dual heatercores a good idea?

[rxtrom]

Hey guys, got yet another Q for you, i have been doing some searching, but have not come to a conclusion yet about it. I want better temps than my 39C load. So i was thinking about getting another rad/heatercore. My system would go like this

Pump -> RES-> HC-> HC-> CPU - > Pump

I was also thinking about useing the 2 HC's as one and getting Y adapters

Y adapter -> HC HC-> Y adapter

AND

FAN-> SHROUD-> HC -> (siliconed) HC-> shroud ->FAN
because i do not have the room to mount them seperatly

LMK your thoughts thanks ROB
(probably going to get a WB for the 9700 pro soon too)

[Korndog]

well.. a second HC would more likly aid your system dumps more heat the the HC can handle. if that was the case, your temps would increase continiously. Since i doubt that's the problem, an extra HC may actually increase temps due to decrease in flow rate.
if you were to upgrade to a larger pump as well as a second rad, then you would get results

[E-Licious]

Originally posted by rxtrom
I was also thinking about useing the 2 HC's as one and getting Y adapters

Y adapter -> HC HC-> Y adapter

Yes to this.


Originally posted by rxtrom
FAN-> SHROUD-> HC -> (siliconed) HC-> shroud ->FAN
because i do not have the room to mount them seperatly

Not sure how well this will work. It'd prob still be better than one heatercore though.

[E-Licious]

Originally posted by Korndog
well.. a second HC would more likly aid your system dumps more heat the the HC can handle. if that was the case, your temps would increase continiously.

Huh?

Originally posted by Korndog
Since i doubt that's the problem, an extra HC may actually increase temps due to decrease in flow rate.

Not if he puts them in parallel. That would increase flow rate.

[rxtrom]

i have a maxijet 1200 295GPH, what would be an upgrade from this?

[slater3333uk]

I think you would be better off with

Res -> Pump -> CPU -> HC -> HC -> Res

You want the pump drawing directly from the res and straight into the cpu block right?

i would recormend against it as the decrese in flow rate will attuay cause a rise in temps. Dual rads will only be any good in a system that has a high heat load on it eg. peltiers or dual cpus.

[Korndog]

Originally posted by E-Licious

Not if he puts them in parallel. That would increase flow rate.

adding stuff to your system, regardless of how its added, will decrease flowrate.

Originally posted by E-Licious

Huh?

heater core removes heat, obviously. if can't remove all the heat dumped into the water by componants by the first cycle, then obviously the Heatercore is inefficiant for the amount of wattage your system produces. If your Heatercore can remove the the heat from you water efficiantly a larger or a second heatercore will not remove more heat then the first.

[Cyrix_2k]

I'd say a Hydor L30 or an ehiem 1250 would be an upgrade.

BTW, I agree with e-licious.

[Korndog]

Originally posted by rxtrom
i have a maxijet 1200 295GPH, what would be an upgrade from this?
you may want to consider a danner mag for an upgrade
the maxijet has a 69" max head
the mag 3 has a 126" max head
so although the mag 3 is only 350gph, which isn't much of an increase in max flow, the head is almost double which means with pressure it'll outpreform the maxi by a large margin.
just for example, the maxi jet would probably get like 60-80gph with two heatercores, two waterblocks, and 4 feet of tubing. if u replaced it with a danner mag drive 3, you would be getting about 200gph in the same system. Even upgrading to a danner mag 3 without adding a HC will improve your temps, expectually if you have a block that benefits from it like the WW, cascade, or RBX.

[E-Licious]

Originally posted by Korndog


adding stuff to your system, regardless of how its added, will decrease flowrate.

Incorrect. When you have two things in series, the total resistance is greater than either one. When you run two things in parallel, the total resistance is less than the resistance of either one. The exact equations for 2 resistances in an electrical circuit are:

Series: R1 + R2 = Rtotal

Parallel: (R1*R2)/(R1+R2) = Rtotal

Do the math and you'll see that the total resistance is way lower in the parallel case. I'm sure that the equations for a fluid system are equivalent to these. Lowering the total resistance will increase flow rate for the system.

But, I wouldn't advise running anything in parallel with the CPU block. You get more flow in the system, but you get less flow in each individual branch. Think of it this way, if total flow in the system is 1 GPM, you can't have more than 1 GPM in either of the branches, unless one branch has negative flow rate (going the other direction). So, since you want as much water going through the CPU block as possible, I wouldn't put anything in parallel with the CPU block. Some do, but I wouldn't.

Another advantage of running two rads in parallel is that both rads are getting the hottest water in the system. We know that heat dissapation is fastest when the delta T is highest, so you want the hottest water to be going through both rads.



Originally posted by Korndog

heater core removes heat, obviously. if can't remove all the heat dumped into the water by componants by the first cycle, then obviously the Heatercore is inefficiant for the amount of wattage your system produces. If your Heatercore can remove the the heat from you water efficiantly a larger or a second heatercore will not remove more heat then the first.

This is true if the water temp at the outlet of a single radiator is the same as the air temp. However, this is hardly ever the case. The lower the delta T, the slower the heat dissapation. Need I explain more?

[E-Licious]

Originally posted by Korndog
just for example, the maxi jet would probably get like 60-80gph with two heatercores, two waterblocks, and 4 feet of tubing. if u replaced it with a danner mag drive 3, you would be getting about 200gph in the same system.

Can you please show how you are getting these values?

[Korndog]

Originally posted by Cyrix_2k
I'd say a Hydor L30 or an ehiem 1250 would be an upgrade.

BTW, I agree with e-licious.

eheim has a 79" head and hydor has 75"
not much of an upgrade, although these pumps are better quality then the maxijet. i would still go with the danner mag 3

[FIZZ3]

I'd prefer a kingsize heatercore over a messy dual setup if I wanted increased cooling capability from my system. Less messy, less restriction, less unneeded loop length.

You can only go so far with watercooling though, keep that in mind. 39C load is not so bad IMO (can't give detailed info here, because the temp measurement is fairly uncontrolled ;) ), and without active cooling you have a bottom of ambient plus a few degrees anyway.

Also instead of multiple fans on multiple heatercores, you might want to use stronger ones on the existing core instead. Airflow over a radiator helps a lot... but of course you'll have to face the extra noise then too.

[Korndog]

Originally posted by E-Licious


Can you please show how you are getting these values?
its just an example.
http://www.reefcentral.com/calc/hlc2.php

please keep in mind that i'm simply stating facts. these are not my opinions.
if u want my opinion, heatcores, expectually the large 6x7 ones, are more then enough for a cpu, gpu, NB water cooling system as long as u have adequate airflow. putting your money and effort into a stronger pump and a better waterblock would show better results.

[Korndog]

using the flow calculator set at 2ft vertical, 1ft horizontal and 3/4" tubing the Mag3 gets 321gph and the maxi gets 52gph, i guess i was too generous with my previous example.

[E-Licious]

That's a handy calculator.

Stating facts is one thing. That's ok, but when you make an incorrect statement, I feel the need to correct it if I can.

[Korndog]

Originally posted by E-Licious
That's a handy calculator.

Stating facts is one thing. That's ok, but when you make an incorrect statement, I feel the need to correct it if I can.

of course, i know what you mean. the previous numbers were from my own old calculations (sigfiged) with elbows and y connectors included, i was probably off because i used the flow graph instead of that calculator, so i was using it as an example. no one should expect similar GPH on their system, but you get the idea.

[Graystar]

Two heater cores with the "Y" is the best way to setup what you have.

The pump is fine.

[E-Licious]

Originally posted by Korndog


of course, i know what you mean. the previous numbers were from my own old calculations (sigfiged) with elbows and y connectors included, i was probably off because i used the flow graph instead of that calculator, so i was using it as an example. no one should expect similar GPH on their system, but you get the idea.

I was talking about the parallel resistance

[pelikan]

If I were you I would not be concerned with those temps. I'd be busy raising my vcore.

If you want more radiator power why not just get a hc that fits two 120's like a caprice?

[rxtrom]

i don't have room for that size

[Malpine Walis]

It has been a long time since college but the formula that I remember for parallel resistance is 1/R1*R2. I could well be wrong on that and plumbing variables will play a bigger role in water cooling that with electrical connections. Whatever the math is, putting the cores in series will increase the length of your system and drop the flow rate.

However, parallel heater cores will have another advantage over series. In a parallel setup, the entry temperature will be the same for both cores. So both cores will have the same “cooling profile”. Therefore, both cores will drop the coolant temperature by the same amount.

In a series setup, the second core will have a lower entry temperature because the water has already been cooled from passage through the first heater core. So the second core will not have the entry temperature of the first core. Therefore it will not be able to provide the same amount of cooling as it would in a parallel setup.

To summarize:

Parallel, both cores do the same amount of cooling. Total length of the system is minimized, thus reducing friction and providing a higher flow rate.

Series, the first core does most of the cooling. Total length of the system is greater, thus increasing friction and dropping flow rate.

[E-Licious]

Originally posted by Malpine Walis
It has been a long time since college but the formula that I remember for parallel resistance is 1/R1*R2.


I assure you the formula is (R1*R2)/(R1+R2). If we had two equal resistors, and put them in parallel, the total resistance would be half the value of one resistor. Plug values into the equation and you'll see.

Originally posted by Malpine Walis

Parallel, both cores do the same amount of cooling. Total length of the system is minimized, thus reducing friction and providing a higher flow rate.


It's not so much that the total length of the system is reduced, or the reduction in friction that provides a higher flow rate. It is the increase in cross sectional area at the parallel portion of the system. Look at it this way, the two rads have 1/2" diameters. If we have one rad only, the area that the water has to flow through is 1/16*pi inches. If we put two rads in parallel, we double the cross sectional area to 1/8*pi inches. We are very aware that larger diameter tubing increases flow. Putting components in parallel increases flow by the same principle.

[Malpine Walis]

As I said, it has been a long time since I was in school. I looked the formula up on google and I couldn't find either formula. Oh well on that.

As far as system length having an effect on flow rate, what would the effect be if you compared two radiators is series aganst only one radiator?

I would expect that you would have a lower flow rate in the series setup vs. the single radiator. Also the second radiator would still have the lower entry temp that I noted above.

[E-Licious]

Yup, those points are correct.