Coolant temps question...

[Boonstick]

I've just added a second graphics card into my loop and I am wondering if my coolant is getting too hot.

Im running 2x 780Ti and a 4670K oc @ 4.6ghz.
I have 1x XSPC AX240 & 1x XSPC AX360 rad with fans in push.
My loop is pump -> cpu -> crosschill motherboard -> 240 rad -> gpu -> gpu -> 360 rad -> pump

Running Heaven benchmark my GPU's get up to around the 42C mark. Before it was about 34C. Ive adding in a coolant thermometer and on full load it reads about 38/39C. Should I be worried about this? Will I have to add more rad capacity? I knew it would get hotter but I'm just a bit concerned about the coolant temps...

Thanks in advanced!

 

[remember300]

are all your rads push pull?

 

[Bartacus]

I don't think the GPU temps are a problem, as 42C is excellent. Not sure about the coolant temp though. It isn't a problem for sure. What are the ambient temps in the room?

 

[Boonstick]

are all your rads push pull?

No, I can only have them in push as I'm very limited on space. I may have to change the fans from the SP120 quiets to the performance ones.

It's at about 22/23C. I just did some more testing (played a game for 3mims) and the coolant got up to about 35.5C. I may take the crosschill out of the loop as I don't really think it's doing anything useful except adding heat into the loop

 

[Remmy]

The main thing that's important really is the temperature of the components, and as long as their temperature is acceptable then the coolant is doing its job fine, and I wouldn't worry about it.

 

[remember300]

there is alot more heat being generated, so the fact there is only a few degrees in it does not seem so bad and the temps are still lower than some air solutions, whats it like under heavier load?

tbh those temps dont worry me to much considering, but if you wanted to ramp up the air flow then go ahead no harm i guess

 

[Dragolu]

An ambient to coolant delta t of 10 c is considered acceptable. More rads will lower your temps but in the end its what you are willing to live with. You are way below the tolerances of any of the components so it is not a problem.
As it has been said previously the order of the components is all but irrelevant.
Your temp rise is driven by the GPU's as they are used to running form 60 to 90 c on air. The CPU usually runs lower and the crosschill is probably not adding much heat. You could separate the loops and tolerate higher GPU temps but you loose overall cooling performance.
I would leave it as is but you can add more rads if you want a higher cooling capacity and lower delta.

 

[SuB]

Temps seem fine to me dude. Nothing to worry about. You'll know when you hit the thermal limit of the rads because your water temps will go up MUCH higher and your components won't level off temp wise.

 

[Boonstick]

Thanks for all your advise guys.

Yeah I'll just have to see how things go. I guess I've been used to an ultra cool, ultra quite rigs for the past few months.

I can't do anything about the amount of rad capacity and id be pushed squeezing in more fans. If it comes to that I'll have to change my case. I'm using a Corsair 540 Air and it's getting seriously tight in side.

Just a thought...

Currently the front mounted 360 rad has 3x SP120's pushing into the case. I can't get it in push/pull as the 240 rad in the roof is too close to the 360 to allow for another fan. If I was to spin the front 3 fans round into pull and only mounted 2 more sp120's on the mounts that I can get to, would that make much difference to the temps? I guess it would be worth a try but I don't really want to buy more fans only to have to send them back.

Have a look at my build log to see what I'm on about with the fan clearance. Pics are a little out of date

http://forum.overclock3d.net/showthread.php?t=59440

 

[qhfreddy]

I may take the crosschill out of the loop as I don't really think it's doing anything useful except adding heat into the loop

I doubt it would add much heat, but the biggest thing is probably the resistance it adds.

 

[Boonstick]

I doubt it would add much heat, but the biggest thing is probably the resistance it adds.

I think my pump is man enough to handle the resistance to be honest.

I've messed about with my fan configuration and that has seems to make a fair bit of difference. May have to replace all my fans tho

 

[qhfreddy]

I think my pump is man enough to handle the resistance to be honest.

I've messed about with my fan configuration and that has seems to make a fair bit of difference. May have to replace all my fans tho

It's not really about whether the pump is good enough, the reduction in flow rate through a highly restrictive area will usually scale with how much flow there is. It should reduce your flow rate by approximately the same percentage regardless of how good the pump is.

 

[FTR.4500MHz]

One thing to watch out for tho is that your coolant doesn't go above the recommended operating temp of your pump, other than that your fine with your actual liquid temps imo.

i.e for a D5 60° is the limit.

 

[Boonstick]

It's not really about whether the pump is good enough, the reduction in flow rate through a highly restrictive area will usually scale with how much flow there is. It should reduce your flow rate by approximately the same percentage regardless of how good the pump is.

Ok, I see. So I guess what you're are saying is that regardless of what testing I do with the pump on different speeds, it will make next to no difference if it is a highly restrictive loop?

One thing to watch out for tho is that your coolant doesn't go above the recommended operating temp of your pump, other than that your fine with your actual liquid temps imo.

i.e for a D5 60° is the limit.

Yeah, my coolant and pump operate at 60C and I'm well below that...

 

[qhfreddy]

It will make a difference, but lets just make some numbers up to explain.

Say just the pump (no waterblocks at all) is 100L/min.

With the waterblocks (excluding the motherboard your flow rate is 50L/min.

When you add the motherboard your flow rate goes down to 25L/min.

Now lets say you turned your pump up to 200L/min. All the values above should (approximately) double. 200, 100, 50 respectively.

What would not happen is that you subtract the flow rate, so you would not get 200, 150, and 125 from adding 100. Instead you are (again approximately) doubling the values.

Agaim, these are just approximations, and the scaling will change depending on the exact pressure the pump outputs. A high pressure, lower flow rate pump will most likely be affected less by the restriction than a lower pressure, high flow rate one.