Can I get my temps lower?

[johnb]

I've been building my computer for several years and finally have it right where I want it in terms of power, cooling, etc. My "issue" is that my temps seem a little higher than I thought they would they would be, given the watercooling hardware in my loop. I realize my "issues" is extremely nitpicky, but for the amount of time I have put into my system, I thought I would have seen an improvement in my temps.

I recently did an overhaul on my loop, adding a new reservoir and much bigger pump. My old loop was an all in one XSPC solution and ran just my proc in a Rad->cpu->Pump/Res configuration using 7/16 ID x 5/8 OD tubing. My cpu temps were great and I never had a problem.
Old loop:
XSPC RX360 Rad
XSPC Raystorm CPU Block
XSPC X20 750 Pump/Res combo
Old Temps:
CPU Idle: 22C
CPU Load: 40

Then I wanted more, so I bought a new liquid cooled R9 290x. Knowing that I would need some extra cooling between my cpu and gpu, I bought another radiator to put in between them: an XSPC AX140. Adding the new radiator did not affect my cpu temps, but my gpu would idle around 28C, which still isn't bad, and I realize it isn't.

So, knowing that I had too much running off of the little res/combo, I decided it was time for an upgrade and purchased a swiftech MCP655-PWM pump as they had great reviews everywhere as being one of the best pumps on the market. So, my new loop goes something like this: Pump->360Rad->CPU->140Rad->GPU->Reservoir->Pump->... With this new setup I thought my temps would definitely drop by a few degrees, but instead my CPU increased by 9C
New Temps:
CPU idle: 31C
GPU idle: 29C
CPU load: 55C
GPU load: 62C

I had my pump using the PWM setting going off of the CPU fan controller on the motherboard. Thinking that was the problem, I disconnected it, letting the pump run at 100%. That only dropped my temps by 1C in both idle and load situations. My rads have been cleaned and rinsed along with my CPU waterblock. I have been running the new loop for about a month now, all bubbles are out and my system is stable otherwise. Am I missing something? Thanks for any help...

Setup:
CPU: i7 3960x @4.8 - 1.4v
GPU: Powercolor LCS R9 290x @ 1200/1500
RAM: 16GB Corsair Dominator @ 2133
PSU: Corsair AX1200i
MB: Asus Rampage IV Extreme
CPU Rad: XSPC RX360
GPU Rad: XSPC AX140
CPU WB: XSPC Raystorm
GPU WB: EK-FC
Pump: Swiftech MCP655-PWM

Here is a pic of my system below:

 

[King107s]

I get about the same temps with a GTX360 and 3 delta fans.... So I doubt you'll get a lot lower. Nothing wrong with the temps though! Very comfortable temps for the CPU & GPU....

 

[ATMINSIDE]

Not sure why you're worried about those temps, they're just fine.
You've got two components that dump a LOT of heat when OC'd the way they are.

 

[RnRollie]

Unfortunately the numbers/temps dont mean anything...

Because your Ambient temp is missing... for all we know, when you measured CPU 22 °C idle / 40 °C load on the "old" loop it might have been freezing -10 °C in the room, and in that case 22/40 is BAD! Likewise, when you measured 31°C idle / 55 °C Load , the room temp might have been 30°C , and that means those numbers would be fantastic.

All this aside..., achieving a target loop temp of ambient + 10°C with a 360+140 should be easy enough.
The only way to get closer to ambient would be MOAR raddage & MOAR fannage , like a couple of MO-RA3's

 

[GTXJackBauer]

Welcome to OCFs!

Nice looking system you got there and job well done imho. You're temps look fine but as stated it also depends what ambient temps you have. It could have been cooler before or maybe the same as now. Dumping a 290 in the loop w/ that CPU will dump some serious heat. You have enough heat surface. I would mess with the speed of the fans and see if you can shed a few degrees.

Also, the 800Ds are known to have somewhat decent air flow but not that great by design. I had almost the exact setup before. (120+360) My temps were fine but did notice the air floow wasn't the best. I would replace the bottom fan that feeds the entire case to something of high CFM but silent. That might knock off a few degrees.

Those temps were done under what tests and length of time? In general we recommend Prime 95 for 30 mins on the CPU and Heaven Benchmark or Furmark if you want the extreme readings from your GPU. The only thing I can think of coming close to maxing your loop in real time would be folding or benching both at the same time. Other than that you should be good.

Unfortunately the numbers/temps dont mean anything...

Because your Ambient temp is missing... for all we know, when you measured CPU 22 °C idle / 40 °C load on the "old" loop it might have been freezing -10 °C in the room, and in that case 22/40 is BAD! Likewise, when you measured 31°C idle / 55 °C Load , the room temp might have been 30°C , and that means those numbers would be fantastic.

All this aside..., achieving a target loop temp of ambient + 10°C with a 360+140 should be easy enough.

+1

The only way to get closer to ambient would be MOAR raddage & MOAR fannage , like a couple of MO-RA3's

Aren't you being silly now.

 

[johnb]

I appreciate all of the reply's along with the welcome.

My ambient temp is 20C (about 69F). I was just shocked that this new pump wasn't able to produce better temps (or at least equal) than my old pump/res combo (My CPU is now +9C warmer with the new pump). Especially since its 1200lph (vs the X20's 750lph).

Does anyone think my pump may be inhibited due to the 90° adapters on the pump?

 

[EarthDog]

Anything over 1.5GPM flow rate is diminishing returns. So if you old pump managed that rate, a new pump wouldn't help a measurable amount.

 

[Tatsugen]

Martins has a good review on the effectiveness of 90 degree adapters here. Pretty much, every 1x 90 degree bend you have in there, adds around .05 C temps increase. So, you'd have to have around 20-30 of them to see a 1C difference without them.

Thanks for the post with your temps. This will help me out a bit.

 

[wagex]

those temps re more than fine... id even turn down the fans you can spare a few degrees to get the quiet

did you remount your cpu block since you took temps the first time? if so thats probably the issue... a pump isnt gonna make more than a few c difference, if any at all,

 

[lkrose78]

...Anything over 1.5GPM flow rate is diminishing returns. So if you old pump managed that rate, a new pump wouldn't help a measurable amount.

+1 Pumps don't really impact temps unless it's below .5GPM, with the sweet spot being 1.0 ~ 1.5GPM

...Pretty much, every 1x 90 degree bend you have in there, adds around .05 C temps increase. So, you'd have to have around 20-30 of them to see a 1C difference without them...

Again, not really an impact. My ambient is 3c over room temp and I have 660mm of rad being cooled by 600~800 rpm fans with the pump running at roughly %30 or 1400rpm which puts me somewhere around 1.0 ~ 1.5GPM. My idle is 25 average across all cores (3rd in the loop) and my 760's both sit at 22 and 24 (1st and 2nd in the loop).

Your temps are fine.

 

[EarthDog]

Room temp is ambient temperature...