Thermochill = OOS

[Cathar]

Wouldn't running two of them at 26v negate any sort of gains with the flow rate?

I was under the impression from your tests that anything past 18V would be more harm then good.

I don't know what our good friend Turtle1 has up his sleeve. He just asked me "What could I use to do this?", and I answered.

What he seems to be attempting is well beyond the norm of regular PC water-cooling systems, so I'm not making any assumptions at all as to what's going to work out best for him.

Given enough radiator capacity, the heat dump penalty of 2 x RD30's @ 26v can be overcome, but we're likely talking about needing 8 x PA160.1 to do that, or maybe 4 x PA120.3's, to achieve a balance point. Maybe that's what he's got? Who knows?

All I know is that from what little he described of his system so far, that it's not your usual PC water-cooling setup by a long shot, and a lot of what hes delving into doesn't apply to the standard PC water-cooler. While it's not a different set of rules, the limits are different.

[Turtle 1]

First on the subject of the high HP racing engines. We don't run our pumps at the same ratio as a street engine . Mine is reduced to to 50% of what stock is so it runs at about 4,250 RPM . We also use speicial pumps that prevent cavataion both water and oil pumps. So thats really not part of the equation with the use of restrictors.

I have used for pumps so far the rd20 /rd30 and the alphacool ap1510.

your correct as is normally the case that the RD30 in paralell will get me to 10.6 gpm but 2 of these dump a lot of heat into the system and are noisy as hell. 2 rd20 are good for 8gpm. but still to loud for me than I have to convert 12v to 24v on the RD20/30 which is ok.
Right now the Alphcool AP1510 is what I am happiest with . because of the v.control unit . It pushes around 1500Lph which is right at 8gpm . which will in fact do the job . But I really would like about 10gpm and cut v back so as to have a little overhead. the 1510 has excellant head.
I did have to resize the 160 for 3/4 inlets and outlets to maintain good flow however. Which is OK . I am ripping off that design because we can make them cheaper than I can buy them . When I can find them . Than we also make them showy looking .
I won't be reselling these as I wouldn't feel right about riping off someone elses design. They will just be used in our gamers.

Besides with the added work of making them show piecies I couldn't compete price wise.

Now off topic . I laughed pretty hard on the story you told about the nurse ar [H] . I have a good one also.
When I got injured in 82. broken left shoulder completely dislocation of the left knee (only thing holding my leg on was some skin )ruptured L4 and L5 in lower back and one hell of a headache. I had this Candy stripper bring my food . Well she would come in lay down my tray and leave . Here I am with only one usable arm/hand . Well I didn't eat much first 3 days anyway.

The 4th day she brought it in and was heading out the door and my tray hit the door befor she got out. Dam the doctors were coming in to see me at that same time. He yells at what the hell is going on . I yell back in not so good wording . LOOK AT MY---- bread and My GOD --- milk . Bread isn't buttered and milk isn't opened. They went out in the hall there was some yelling . Doc came back in and said please except our apologies she no longer works here.

[Turtle 1]

On the 1510 . 1 pump pushes 1500lph= just under 4gpm. 2 in parallel is just under 8gpm.
Because of the very high flow Flow control unit were using . I have found that 4 160's modded to 3/4 inlet outlet. really doesn't restict the system much. Cathar knows better than anyone that the 160's have enormus flow capabilities use of 3/4 inlet outlet ID only expands that flow capability.

So 3 160's will cool the system and the 4 th one was added to take care of the pump heat dump with a little extra left for the 6(dual socket) or 5 cooling circuits.

[Cathar]

First on the subject of the high HP racing engines. We don't run our pumps at the same ratio as a street engine . Mine is reduced to to 50% of what stock is so it runs at about 4,250 RPM . We also use speicial pumps that prevent cavataion both water and oil pumps. So thats really not part of the equation with the use of restrictors.

Okay. If I were in charge of it myself, my brain would be working overtime to find the explanation, and I'd be sticking all sorts of diagnostic equipment on the cooling system to figure out what's going on. It's well established in physical theory that greater flow equals a higher reynolds number, equals a higher h (co-efficient of thermal convection). If that's not holding true in some specific scenario, then there's something else going on. I'd be curious to know what. Might even be some degenerate flow-regime scenario with too much flow, and when the flow transitions from laminar to turbulent in various cooling paths, things might start flowing differently. Where flow was being divided in the right proportion to various engine locations at lower flow, it might become imbalanced, increasing temperatures, and therefore metal expansion rates, which then causes increased frictional losses, which in turn generates dramatically more heat.

There'd be some reason for it, but it's certainly not going to be the "simple" answer of more flow is bad. It's not. There'd be something else going wrong, maybe because the system wasn't originally engineered to handle laminar->turbulent flow transitions at super-high flow rates? Maybe because of something else altogether?

I have used for pumps so far the rd20 /rd30 and the alphacool ap1510.

Ok, so you're also after quiet 8gpm operation?

The Alphacool AP1510 is decent as you say. What's your current problem with it? Is it not suiting your needs?

[Turtle 1]

Well the 1510's are ok but to achieve that flow rate I have to run them at 24v. I would like to have a little overhead. You know a small safty net.
I was just hoping you might know of a pump that I haven't heard of.

Its still a toss up between the rd20 and the 1510 . both compare about the same . Noise issue we pretty much took care of . With 6 circuits recieving 8gpm . is really cutting the flow per circuit.. Yes the nb/sb/ and regulators don't need that much flow I still would like 1gpm threw that circuit and the HD circuit. that leaves me with 1.5 gpm to the 2 gpu and 2 cpu circuits .

On the five circuit setup only 1 cpu circuit . were really happy with the results .
But our promised hardware from a certain company hasn't arrived yet. So I haven't really beenable to give the the 6 circuit system a valid test yet. I believe well be ok . But until I put a hard load on the 6 circuit system for 24 hours I just don't know yet. I am just trying to be prepared for the worse case .

[Cathar]

Well the 1510's are ok but to achieve that flow rate I have to run them at 24v. I would like to have a little overhead. You know a small safty net.
I was just hoping you might know of a pump that I haven't heard of.

Look, it's a long-shot, but it's the "big-brother" of the old Swiftech MCP600. Made by the same OEM manufacturer.

It's the EW-72025-95 (aka K-72025-95).

Variable speed, 5.3gpm, 24VDC, 36' (~11.5mH2O head).

Don't know how loud it will be. I haven't tried it myself, but it's about the only other pump that I can think of that would suit your needs. Would need 2 in parallel.

The other possible pump, that'd do it all in one, would be the Panworld NH-40PX-D. 12vdc. 8gpm. ~4mH2O head.

Alternately, grab four of the Laing D5's, and hook them up in 2x2 format. i.e. A parallel pair of 2-pumps in series. That'd net you some serious flow & pressure, AND you could dial them into whatever speed you wanted.

My only concern is this. You're trying to pump 8gpm through the radiators at the moment. What's going to happen when you add some CPU water-blocks in there? Or are you running a separate radiator loop to a shared reservoir to the component cooling loop?

Seriously. I don't know what you're really trying to achieve, and without further info, I'm just taking stabs in the dark. If you want to run it by me properly, then take it to PM, and I'll keep it private, and you can share with me the full details of what you're doing, and I'll propose something. I really don't want to continue this game of me guessing what you want. No fun for me.

[MotF Bane]

FrozenCPU has them!

I think Frozen usually does sales on holidays... but I'm not risking that two day wait when they only have six.

[kinghong1970]

i purchased the 160 and 120.2 from Thermochill directly...
not too bad... DHL shipping and all but took only a couple of days more...

need to check my credit card bill to see how much it actually cost me tho...

[Turtle 1]

Actual cathar you have really helped me alot already. Sometimes just setting a talking about something with someone you trust a light goes off. Your last post did that just for me . And I slapped self on head. I already linked you to the flowcontrol unit I am using. So when you said 8ghm might be a bit much for 4 rads which from what I already told you about 3.2gpm is what seems to be the best for 1 160 Since I am using 2 y blocks anyway.

There is no reason at all that I can't run 1 pump threw 2 160s . with the Y block after the rads. . Since I alread sold 3 gamers with 5 circuits. I can easily re plumb them . Running just 1 pump threw 2 160's cuts the restriction threw rads considerably and the heat dump should be smaller also. . I can see clearly now this will be much better than 4 160's in series . 2+2 rads inseries is much smarter and will prove to be more efficient .

Thanks so much for your imput and the links to those last 2 pumps. You have expanded my options and I believe solved my problem . AS always it turned out to be a simple solution.