clouding always occur, its not algae lai, its some kind of chemical or others i think, cause if use hydrx sure cloud super fast.... i heard tht use biocide + distilled water is the best way to prevent clouding, i think most ppl facing clouding tubes problem, even how clean and pure is ur loop, clouding oso occur, if u wan ur loop look nice and cheat urself tht no clouding occur, try use UV dye

no la... recently finish all midterms test and assg, so fork out some free time to test all the setups lo, not disassemble whole loop ma. just part of the loop.

still dont understand why clouding happen...before this i make an experiment with chlorine gas go through pvc tube...in a minute the tube become clouding... it is not clear as ori before experiment... so i sure that cl2 cause it not algae...

need to check which chem compound can cause it happen..

why did you pass cl2 in the first place?
normally our loops dont even have chlorine inside unless youre not on distilled water.

that during study la...at school not during setup wc la....compound group 7 show it reactivity at room temp with pvc....

it is posible on the compound of coolant itself..

Simply because the caveat for using a double pump setup is for complex loops which will slow down the flowrates and drop the pressure rates down by the time the water returns to res. The reasoning behind becomes clearer as I explain:

A pump like all devices is a device that requires energy to perform work. In this context, work is done to not ONLY pump the water, but ALSO to draw water from source. The more difficult it is to draw water from source, the more work required for pump to do that task and reducing available work to pump the water. This is based on the understanding of Newton's Law which is as follows

energy supplied (electrical) = total work done + energy loss (noise, heat, kinetic)

where total work done = work done to draw water + work done to push water

Once that is understood, we can begin to see why a dual inline pump setup gains from. The primary pump's purpose is to draw the water out of the res and supply water to secondary pump. Since there's ready available pressure at secondary pump's inlet, less energy is wasted on drawing and more available to supply head pressure at maximum quoted specs which is vital once you start going berserk and fit a CPU, NB, Dual PEG water blocks, plus a hewg rad to dissipate the heat dump effectively.

And as for fans, the thumbrule that higher cfm is better is not really applicable here. I once posted an article from XS which does state there are TWO type of rads. It be too long to explain here but if you're using a low restriction airflow rad, a lower end fan will still do the job nicely. Trust me, I've tested that theory out on a BIP and MCR220 and came to that conclusion. In fact I'm using some dumb TT SMartLED running at low rpm and yet still get good temps about 42C max on load

and BTW, if you SRYSLY want flow restriction, try a Storm's Jet Impingement nozzle. A Nirox P2800 would be pissing rat's piss on it. Or ask uzairi and his old cascade style waterblock

This post has been edited by almostthere: Sep 3 2008, 07:09 PM

icic, now my pump juz sitting below the cpu block, using stacker in an inverted atx style.... the tubings from pump to block only around 8-15cm long... using 7/16 tubings here for my secure connection between barbs and tubings...

might get thermochill pa120.2 or feser radiator if i can sell some of the old wc parts.. .

This post has been edited by lichyetan: Sep 3 2008, 07:42 PM

Offer me one of your rads and pumps ler....ekekeke

24/7 setup 3.6ghz..
intel burn test 5 loops... ambient temp 28C.. =)

Max Temp : 59C
Min Temp : 34C

This post has been edited by IcEMoCHa: Sep 4 2008, 10:54 AM

I rarely clean my loop, 2-3 times in a year i guess. Im using Petronas batt water alone without any additives, better and less hassle.

Correct, both blocks are damn restrictive(im using Storm V1 currently) and changing from Nirox P2800 and a Laing D4 would make a really huge difference. Using a high Hmax pump is needed to overcome the restrictiveness.

Your current setup is enough bro, just need some tweakings and adjustments.

my current setup,
Spec:
D-Tek Fuzion v1 with 4.4mm Nozzle,Swiftech MCP-650, Swiftech MCP-655B, Aquaxtreme MP-01, Swiftech MCR-220QP, Swiftech Micro Reservoir

any tweakings needed ? now getting 13-14c delta for 3.6ghz Q9550 @ 1.36v core voltage + 4850 stock

No offense but still abit overkill. But temps how? Smiliar to last time or worse? Both CPU and GPU. And btw, for such a short loop (it is short), I think you can stick to one D5 since the gravitational flow of the water from res would ensure constant supply and reduced suction power loss (which is almost negligible in your case)

And have your tried mounting the rad externally?

This post has been edited by almostthere: Sep 4 2008, 07:05 PM

the temp for Q9550@3.6ghz@1.36vcore + 4850@ stock,

both CPU and GPU @ same loop and stressed...

10 mins stress time

almost same as parallel setup of radiator and better thn series setup...

but i think with this setup, i mean radiator at end of the loop, mayb another radiator in parallel will make temp even better since the block are now getting enough pressure and flowrate, but i planning to sell another rad d... lol

i think i'll stick with 2 pumps... nampak ganas but i think with 4.4mm nozzle, its also kinda restrictive according to xtremesystems and martinsliquid.com

*addons

the temp after load drop instantly as i stop the stress, but 1c-2c higher thn temp b4 fully stressed, gpu back to same as previous idling temp as soon as i stop the application...

if aircooling or previous setup, i guess the temp wont drop so fast...

This post has been edited by lichyetan: Sep 4 2008, 07:29 PM

Neh, I'd doubt you'd gain much more by adding the other rad back in TBH. What I'd suggest is exposing it in the open so that it's able to thermally dissipate into free air easier and pair it with the optimal set of fans for it. Thermally you can only dissipate that much heat so as long as heat absorbed from surroundings isn't of considerable amount. Just do a simple feel test of the end tanks. If you can hardly feel any warmth at the outlet end tank, it's already doing a good job

And in terms of restrictiveness, rest assured, with that short a length from pump to block, you'd won't notice the difference between a single and dual pump setup. I'd say for your case the losses in terms of available head pressure is at most about 2-3% (There's always the overvolt option but bear in mind beyond 12.4v AFAIK is rather pointless). Like I said, a dual inline setup is for total WC solutions (e.g CPU+NB+SLI/X-Fire)

Reply to addon: There lies the possible tweak. Fix the heat dump still stuck at rad. Your scenario indicates that there's a certain amount of retained heat in water.

This post has been edited by almostthere: Sep 4 2008, 07:38 PM

ok, thanks for the advice, but according to this estimator over at xtremesystems

dual pump keep me above flow rate of 1 gpm while single pump doesnt with 4.4mm nozzle... and according to tht guy, 1gpm above to keep the cooling solutions at optimal efficiency i guess, gotto check the web again...

mayb i'll stick with dual pump so tht can add another GPU block in the future if i crossfire or NB blocks for total wc solutions...

*actually i wanted to try D5 only by taking out D4, but lazy d since temp now already quite statisfying, mayb sell the mcr-220qp and other old parts to recover a bit $$ after losing so much $$$ on overkilled parts...

This post has been edited by lichyetan: Sep 4 2008, 07:45 PM

liechy, PM that martin's flow chart calculator. Something's not right here

UPDATE: Checked the chart out, the assumptions using that excel sheet are abit wild and inconclusive. But answer me a few questions

1. Why your D5 is running at setting 4 instead of 5? Difference is by about 0.15GPM

2. Do you understand what the flow rate calculation is for? It's an average BTW

This is because alot of the test data is not yet concluded based on my checks. So to say it's a good measure is quite iffy. And I haven't checked that test data for the nozzle kit yet (XS seems to be down ATM)


I'm not disputing your theories on your setup but if I were you, I'd retest personally all parameters to decide on an optimal point. In fact since I'm of the simplicity sort of person, I'd ditch the two pumps and get a DDC 3.2 with Petra's Top (or whichever good un you can obtain) installed. That alone is worth 1.20GPM at the very least. That's 0.1GPM less then a dual inline D4 setting 4 setup. Play around again with the chart and PM me if you don't understand anything

This post has been edited by almostthere: Sep 4 2008, 09:47 PM

My load temps using 4x sp2004 at 3.7ghz

This post has been edited by uzairi: Sep 4 2008, 09:19 PM


Attached thumbnail(s)

oh... my D5 is MCP655-B which the version without the speed control, so its basically the basic version, its rated at speed 4 of the mcp655, and also D4 is actually same spec as MCP655-B, cause if u notice, MCP655 have higher head pressure thn MCP-655B so as to D4...

yup, i know its overall flowrate... but to get overall cooling correct, average flowrates also takes into account rite ? i mean if cpu block functioning at optimal range, my gpu block might suffer or my radiator didnt work as efficient as higher overall flowrate rite ?

I've been thinking to get the DDC3.2 before, but due to its reliability problem, i gave up the plan and ended up on dual pumps(D4 & D5 basic).

about the nozzle kit, martins over at xtremesystem there tell me to give up 4.4mm and use the larger nozzle instead if running multiple block... but its all about overall performance, if wan more perfomance on the cpu block, 5.5 and 4.4 mm are the choice, if wan overall performance of the whole loop, use washer only instead of nozzles...

so the main reason pushing me for dual inline pump is i want the 4.4mm nozzle yet wanted to use more restricitive mp-01(compare to MCW-60, i dont like the mounting style of mcw-60 to be honest) on my gpu, so ended up bought another D5 basic to pair with my original D4 to gain higher overall flow...

*addons:
isnt dual pump give us more safer loop, i mean if one pump dead(*touchwood) the CPU wont overheat as the backup pump still running rite ? its more redundant setup rite ? just my idea, dunno whether u all agree onot...

This post has been edited by lichyetan: Sep 4 2008, 10:55 PM

It's getting late and I still work tomorrow but here we go

The overall flow rate calculation is an average, not displaying effective flow rate at entry of waterblock and post block.. What I'd rather do is to use the use the chart to calculate what's the flow rate post block. That's how I worked out your flow rate. That's the figure post outlet of your fuzion and if that's good enough for your MCP60, then it should be fine. That's how I view the table's usefulness.

As for the nozzle kit, you have to ask yourself how much your willing to give up just for that extra squeeze. I'd say for such a short and very low pipe restriction due to minimal bend, you might be better off without the nozzle. At such length, it's likely you're still getting laminar flow so you're having jet effect at inlet. That nozzle might worsen things since at such length, it might be just causing back pressure, hence reducing effective flow and effective head pressure at inlet. that nozzle imho would suit best for normal nearly 1ft length arcing tubing in ATX casings since by about 8-9", the flow would be experiencing turbulent flow internally hence slow down of flow rate thanks to turbulence resistance and wall friction combined. That's where the nozzle's purpose kicks in as it accelerates the slowed down water flow. Because remember, the Fuzion still utilises a flow thru method where flow rate is more important then flow thru pressure. Flow thru pressure only applies for jetting based blocks

I agree with the MCW60's mounting method being a pain but it's not much different then mounting any normal waterblock if you asked me. In the end, you have to really ask yourself, at what cost are you going to pay to get the best? Would it be the cash spent experimenting the various blocks or the time and effort to tweak and tweak to get that edge?

Dual pump not necessary safer to be honest. You want safe? Go AC based instead, the ultimate in safety (Japan made Iwaki's anyone? LOL). At least you know the bugger works or not before you power up. And the DDC 3.2's problem you wanna know why? Very simple really. It's a finely engineered pump which utilises the 3 pin connection for mobo's. The problem with that is that these motors bloody hate variating voltage rates which mobo outputs are notorious for. Feed it with clean current at a fixed rate and it'll work for quite sometime.

IF you wanna really learn liechy, I'd suggest you'd take some time and get your hands dirty in experimenting various setups until you get your best setup. And you have to try every variable down to mounting technique and thermal paste application. Trust me, the biggest tweaks lie in the least likely places

and BTW, I know how your rigs looks like because of this



and I was away ahead of you guys

This was back when Athlon XP ruled









This post has been edited by almostthere: Sep 5 2008, 12:38 AM

Bro, no pics of the ghetto bay reservoir ar ?

as per requested