Chuck, it's a good suggestion, but I don't want to add a third pump and loop to the system. I certainly could for the cost of the couplings, but the extra tubing, pump and paraphernalia will obstruct airflow. One reason I'm replacing the 1KW with an 860W PSU is to allow for more airflow on the two triple rad's.

I realize it's an unconventional solution, but if the couplings aren't too bulky and prove as effective as their stats would indicate, I think this is a solution I may use in my other wcooled boxes to ease servicing loops -- draining, swapping out blocks, etc. Sure would make life simpler. This is one way to -*test*-('") out a new approach to servicing wloops. If it works, great. If not, I'll have the 3/8" baby barbs on hand to muck up my nice EK blocks.

My second loop is ridiculously underutilized. I want to get a 790i just so I can watercool another 8800GTX XXX and make the bloody thing work for a living.

We simu-posted (new Merc word) Get rid of the pump on the second loop and replace it with a native 3/8" pump. Sell the pump, help a poor geek out and more than cover the costs. Kwikky may be in the market soon.

Yes, but then I lose the pleasing aesthetic symmetry of dueling MCP655's. And we couldn't have that, now. I could have sold my second 8800GTX XXX to you, had I known you were in the market -- complete with a zippy Stealth waterblock. It sold Friday to a fellow in Paris, France for enough Euros to cover the new 3870 x2.

Just so no one thinks I've entirely lost it, I didn't buy the $300 couplings -- I bought $75.00 worth of couplings (1/4 the cost; should edit that). Of course, overnight FedEx added $40 to that . . . it's been a rough week. What can I say. I had to drain the second loop to swap out video cards, so guess that was one way to go about it; the hard way!

After some canon shots back and forth -- mainly "forth" -- Super Talent is replacing my RAM with the second gen of STT Project X D3-1800 2GB (2x1GB). The 1866 RAM was, well, not cooperative, shall we say; spewing out a Technicolor morass of MemTest86 error messages at a lowly 1800MHz. No surprise Super Talent dropped that dog from their product line.

New Egg wouldn't accept a return for exchange or refund. MA Labs (owner of ST) initially refused to accept an RMA on their own product. This, after sending me two RMA forms to complete -- not bothering to check New Egg's policy that the end user proceed through mfr for warranty coverage. So far, I'm seriously underwhelmed with both Super Talent and MA Labs.

I hear ya and understand.

Anyone that has had to drain, swap waterblocks, replace Tygon or otherwise service a water loop knows it's a messy and sometimes risky proposition. I've yet to see any system use non-spill couplings. In exploring the option, the two disincentives are possible impact on PSI and flow and cost. The more perfect the seal, the more precision involved in making the couplers, the greater their ultimate cost. I chose to start with the best quality, medical grade coupler I could purchase in order to -*test*-('") efficacy: if the best doesn't work, then inferior grade solutions most likely will fail, as well.

The other issue is PSI drop. I've included a link to coupler specs with a graph showing PSI change given a starting PSI and GPM rating. The DD5 or MCP655 pump is rated at a GPM of 5.30 at 50 PSI. Using the linked graph, I concluded the coupler would contribute a maximum 10 PSI drop. My 'consulting chemist' and significant other disagrees, claiming the drop is significantly less than 10 PSI. Here is Part I of her explanation:

For the -*test*-('")ing experiment over the ranges -*test*-('")ed (p: 0-100 psi, T: 32-160 F). The chart shown is for a set series of Äp and flow rates. This DOES NOT mean the flow rate uniformly drops, regardless of which end of the coupler is in and out.

Fluid dynamics, the discipline of fluid flow, is the discipline with which one must draw to evaluate the scenario under consideration. Fluid dynamics can be applied to both air and liquid flow: air is a compressible fluid, some fluids are semi-compressible, others incompressible. Many liquids vary in compressibility states as a function of temperature and pressure: think of the tea kettle as it comes to a boil, or liquids change in thickness as they cool. The rate of flow of the fluid is termed fluid velocity or fluid flow, measured in volume of fluid per unit of time (e.g., gallons per minute, liters, per hour). Fluid flow depends on energy applied to the fluid (e.g., driving force from the pump), pressure, temperature, and friction drag from the environment (e.g., pipe, tubing, conduit) through which it is flowing. In many cases, the fluid flow rate will vary over time, as these variables alter flow to individual degrees. The (rate of) change of volume in time at a particular location is termed flux. When a set amount of volume past a particular point encounters a change in the environment (e.g., change in pipe or tubing diameter, change in pressure and/or temperature, change in friction), one or more of the other unconstrained variables will change.
Here’s some things to consider:
• The fluid is flowing in a “closed loop system”: that is, the loop begins at the pump and returns to the pump, with component tubing and process units in between.
• The fluid is intended to cool and heat. The fluid itself has specific properties: density (“how much does a given volume weigh?”); viscosity (“thickness of solution”); heat capacity (“how much heat can it pull out and hold”), compressibility (“can it be squished?”). These physical properties vary over pressure and temperature.
• The tubing has properties, too, material properties that define how much, how fast, and at what degree of efficiency the fluid flows through it: internal diameter (ID), friction drag (“does the interaction between the tubing and the fluid slow down the passing fluid?); resilience (does the fluid degrade the tubing material over time, thus changing other physical properties over time?”)
• The couplers have properties analogous to tubing: What fluids can flow through the coupler without degrading it? Over what p, T ranges does it optimally operate? How does the pressure change across the coupler in a given direction, at a particular flow rate?

This is the scenario under evaluation:
• Fluid flowing in a closed system ejected from a mechanical pump (linked to a reservoir) (1) at a particular (i) delivered pressure and (ii) volume; (2) through tubing of different sizes (3) linked by step couplers and (4) across heat sources (block) and exchangers (radiator).

Stay tuned for an analysis, in fluid dynamics terms, of Jason’s la-*test*-('") configuration.

Marion, thanks for understanding and indulging, our particular (peculiar) hobby. It is greatly appreciated as most of us are just guessing. LOL.

I don't know how you could -*test*-('") this without some kind of meter in the fluid. I guess all you can do is get a good set of temp readings, including ambient air before the inserts and then another set afterwards. I think it is worth it regardless as it will make system maintenance a lot easier. Also, if you buy high quality stuff it will last for years. Like any watercooling setup, the initial investment is high but the stuff lasts forever. My first watercooling kit is plugging away happily in the dual loop rig. If I end up going to a single loop in the future then it will find a home in the wife's rig.

Koolance has flow rate meters, but you'd have to put one in the 1/2" part of the loop and one in the 3/8" part of loop. It would run you about $55 for two meters and two sets of compression barbs - the best.

Koolance flow meter

More trouble for you, but fun for us if you did it.

Merc and folks, it is my pleasure to help. Jason lets me ** touch ** the vaulted water cooled behemoths when doing maintenance work that requires my second set of smaller hands or involves chemicals. My motivations: (1) One day, we will again have a dining room table at which we both can sit, at the same time, and (2) maintaining mental health in the household.

I heard Jason screaming "oh shi**, the tubing barbs leaked AGAIN!!!" via long distance (OK, so it was via email). As these SOB tubing barbs seem to be a failure point, switching tacks sure seemed the way to go. Actually, these connectors are so cool . . . and MUCH more reliable than damned worm clamps around tubing barbs. Designed specifically for low pressure (e.g., 250 psi or so), these connectors are perfect for a myriad of specialty applications. There are even coupling units that one can apply to a nozzle, and "plug and play" a variety of other tubing to it (think of garden hose at the spigot connectors, on steroids). God knows I wish these connectors existed when I was plumbing analytical instrumentation some 25+ years ago .. would have saved much skin on my hands from leaking chemicals during reservoir, fluid and tubing changes.

As for this being "DoD level pricing": well, at first blush, the approach certainly seems like it is. However, the connectors come in all sorts of materials which vary greatly in price. I've read the posts regarding switching pumps and going solely with 3/8" tubing throughout the loops. The radiator units and reservoir have 1/2" connection barbs, so there you go. Also, spending money for more pumps would not seem necessarily prudent at this time, maybe once this sucker is stabilized and a second rig is undergoing an overhaul.

Also, I'm up for getting a fluid flow meter and in-line sensors at some point, but not now (kind of expensive to put sensors in all the critical connection points.) I'll look into the Koolance suggestion. If anyone sees reference to a hand-held, affordable meter that measures flow outside the tubing (e.g., Doppler type meter, which we use for manufacturing plant piping flows), then that would be useful and surely informative. I think. Actually, I have no idea for what else we'd use such a meter, maybe keeping tabs on the garden hoses during watering ...

More later, once I slog through calculations for Jason's rig.

Marion

In other words: You dont want the hose to pop off on the 1/2-3/8 on the NB? LOL

I've actually got a Maximus Extreme sitting here next to me, and first thing I looked at were the fittings on the NB and the Asus converter kit, and I LOLed at how big of a POS they were...The barbs on the NB are non-existent, and the converter kit has just slots machined in the plastic to act as a barb... Crap..

I dont think you have to worry too much about the barbs on the NB cooler tho..Its the 3/8-1/2 thats the problem... Have you tried putting some silicone or something on there, and let it set overnight?

Have you seen the snazzy quick-release no drip 1/2" to 3/8" fittings Jason is getting? Wow - nice. I think Marion helped him locate them.

Oh yeah the garden hose coupling

Yeah it does look VERY cool... I might try to have a look around here in Aus for stuff like that, might come in handy one day... That being said, the hose might still pop of the NB barbs no matter how flash the coupling...

I think I'll install one of those flow meters the next time i do maintenance on my loop. Maybe a year from now? The quick disconnects are definitely a nice thing to have. I saw one coupling that was comprised of a barb end and a threaded end. Here are some ideas I have been thinking about when I saw that:

1. If those could be screwed into the various blocks I think you'd have a really convenient setup for future upgrades and cleaning. The blocks are the items most likely to get clogged and they could be easily cleaned out whenever you want.
2. You could hook the loop up to some kind of portable pump and flush it out with a few gallons distilled water then power fill it, plug the blocks in and top off.
3. Swap out blocks easily for review and -*test*-('")ing.

They would change the whole way you deal with your loop. It would be easierthan air cooling.

Paul, I realize it's the change in PSI created by stepping down ID that's the problem. The converter couplings are engineered for pressures much higher than that created by the DD5 pump -- up to 120 PSI. I don't think the pressure spike will be an issue.

I actually located them all by my lonesome. This is definitely Principal Chemist Marion's area of expertise and not mine, but I wanted the experience and chance to develop contacts with the various manufacturers and supply houses. Mission accomplished.

I called all over god's green earth to find a supplier than would sell to me direct in small quantities: McMaster Carr, VWR, MSC and Cole-Palmer were those distributor/resellers on my short list. Cole-Palmer was the only supply house with these particular connectors in stock.

Marion helped me after the fact to ensure the coupling would be suitable for the application.