Project Metaversa: LCD goodness (small update 17/07/05)

I had the same problem with Vortex's forming on block. I thought it was due to the Spiral design of the block. Someone sugested the hose round to the other way on the inlet and outlet. Hay presto it work

Just a thought

 

No can do... The block inlet and outlet actually have to be that particular way around for the thing to work. Upon entering the block, flow is split into two S-shaped channels which merge at the outlet. The channels are cut in such a way that the two streams there merge in a whirl, rather than colliding, to facilitate a smooth exit. As such water exits in a spin, hence the vortices. Since I got all the air bubbles out and topped up the res to three millimeters below the rim, these are less visible, btw. Flow is now smooth and without bubbles.

Leak testing is complete now. Max pump temperature is 33C. Interesting since other people tend to report 53C on this type of pump. Perhaps it just doesn't like being constricted at the inlet, something which I have been at pains to avoid. I may still look for something with a longer lifespan though.

EDIT: Of course I also tested the flow switch. This closes about one second after the pump activates, and opens one second after it turns off. Nice.

 

Ummm MTBF = Mean Time Between Failures

Wonderful job on the project Nexxo! Nice attention to detail.

 

I was thinking of getting a resevoir like that one, but I have some interesting res mounting issues... I may need to mount mine sideways, is this at all possible? And did the tubing come with the pump or is it custom ordered? I searched "tubing CM30P7-1 Metaversa" and the forum searched said no hits. I then searched for "Metaversa" and it returned no findings. So searching was out, asking as a last resort

 

The Aquatube can be mounted in any position, including sideways. You can even buy a nice stainless teel mounting plate for that, that fits in two 5.25" bays! Check with Pug; his project Hexdragon has that setup.

This being a built-from-scratch circuit, the tubing was ordered seperately: Tygon 3606, 3mm thick walls, 3/8" ID tubing throughout and 3/4" ID tubing for the inlet of the pump.

 

However, having a MTBF of 4,000 hours doesn't mean that it will die at 4,000 hours.

If they test two pumps, one might die right away and the other may run 8,000 hours. This yeilds a MTBF of 4,000 hours but that doesn't mean at 4,000 hours your pump will die.

 

True... true... Good point. The official MTBF is closer to 2500 hrs. but in reality (and with some TLC) they can last longer... I'll see how long it lasts. I reckon replacement is due in 2 years and by that time, Johnson may have a brushless version of the pump (they're planning one now but don't hold your breath). Meanwile I just chill on!

 

Sorta OT I guess, but why does every manufacturer make their awesome looking parts out of aluminum? Is it a cost issue?

If Cu blocks are better at heat dissipation, why not use those?

Does it even make a difference in terms of the Al vs Cu waterblocks/radiator/parts? I mean it should, but is it noticeable?

 

*puts money back in pocket, continues using old components*

I think I'll wait around for that brushless Johnson pump. I like to spend more on those things that are gonna last you, like speakers, monitors, keyboards, mice.

Mod on! Can't wait to see the benchmarks of this beast.

I say you break 20k in 3dmark01, and you might break 7.5k in 3dmark03
And you better be overclocking

 

Your pump temps are pretty good right now. How much do you think it will increase once the system is running on a complete computer? Just a little curious about that. Looks great BTW...

 

Cant stop thinking about that brushless Johnson pump

Too bad the only one out now is a brushed version

Looking good as always Nexxo. You've got one of the most high-end watercooling systems I've seen, and It's so damn clean and professional looking.
Great job!

 

@ Spotpuff: Weight is probably a factor. You can't hang a brick onto a mobo... Also, copper is harder to mill. Still, most blocks nowadays have a copper base, and most rads are copper. There are some thermal considerations regarding why copper works better on blocks, but alu may be good for fins, but frankly I can't follow that whole discussion. It gets very technical...

@ the Sub: I suspect in a closed case pump temps will rise a bit, but as you can see it is purposely mounted behind the 120mm fan, so that it can receive some cooling from its air stream.

 

IIRC aluminum is better for dissapating heat but copper holds it better. A copper base with aluminum fins seems to be the best solution.

I am, however, NOT a physicist, so I will be quiet.

 

Jeez, 12 days of no updates and I find my log on the third page! Things sure keep moving in the modding community... OK, tiny update: given the fact that the Johnson CM30P7-1 makes a bit too much noise for my liking (OK, it's not loud, but louder than my PII 266. Kinda defeats the purpose of watercooling), I've ordered a Davies-Craig EBP pump:



15000+ hrs. MTBF, and practically identical in shape to the Johnson so I don't have to change internals or fittings around at the last minute. Lets see how this sucker performs...

Starbuck3733T has come up with a way of accessing my SMbus without depending on a SMbus connector being present on the mobo (which it isn't on mine ), so I may splice in some in-line temp sensors to monitor coolant temps in and out the blocks, via MBM. A small surgical procedure.

 

Nexxo, where did you pick that pump up? is it a stand alone pump too? cos the name is slighlt ambiguous 'Electric Booster Pump'...
Oh, and how much was it?
I was gonna have one of those MAX1668 chips off Starbuck3733T too, but my current G'byte m/b doesnt have the header, which would mean going the route you are.

This is looking good - following over @ WizD aswell so i don't miss nething!!

b4 i forget too (i know this is going back a while, but..) where did you get your Aquatube bored out for larger connecters? Was it local? only reason i ask is cos b'ham is my home out of term time and i'm getting an AT for my project in the summer.
Thanks

.: Fillip :.

 

The pump was ordered from thinkauto.com and costs £90,-- (yeah, yeah... ).

I had the aquatube bored out by an acquaintance at a Birmingham University metal workshop. A bench press drill does the trick. More critical was cutting the thread with a tap and die set. This is harder than it looks and needed experienced hands in order not to mess up the thread. I got access by virtue of the fact that I supervise trainee psychologists and do some lecturing for the Uni, but I'm not sure anyone can just access that shop. However any metal workshop should be able to do this for you, for little charge.

 

Looking good Nexxo That looks like one nifty pump.

I look forward to more updates


P.S. -- What you are doing for Brian over at ProCooling forums is really great. It's amazingly nice of you to go out of your way to type up those long responses and try to put Brian at ease. Best of luck to both him and you.

-OMP

 

oh right - well, i'm not paying £90 for a pump, can't believe you did!! lol.

I've got a couple of mates at b'ham uni but doubt i'd be able to get access to their workshop, my old school may allow me though they have bench drills and milling equipment aswell as tap + die tools (dunno about sizes) which i used when was student there.

I'll see if i feel brave when i get my AT in summer, otherwise i'll ship it off somewhere for the work, thanks for ya help.

.: Fillip :.

 

What is the fitting of tubing that you're using? Is it 5/8" and 3/8"

The pump that you got is damn power !! I can't wait to see how it rocks in your system !!

 

Tubing is 3/4" ID from res to pump inlet, and 3/8" ID throughout the rest (given that 1/2" barbs are generally really 3/8" ID anyway).

Pump arrived today; looks freakily like the Johnson only shorter and with plastic housing, not metal (bummer. But the MCP600 would have been too major surgery. And it's still a bit noisy). Let's see if it's any quieter this weekend!