Gratuitous Bling?
 

Now I do hate the term in general but what I'm about to post probabably counts as 'Bling' in Procooling usage. Either way I thought it might amuse.

What we have here is the Alphacool MCX minicoolers
http://www.veryrisky.co.uk/webimages/miniblocks04.jpg
http://www.veryrisky.co.uk/webimages/miniblocks07.jpg
http://www.veryrisky.co.uk/webimages/miniblocks12.jpg
http://www.veryrisky.co.uk/webimages/miniblocks24.jpg
http://www.veryrisky.co.uk/webimages/miniblocks28.jpg

Ok now the disclaimers:

• I know this isn't necessary and a set of ramsinks are quite sufficient
• I know it is proabbly a bit restrictive, but not too much as it is split into 4x3mmID from each 7mm ID outlet of the Antartica block.
• This stuff is at the end of the loop with the GPU before the CPU. There are dual lines back to the res.
• There's a half-decent pump running this, the AP1500 on 12-24V variable which is supposed to have maybe 3m head.

I think in the medium term I may take it off the CPU-GPU loop and have an Aux loop with just this, the NB and an 80mm rad.

You should do all the mosfets and other VRMs on the motherboard too and nb and southbridge and NIC controller etc etc until the entire case is filled with hose. Maybe cool each individual memory chip on your DDR too; that should fill up some space :)

Looks like a job for Flourinert (sp?).

So did it improve performance?

Good question !?!

I wouldn't imagine so. This is just an amusing diversion, I'm afraid. However I do plan to pull if off the main loop in the future to mitigate any impact on the primary cooling effort. Alas the build wasn't complete beforehand (or in it's current state) so I can't give temps etc.

Like the idea but not the implementation.... this isn't aimed at how you've fitted it btw... I jsut think they'd have done better to make the manifold blocks into a single larger unit with a single inlet, splits to all the small blocks, return to single large manifold, outlet back into rest of rig... just don't like the idea of having those 4 manifolds hanging in mid air suspended only by tubing.

As/if tubing/coolant heats up, would the small bore hose not be likely to collapse on themselves under the weight of the manifolds pushing down on them?

Should the smallbore soften and begin to sag, likelihood of manifolds in current position shorting out components?

How tight is the 3mm tube over the fittings - ie: would an excess of pressure cause any of the hoses to pop off?

Good points but it's not as scary as it looks. The splitters are really hanging off the bigger tubes rather than resting on the small bore. The hose I use is Legris 10/7 Flexible Polyurethane Polyether Tubing which is stiff enough to take some weight on the end so hopefully it isn't resting much weight on the small stuff which is Tygon R3603 5/3mm. That said I probably should replumb so every thing is stilling in the right spots.

There are fixing holes on those manifolds, though I didn't use them. The fact that I was using them in the parallel section makes it all look a lot more complicated. As the splitters have a second, blanked G1/4 at the end, you could chain them together for an 8-way with a single pipefitting.

As for the hose being blasted off, I guess its an issue which I'll look at if I ever get a DDC into the rig. However if takes a fair pull to get the stuff off as it stands so hopefully it will behave. You guys can probably tell me a lot more about Tygon than I know, the concern would be if it lost its elasticity.

Of course they could have used legris 5mmOD push-fittings and PUPE tube but that would have been harder to work with

haven't laughed so hard in months, had all the employees to my computer to take a look
Gabe will be heartbroken, it will be impossible to make something with more tubing connections than this;
amazing this WCing business

Oh chin up
At least get him to TRY

maybe take it one step farther and put little peltiers on each of the blocks..
super cool the ram :evilaugh:
I don't know how hot the ram get's but i will soon...
i ordered a Gainward 6600GLH video card...
I'll check teh temperature on the ram when i get it and make considerations from there...

Those are nickle plated what?

I ask this question every 4mos or so:
has anyone documented a higher overclock by WCing the memory chips ?
(i.e. demonstrated an improvement over air cooling the same chips)

Never tried water cooling the memory, but I got a ~12Mhz improvement in memory clock rate when I put an air duct in to cool the RAM. Blocking the duct off put it back at the original clock limits. Since the ram is not producing a lot of heat it seems a little judicious air flow will get the ram temps close enough to ambient that you'll get little noticeable advantage with water cooling.

I think the reason you don't get an answer is because not to many people watercool their memory on their video card.
also that would need a before and after overclock as well once somwone does that.
If the video card i get has the ram running very warm to hot... i will try it and let you know if i get a better overclock.
I still think tec based ram cooler for video cards would be cool :D

oh I get answers, always the same - no benefit
if water cannot be justified, how so a TEC ?

from what i have read... ram doens't run all that hot...
so if watrer cooling ram only brings the temperature down by 5 deg then i can undertand why there would be virtualy no improvment in overclock.
But if we have a 20-30 deg drop in temp for the ram... I think that should have some impact on the overclock.. not?

only by virtue of absorbing some of the heat in the secondary paths from the GPU

Now i didn't know that...
I'll have to do some more reading on that...

LOL, now that's exactly the response I was hoping for. I completely realise this is by no means a rational piece of cooling design, but nothing I've bought for the box in months has given me quite so much pleasure!

I'm sure its not going to be too hard to increase the, um, convolution(?) level a bit. I originally considered running all eight lines back to the res seperately but couldn't face the resulting difficulty in taking it apart in the future. However I'm sure someone will come up with dafter setup before I have the enery to take the next step. Obviously TEC cooling would be a, er, equally logical extension, if only in order to add all that extra cabling :drool: I'll leave it so someone else however.....

I suppose I should add that I will continue to recommend GPU-only cooling blocks and ramsinks if anyone asks on the forums. However at the moment there's a fad for these huge GPU and ram blocks which seem a bit silly, if not as silly as this effort!

We've tested PC-4000 memory in a case that has 0 airflow with all major components water-cooled (so there were no fans in the case) and running MemTest resulted in the memory getting hot enough to start producing errors. What happened is the memory would pass ~3-15 times successfully than start failing slowly with an error or two every odd pass after that till it gets progressively worse and started throwing errors up each pass. Putting a fan in front of the RAM resolved the errors progressively (but very quickly, within a pass or two) till no more errors were generated. This was reproduceable on 3 systems.

The systems were running the PC-4000 (DDR-500) memory at 3-3-3-8 200MHz, so it was underclocked to a certain degree.

Anecdoctal, but with that information in hand it appears cooling the memory more efficiently can result in better overclockability or just less opportunity for errors...?

Naturally, this isn't the same experience we've had with GPU memory which appears to run fine without any airflow.

ROFL!!!

Its 6 minutes later and im still laughing... im very happy i clicked on this thread... I think i have to make this my background.

Call Binford
Call Tim Taylor

Its the brand new Flowkilller 5000000.



EDIT: If you notice bill said that there is no improvement from air cooling. not no cooling at all..

Im sure you will see some gains from having the chips cooled compared to blazing hot. I have noticed it on myx800xt. but the amount of cooling it takes to stop the erros is a minor amount of airflow from a 120mm fan in the genera area of the card.

revive, found this:
http://watercoolplanet.de/index.php?open=2&show=101
"review" ?
Point to "mouting" and "pressure" testing.
Still interesting.

did not translate the article
BUT
I know a bit about thermal testing and the lack of insulation precludes anything meaningful from this test
(it WILL demonstrate cooling, but I don't think they have a calibration setup that would enable quantitative conclusions - yea, the H2O did better than that too dense hfs; BFD)

why is the test setup in a briefcase ?
saw no instrumentation either, perhaps in the text

Let me just clarify just for anyone that speed-read the original post.

1. I don't think this is in any way likely to improve the performace of the system,
2. I'll hazard a guess it wouldn't have added too much restriction given the configuration of my system.
3. I still think its pretty 'fun' to look at.

Burp (bump with attitude).
http://www.ocmodshop.com/default.aspx?a=249

difference. 3/8" or 1/2" .
same effect.

From my experience with gfx overclocking and cooling:
1. Running stock - no point in doing anything
2. Ocing without volt mods - improved case airflow entirely sufficient, mem. cooling pointless.
Even TEC setup without volt mods was fruitless in case of GPU (only single digits increase on soft modded R9500->R9700, all pipes functional) not to mention no increase in mem. Oc
3. Ocing fully volt modded card – GPU overclock went up in leaps and bounds, memory oc increased 50% after adding ram sinks (50% more than without).

Suggestion? Pump some good juice through that system and it should pay you back for all your excellent efforts with some nice clocks :)
Kudos for having balls to try it and actually making it work
P.S. Two separate loops sounds like a splendid idea!

Jabo

do you thing you'd have picked up any more by WCing the ram ?

on my 9800pro, I would have said no way. The ram was warm to very warm max with no heatsinks.

On my 6800nu, I'm not so sure. That thing is so hot that almost the entire pcb is too hot to touch. The memory absolutely smokes, and I would think adding heatsinks just about anywhere wouldn't hurt. Not the memory specifically in that case but just the whole damn card.

I still can't get over the 6800 properties screen where the auto-shutdown temp defaults to 125 C or something. Graphics cards are just nothing like cpus these days.