345ac High Temps and mounting questions

CPU water cooling blocks...
User avatar
MUGEN
Tactical Bucket Splasher
Tactical Bucket Splasher
Posts: 212
Joined: Thu Jul 02, 2009 9:11 pm
Location: US

345ac High Temps and mounting questions

Postby MUGEN » Wed Oct 07, 2009 11:54 am

I was expecting lower temps then what I'm getting, at least around 10c less.

PC specs
I7 965@3.875 with 1.3Vcore (I'll be going back to 3.6 since I don't need the extra speed/heat)
GTX 295 dual PCB@ 670/1458/1188
OCZ Freeze used on CPU and GPU's

ERM-2K3U setup and settings
all 3/8" ID fittings/tubing
ERM-2K3U-> QDC (High Flow) M/F -> 3.5ft of tubing-> CPU 345ac -> gtx295waterblock ->3.5ft of tubing -> QDC (High Flow)F/M -> ERM-2K3U
Fan speed 1 pump 1

Room temp 74F
Idle
GPU1 38c
GPU2 36c
CPU 38C (avg of the 4 cores)

under load
furmark stability test 3 minute run
GPU1 around 49c
GPU2 around 47c

prime 95/realtemp warm up test(about 5-10 mins)
CPU 67c(avg of the 4 cores)

My questions are
1. Does it matter which way the CPU block is mounted, outlet port at the top/side/bottom?
2. Would 1/2" tubing/fittings do anything?
3. What pump speed should I be using?
4. Would upgrading to the 350AC be worth it?
ERM-2K3U + HX-1320
i7 965@ 3.6ghz CPU-360
MB-ASP6WR
GTX 295@ 670/1476/1188 VID-NX295
6 sets of No-spill QDCs
EK-DCP 4.0 Pump

Izerous
Rain Maker
Rain Maker
Posts: 1185
Joined: Tue Sep 16, 2008 10:34 am
Location: Edmonton, Canada

Re: 345ac High Temps and mounting questions

Postby Izerous » Wed Oct 07, 2009 1:52 pm

Shoudl be doing origonal tests with pump speed on max. If temps remain that high almost seems as though block for cpu is not mounted correctly, ie not correct amount of thermal paste etc.

EDIT: missed the part about the fan speeds also being at one. Tim's response better covers the issue in better detail.
Last edited by Izerous on Wed Oct 07, 2009 3:17 pm, edited 2 times in total.
Corsair 800D
2xPMP-450S > Amd FX8370 > Asus CH4-EX > AMD 6990 > AMD 6990 > 2x1080 rads > return
32GB Ram

User avatar
[Koolance] Tim
Moderator
Moderator
Posts: 451
Joined: Tue Sep 16, 2008 9:34 am
Location: WA, USA

Re: 345ac High Temps and mounting questions

Postby [Koolance] Tim » Wed Oct 07, 2009 2:27 pm

The CPU-350 would perform better than the CPU-345 with that unit. However, your temps are high for the CPU-345. You will need to increase fan speed-- I would recommend the auto setting. That heat load is too large to keep the fan stuck in mode #1. Mode 1 (all the lowest manual modes, actually) are intended for customers who need the least amount of noise, with temperatures being secondary. The pump could also use a kick up if the fans don't adequately cover it.

Try to get the temperature of the CPU block. The ERM-2K3U includes three sensors that can be used for this. The value should be about 5-10°C below the CPU core under load.

If the CPU core is much hotter, it indicates: 1) a water block contact or thermal paste problem, 2) a faulty CPU core reading (typically from uncalibrated 3rd-party software), or 3) a faulty/loose CPU block temp sensor.

If the CPU block and core are within 5-10°C and the system is still too hot, it indicates: 1) insufficient radiator airflow, 2) insufficient coolant/pump speeds or line blockage, or 3) too small of heat exchanger for the overall head load.

User avatar
MUGEN
Tactical Bucket Splasher
Tactical Bucket Splasher
Posts: 212
Joined: Thu Jul 02, 2009 9:11 pm
Location: US

Re: 345ac High Temps and mounting questions

Postby MUGEN » Wed Oct 07, 2009 4:03 pm

I'm now running at 3.6ghz with a 1.264 Vcore. I have a sensor hooked up and usually set it to auto and i've only seen it bump the fan speed up to 2-3 once and that was when the room was around 85f and I was doing some furmark stability tests or something.

Fan speed 10 pump 10

Room temp 74F
Idle
CPU 37C (avg of the 4 cores)
Sensor 1 28c

prime 95/realtemp warm up test(about 5-10 mins)
CPU 63c(avg of the 4 cores)
Sensor 1 29c

Looks like the sensor must be bad because it's on that block perfect. I'll go connect another and touch it to the same spot where the bad sensor is to see if that's it.


My questions are
1. Does it matter which way the CPU block is mounted, outlet port at the top/side/bottom?
2. I saw one of you Koolance guys posted that the max distance of the thumb screw to the motherboard should be 15-16mm for the 350ac block, is this the same for the 345ac?
ERM-2K3U + HX-1320
i7 965@ 3.6ghz CPU-360
MB-ASP6WR
GTX 295@ 670/1476/1188 VID-NX295
6 sets of No-spill QDCs
EK-DCP 4.0 Pump

MCP
Water Shotty Skirmisher
Water Shotty Skirmisher
Posts: 46
Joined: Tue Oct 21, 2008 1:30 pm

Re: 345ac High Temps and mounting questions

Postby MCP » Thu Oct 08, 2009 9:28 am

Those numbers could indicate poor heat transfer between the CPU and waterblock. Have you verified that the block is mounted correctly and is making good contact with the CPU?

User avatar
[Koolance] Tim
Moderator
Moderator
Posts: 451
Joined: Tue Sep 16, 2008 9:34 am
Location: WA, USA

Re: 345ac High Temps and mounting questions

Postby [Koolance] Tim » Thu Oct 08, 2009 10:08 am

MUGEN wrote:My questions are
1. Does it matter which way the CPU block is mounted, outlet port at the top/side/bottom?
2. I saw one of you Koolance guys posted that the max distance of the thumb screw to the motherboard should be 15-16mm for the 350ac block, is this the same for the 345ac?


1: In a closed loop, not at all.
2: The CPU-345 is a couple mm shorter than the CPU-350, so you might get a smaller distance.

User avatar
MUGEN
Tactical Bucket Splasher
Tactical Bucket Splasher
Posts: 212
Joined: Thu Jul 02, 2009 9:11 pm
Location: US

Re: 345ac High Temps and mounting questions

Postby MUGEN » Thu Oct 08, 2009 10:48 pm

Just got done taking the cpu block off.
Things I found.
-Too much paste and it was pilled up on the sides but the paste was compressed well in the center and outward
-Small particles in the coolant if you hold it up to the light, not air bubbles.
-the sensor was still on their tight and I tested all 3 sensors by putting them between my fingers and found the sensor that was installed on the block was 3-4F cooler than the other two.
-from the plate to the top of the thumb screws I found 15-16mm on each one and I couldn't turn them anymore by hand. The 350AC is supposed to be 15-16mm and from the diagrams the 345AC is actually 2.1mm shorter than the 350AC so does that mean I should have my thumb screws at 13-14mm?
-While removing a qdc on my ERM-2K3U with a wrench I snapped off the LED light switch when the qdc finally let loose :oops:. I never used it anyway and good thing it's stuck in the off position.


Sensor 1 was put on just like it shows in the picture in the manual and it's on just like the old one.

Fan speed 10 pump 3(3 is the least audible setting I've noticed)

Room temp 76F
Idle
GPU1 38c
GPU2 36c
CPU 38C (avg of the 4 cores)
Sensor 1 28-29C

under load
furmark stability test 3 minute run
GPU1 around 48c
GPU2 around 46c

prime 95/realtemp warm up test(about 5-10 mins)
CPU 63c(avg of the 4 cores)
Sensor 1 31C

So their you have it, absolutely no change after about 2 hours of work. The 345AC and the ERM-2K3U aren't living up to what I expected.

So where do I go from here?
ERM-2K3U + HX-1320
i7 965@ 3.6ghz CPU-360
MB-ASP6WR
GTX 295@ 670/1476/1188 VID-NX295
6 sets of No-spill QDCs
EK-DCP 4.0 Pump

[Koolance] Dean

Re: 345ac High Temps and mounting questions

Postby [Koolance] Dean » Tue Oct 13, 2009 10:21 am

A 30c difference between the block and the core is very significant. What thermal paste are you using? Also 15-16mm is the distance you should have from the top of the thumbnut to the top of the bracket, this shouldn't change for the 345 as its a function of the spring's constant and the desired force.

User avatar
MUGEN
Tactical Bucket Splasher
Tactical Bucket Splasher
Posts: 212
Joined: Thu Jul 02, 2009 9:11 pm
Location: US

Re: 345ac High Temps and mounting questions

Postby MUGEN » Wed Oct 14, 2009 8:28 am

[Koolance] Dean wrote:A 30c difference between the block and the core is very significant. What thermal paste are you using? Also 15-16mm is the distance you should have from the top of the thumbnut to the top of the bracket, this shouldn't change for the 345 as its a function of the spring's constant and the desired force.
I'm using a pea size dot of OCZ freeze. I'm going to go measure the thumbscrew distances with a digital micrometer today and put them to 15mm.
ERM-2K3U + HX-1320
i7 965@ 3.6ghz CPU-360
MB-ASP6WR
GTX 295@ 670/1476/1188 VID-NX295
6 sets of No-spill QDCs
EK-DCP 4.0 Pump

MCP
Water Shotty Skirmisher
Water Shotty Skirmisher
Posts: 46
Joined: Tue Oct 21, 2008 1:30 pm

Re: 345ac High Temps and mounting questions

Postby MCP » Wed Oct 14, 2009 10:00 am

It's possible that you're using too much thermal compound. The OCZ Freeze website suggests using an amount roughly equal to a grain of rice. Using an amount the size of a pea would be 2-4 times as much.

One other thing to check: I doubt this is the case, but did you remove the protective film from the waterblock?


Return to “Water Blocks - CPU”

Who is online

Users browsing this forum: No registered users and 1 guest