|
|
General Liquid/Water Cooling Discussion For discussion about Full Cooling System kits, or general cooling topics. Keep specific cooling items like pumps, radiators, etc... in their specific forums. |
Thread Tools |
03-19-2002, 01:20 AM | #1 |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
Eheim 1060 - should i?
might be able to get a 1060 cheap second hand..so I was thinking if its going to be worth it to upgrade from my current 1250 that i'm using now (more like PLANNING to use once my blocks come ) will the increased waterflow be negated by the increased heat output from the bigger pump?
(using a simple single loop setup, with just one waterblock, pump and radiator - BIX, nothing else) |
03-19-2002, 02:09 AM | #2 |
Cooling Savant
Join Date: Mar 2002
Location: North Vancouver BC
Posts: 234
|
Waterflow = Good
I guess unless it's major overkill and your pump is contributing a lot of heat, but hey, that's why inline is good compared to submersible, heat doesn't go into the water system as much.
__________________
"mooooo" said the cow. ERTW - UBC P4 2.4B @3.01Ghz 167FSB :: Abit IC7 :: 2x256MB HyperX PC3700 :: ASUS 9600XT :: WD Raptor 2x36GB RAID0 |
03-19-2002, 02:30 AM | #3 |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
but the 1060 is 50W compared to the 1250's 28W..won't that generate a lot of extra heat that will end up in the water?
|
03-19-2002, 03:14 AM | #4 |
Cooling Savant
Join Date: Mar 2002
Location: North Vancouver BC
Posts: 234
|
Well if the pump is used as a submesible pump, (under the water), then all that heat will go into the water, 100% of it; however, if you are using the pump inline, (water in tube/water out tube), then not all of the heat will end up in the water, some will go into the surrounding air. But, logically it would make sense that even in an inline setup, there would be an increase proportional to 28:50, although that may not amount to very much, if you want an idea, leave your current pump on, and feel how hot it gets, then you'll have an idea of how much heat is going into the air rather than the water (inline). Obviously more heat will come from the 50w, but the actual increase in the water temperature might be almost negligable.
But if anyone has any specific experience with this, we're all ears
__________________
"mooooo" said the cow. ERTW - UBC P4 2.4B @3.01Ghz 167FSB :: Abit IC7 :: 2x256MB HyperX PC3700 :: ASUS 9600XT :: WD Raptor 2x36GB RAID0 |
03-19-2002, 04:13 AM | #5 |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
yeah..anybody using the eheim 1060? how's the noise and performance?
thanks btw, khledar |
03-19-2002, 03:56 PM | #6 |
Thermophile
Join Date: Oct 2001
Location: Nuu Zeeelin
Posts: 3,175
|
I'd get the 1060, even if you aren't going to use it in the current computer it's a worthwhile investment for the future.
all eheims are virtually silent
__________________
2x P3 1100's at 1400, Abit VP6, 2x Corsair 256mb PC150 sticks, 20gb 'cuda ATA-III, 2x 40gb 'cuda ATA-IV in raid 0. 20" Trinitron. No fans 2x 2400+ at 2288mhz (16.0 x 143), Iwill MPX2, 2x Kingmax PC-3200 256mb sticks, 4x 20gb 60gxp in Raid 5 on a Promise SX6000. Asus Ti4200 320/630. Cooled by Water |
03-25-2002, 03:01 AM | #7 | |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
Quote:
|
|
03-25-2002, 07:33 AM | #8 |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
and if anybody has had any actual experience with this pump it'd be greatly appreciated if you could give your opinions too..
|
03-25-2002, 08:05 AM | #9 |
Big PlayerMaking Big Money
Join Date: Aug 2001
Location: irc.lostgeek.com #procooling.com
Posts: 4,782
|
Magdrives don't produce dangerous levels of pressure. Some people like this but honestly the reson I would move to a larger pump would be to get more pressure in hopes of turbulence and not to get more flow rate.
As for whether to go to ever larger pumps, I personally prefer to remove flow restrictions (drill out the fittings and/or move to larger fittings) and use smaller pumps. That way you are dumping less heat into the case, taking up less space, and spending less money on pumps. With the Maze2 (and I suspect most channel style blocks) there really is no benefit over 1-1.5 GPM of actual flow. Take a look at this picture (Maze2/heatercore/Eheim 1250 with flow adjusted with ball valve): Notice that performance hits a minimum and then plateaus. What happens is that you reach a point where water flow is sufficient to minimize the difference between water temp and waterblock temp. Any increase in flow over this point only serves to reduce the effectiveness of your radiator's heat dissipation. If my pump had been large enough, you would see performance actually decrease at really high flow rates due to a combination of the increased heat from the pump and decreasing ability of the radiator to lower water temps. Huge pumps are for evaporative cooling IMO where you are moving water vertically a goodly distance. FOr an internal setup the 1250 is more than enough pump as long as you have an eye to optimizing flow in the loop. |
03-25-2002, 09:38 AM | #10 |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
thanks for the info!
actually I was thinking of using dual heater cores in parallel. thats why I figured I might be needing this 1060. anyway, if i do get the 1060..how do I "throttle down" the flow rate to a level that's optimum for me? ball valve? how do I do that? advice? |
03-25-2002, 01:06 PM | #11 |
Thermophile
Join Date: Oct 2001
Location: Nuu Zeeelin
Posts: 3,175
|
pH, billA showed that most rads were best at around the 1 - 1.5gpm mark, while his waterblock kept on getting better with more flow. Maybe you should test each component individually.
with dual heatercores, say the flow in each is 1.5gpm, the flow in the block is 3gpm. great performance.
__________________
2x P3 1100's at 1400, Abit VP6, 2x Corsair 256mb PC150 sticks, 20gb 'cuda ATA-III, 2x 40gb 'cuda ATA-IV in raid 0. 20" Trinitron. No fans 2x 2400+ at 2288mhz (16.0 x 143), Iwill MPX2, 2x Kingmax PC-3200 256mb sticks, 4x 20gb 60gxp in Raid 5 on a Promise SX6000. Asus Ti4200 320/630. Cooled by Water |
03-25-2002, 03:55 PM | #12 |
Big PlayerMaking Big Money
Join Date: Aug 2001
Location: irc.lostgeek.com #procooling.com
Posts: 4,782
|
Yes and no Brad. BillA has better resolution of temps, but he sees basically the same thing:
"The trend lately in pump selection seems to be towards over-sizing, with the rationale that the output can then be throttled back to find "the sweet spot" where the maximum cooling (minimum temperature) is obtained. It has been shown that some (few) radiators do indeed seem to have a point of higher performance, but this is not so for most. To understand the watercooling system "sweet spot" concept, another bit of information is needed - the relationship between the coolant's flow and heat transfer in the waterblock. The graph below illustrates this for a popular waterblock, where the difference is shown between the waterblock baseplate and coolant (inlet) temperatures vs. the flow rate. The temperature was recorded with a type T thermocouple embedded in the baseplate over the die area (per AMD's recommendations) with a 77 Watt heat load applied through a 100 mm² heat die. For this test, the inlet temperature was held constant, so the reduced temperature difference is in fact the reduction in baseplate temperature attributable to the higher coolant flow rate. The very significant benefit from higher waterblock flow rates is clear, but this is directly the opposite of the optimum for radiators. The "sweet spot" determined by throttling the pump output is simply the balancing of the compromised performance of both the waterblock and radiator. This point will be different for each combination of components, and will not exist at all for systems where the waterblock performance gain exceeds the radiator impairment; for such systems more coolant flow is always better, much more. " (from http://www.overclockers.com/articles481/index06.asp ) So yes in the absence of a radiator's need to dissipate heat more flow is always better for a block. Even still, if concepts of price/performance (or space/performance) are considered then it is still much more effective to only try to get in the region of optimal performance (1-1.5 gpm for both BillA and my own block testing) and not end up with 1" tubing and fittings and a 1000 GPH pump to try and get that extra little bit of performance. Why? More cooling is always better, right? Well, it isn't possible to completely isolate the waterblock from the heat dissipation vs flow rate curves of the radiator in the real world, hence this whole concept of sweet spots discussed by BillA at length and myself much more modestly above. Ok I just woke up from a nap...coffee and then more of this in a bit. |
03-25-2002, 05:38 PM | #13 | |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
more?? hehe..
Quote:
|
|
03-25-2002, 06:08 PM | #14 |
Big PlayerMaking Big Money
Join Date: Aug 2001
Location: irc.lostgeek.com #procooling.com
Posts: 4,782
|
Sure. Basically you have a situation (for the majority of blocks and radiators) where:
1)Increasing flow increases waterblock performance (this is usually a big increase followed by a plateau or very slowly sloping increase after that). 2) Increasing flow decreases radiator heat dissipation And so the trick is to find the flow rate where the two factors are both as close to optimal as possible. Looking at my data for Maze2 and Eheim 1250, that seems to be around 1.1-1.3 GPM. It therefore makes no sense to put a 1060 into that loop. I would end up just throttling the flow anyway (and only serving to increase the heat in my case from the pump). |
03-25-2002, 06:47 PM | #15 | |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
Quote:
anyway, didn't I read somewhere that the beauty of inline pumps is that the heat produced is hardly gone into the water..instead its mostly conducted away by the air around it? |
|
03-25-2002, 07:37 PM | #16 |
Cooling Savant
Join Date: Nov 2000
Posts: 130
|
phaestus, do you have tests that confirm increasing the flow rate decreases the radiator performance? I would like to see them if you do.
__________________
ondaedg@procooling.com |
03-25-2002, 07:54 PM | #17 |
Big PlayerMaking Big Money
Join Date: Aug 2001
Location: irc.lostgeek.com #procooling.com
Posts: 4,782
|
Yea this article: http://www.overclockers.com/articles481
Here is the easiest to follow figure: http://www.overclockers.com/articles481/dissvsflow.gif BillA is my hero! |
03-25-2002, 08:19 PM | #18 |
Cooling Savant
Join Date: Nov 2000
Posts: 130
|
I have read that article. BillA does do alot of work in this area. Great info. The problem is that increasing the flow rate lowered the radiator's performance less than 10% in some cases and increased their performance in others which leads me to believe that perhaps this isn't really an issue that should be addressed when selecting a pump. So in my opinion, opting for the more powerful Eheim (that is the original topic right?) is not going to hurt his performance to the point that he should have gone with the smaller pump. Furthermore, it gives him headroom to add more blocks for other components (if he feels like gettin crazy) . Technically, you are right that it did in some instances lower the radiator's btu/hr rating. I just think that the difference is so minimal, that it really should be a non issue. Feel free to argue otherwise. =]
__________________
ondaedg@procooling.com |
03-25-2002, 08:27 PM | #19 |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
ph, I was just looking at the graph. it seems like the radiator temps sorta plateau after about 1.5-2gph..
I picked this graph, cos the heater cores I'm planning to use will probably be closest to the big momma scores.. whereas, waterblock temps drop much more with increased flow SO is it not fair to say that with increased flow rate, waterblock temps drop much more than radiator temps go up? ie. is it not right to say that in the end, increased water flow, generally speaking, SHOULD produce better temps? |
03-25-2002, 08:34 PM | #20 |
Big PlayerMaking Big Money
Join Date: Aug 2001
Location: irc.lostgeek.com #procooling.com
Posts: 4,782
|
That depends also on the static pressure of the fans that you use as well though. The bottom two curves on the heat dissipation curve are those more readily accessible with 12VDC axial fans...
|
03-25-2002, 08:37 PM | #21 | ||
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
Quote:
Quote:
|
||
03-25-2002, 09:12 PM | #22 |
CoolingWorks Tech Guy Formerly "Unregistered"
Join Date: Dec 2000
Location: Posts: 2,371.493,106
Posts: 4,440
|
better instrumentation Last edited by BillA; 03-26-2002 at 06:41 AM. |
03-25-2002, 09:16 PM | #23 | |
Cooling Savant
Join Date: Mar 2002
Location: Singapore!
Posts: 124
|
Quote:
|
|
03-25-2002, 10:22 PM | #24 |
Big PlayerMaking Big Money
Join Date: Aug 2001
Location: irc.lostgeek.com #procooling.com
Posts: 4,782
|
I just skimmed through that overclockers article again; was a swiftech block used in that as well? I was thinking it was a channel tye block for some reason. The change from a channel type block to something like the Swiftech may change the improvement with really high flow rates. I have heard it proposed that the Swiftech block and a high pressure pump is capable of actually entering the turbulent regime.
That could alter the performance of the block quite a bit at higher flow rates. I guess that generalizations are something that always come back and bite one in the ass. I was incorrect about the generalizations that I made on the importance of finding a sweet spot for the cooling setup. There are some scenarios where more flow will improve things. For example, if you are using a big momma and are willing to put up with the noise of extremely powerful fans (looks like the ones BillA used were Nidec TA600DC - 960 CFM as I don't believe you can get 0.25 in H2O with the Panaflos) then you may continue to see improvement. That is a lot of noise though. |
03-25-2002, 10:35 PM | #25 |
The Pro/Life Support System
Join Date: Dec 1999
Location: Denver, CO
Posts: 4,041
|
hey Unregistered
You need to put a Http:// in the image links or it assumes the forums.procooling.com/ domain agead of it. Man I think I just heard BillA somewhere around here
__________________
Joe - I only take this hat off for one thing... ProCooling archive curator and dusty skeleton. |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|