goes back to system. I am using the block that comes with u-pipe kit on lower reservoir.

117680

Yeah I thought exactly the same thing when I saw it too.. I even made a comment to them saying exactly the same thing and they said they tried all types of setups including what we suggested but they insisted the one they put into production works the best. I still question that though because i did this mod below and got the thing bled in 10mins. My mod i originally did is exactly what your talking about!

http://w7sp5a.blu.livefilestore.com/y1pNApfNK3QxLyvwSTYXEzBvRfLqGON-UsoSIJ4gLK6EJws_oEfeMd0iI7YlgKXYGjQmaW01WfQThGGv30 wG84m3KpVHc3Imk4_/IMG_1814.JPG

[Koolance]

117680

We tried that during testing and it didn't work as well for us. It's not intuitive, but with the paths joined, air is strained against the small front bleed hole. More "bubble free" coolant is brought in from suction of the lower pump outlet, and these two aspects working together significantly reduces the bleed time for most users. If the acrylic had the two openings separated, I think the top inlet still tends to push too many bubbles down into the reservoir (unless the bleed pipe was extended further down, but in that case it would hit the metal reservoir separating piece unless you were doing what Phatboy did).

Tim

[GMdoubleG]

So I used Martin's trick of flipping the serial pipe to the top inlet/outlet setup and put the block provided in the right side res at the top. It appears to still be sucking in a small amount of air which causes to system to not fully bleed out.

Anyone know of any tricks to get this thing to actually fully bleed the air? I plan on letting it run for at least 12 hours before I start searching out things to do to fix this. Is anyone experiencing it take longer than 12 hours and it actually bleeds out after that?

[GMdoubleG]

Also, I have both pumps down to level setting 2 so the flow rate will be slow enough to aid in the bleeding.

[Jytra]

GM, I think you and I are running into the same issue. I made a post here (http://www.xtremesystems.org/forums/showthread.php?274112-Need-some-assistance-RP-452X2-and-a-lot-of-restriction), and I've been emailing back and forth with Koolance to see what's going on.

If you figure anything out, let me know!

[mlwood37]

Koolance An idea for you tubing in the new bay res you guys have , why don't you Rifle the tubing before bending this will help create larger bubble's which will then float to the top ;) instead of smaller ones that are pulled back though the system....

The idea above only creates smaller bubbles as it being smashed into the bar or base. Rifling could help this problem.

Have a try and if it works you can send me a pint of larger and a smile :).

[GMdoubleG]

Jytra: So I used Martin's trick with putting the block that came with the Serial pipe kit on the top (right res) and put the serial pipe in the bottom inlet/outlet series. Then, I turned the case on its back and let the system run for close to 24 hours. I shut the system off, turned the case back to its right side and then ran the system for another 12 hours before it completely bled. All the while, I had my adjustable pumps on setting number 2. After it was completely silent, I turned the pump to 3, and then 4. Now it runs great and my GT1850rpms are the loudest thing in my system.

It takes a very long time, but I think it is worth it.

[Jytra]

I'll give that a shot after I get a response back from Koolance.

Just one thing though: After putting the acrylic piece in the top and bleeding the system, wouldn't that affect the flow rate?


Sent from my DROID X2

[GMdoubleG]

Are you asking if switching the acrylic piece from the bottom to the top would affect the flow rate? I wouldn't see why. You are just switching where the inlet to the pumps are. The pumps will still be in series and the flow rate should be the same. Someone jump in here if I am wrong.

Also, I have been running my system all day long and it still seems to be very slowly clearing the system of air bubbles. Every day, I have had to add a little bit of water to the res because the level keeps going down just a bit. I also have all the fittings wrapped with tissue just in case I am missing a leak somewhere. But I have ran my PC for three days straight and I can't find a leak anywhere. I am very weary of this because several years back, I fried an audio card because I missed a slow leak in the fittings on my GPU block.

[spacedog]

Just one thing though: After putting the acrylic piece in the top and bleeding the system, wouldn't that affect the flow rate?

Well, if you take the res apart and examine it, you'll find that all 4 internal pathways are 3/8" ID. I can't imagine what noticable effect adding that extra inch of 3/8" tubing would have. Plus, if you're running serial D5s, and are having flow rate problems to the degree where that would matter, then you have bigger problems. ;)

For the record, I modified my 452X2 with the same idea in mind, although mine was a bit more complex:

119696

This took a bit of experimenting, and is kind of complex. On the plus side, it was extremely effective -- I fully bled through 2 GPU blocks, 1 CPU block, and 1 3x120 rad in about 5 minutes, with the reservoir at the lowest point in the loop, and almost no tipping of the case required. I just cranked both pumps to 5, which micro-areated all of the fluid (it looked like milk), and then turned the pumps back to 2. It was crystal clear with no bubbles anywhere in the system (that I could find) a few minutes later. (This is with a rev 1.3 reservoir.)

IMO this product line is a great idea, and has a lot of potential -- especially with regard to its ability to greatly simplify most loops. However, the design still has some kinks to work out. I'm still glad Koolance went to the effort to make this, though, even though I ended up feeling a little bit like a beta tester at the end.

[jayhall0315]

Yeah, I have had the 452X2 version 1.0 and I did get the bleed kit for it. Even so, I really feel like no one Koolance staff member ever took this reservoir and actually set it in a standard full tower case and tested it in several standard configs (dual res, serial, low and high position). It has all the hallmarks of not being designed by a professional engineer with a university education. Part of the problem for all of you is the cavitation caused by going from abruptly from say 1/2 inch inner diameter tubing to 3/8 inch ID pathways in the reservoir. right before the pumps. Generally I think Koolance makes decent stuff but this certainly burned many people out of $130 and days spent trying to bleed the silly thing.