Was I wrong in thinking a MCR320 should handle this?

[WoZZeR999]

My System setup is in my sig. In short, I have a e6750@3.7 with 1.5(real) vcore, a G92GTS at 810/1998 and my FSB of 432 on a 333 strap. I have 3 YL SH's on my rad. When I run Lightsmark 2007, even with my fans at full speed, I notice my water temp going up pretty fast.

My temp probe is in my res, so I don't know if it's just a lot of hotter water going into the res, or if my MCR320 just can't handle it. I plan on buying another inline temp probe (other was Tt, with real tine inlet and outlet, so I took it out of it's metal casing) and get the temp from the exit of the rad. Should a MCR320, with my computer at full load, be able to dissapiate the heat easily?

[BrokeDown]

I'd trust your BIOS temps rather than your res. temps first.

[Eddie3dfx]

Have you tried speedfan/sensorview? It can measure all your temps and create a log.
This way you can see what they are at idle and when you run programs like lightmark.
The water will change 3-5c from idle to load

[WoZZeR999]

I use speedfan, Rivatuner HW monitor, all the fun stuff. I use speedfan for basic temp monitoring, and the radiator fan control. I have another rheo control for the case fans.


I'd trust your BIOS temps rather than your res. temps first.

It's the actual water temp that I was concerned about. I don't know if there is a way to let something that is demanding on both the gpu and cpu like lights mark loop. I may end up letting it sit for a few hours with Orthos and ATITool scan run for a while and see if my water temps rise more than a few degrees. I was under the impressions that the water shouldn't really rise more than a degree or 2 if you have sufficient rad, but I could easily be wrong about that (only Wcing for about 6 months).

I have a pretty warm room, but orthos small FFT loads at about 20c above idle, so I'm sure that it's sucking heat away from the proc pretty good. GPU never raises more than 3 degrees above idle.

[NaeKuh]

It's the actual water temp that I was concerned about. I don't know if there is a way to let something that is demanding on both the gpu and cpu like lights mark loop. I may end up letting it sit for a few hours with Orthos and ATITool scan run for a while and see if my water temps rise more than a few degrees. I was under the impressions that the water shouldn't really rise more than a degree or 2 if you have sufficient rad, but I could easily be wrong about that (only Wcing for about 6 months).

I have a pretty warm room, but orthos small FFT loads at about 20c above idle, so I'm sure that it's sucking heat away from the proc pretty good. GPU never raises more than 3 degrees above idle.

theres a lot that plays in what your saying. We cant tell you exactly whats up without a aprox flow rate, with air temps in and outside radiator.

Also it can take up to 30min for your water to reach max delta. One watercooling loop is not the same as another.

But ive seen coolant temps go higher then 2-3C if you take your messurements from output and input on the radiator expecially on your heat sources.

[WoZZeR999]

I was planning on getting a new temp monitor for my rad output. That would probably give me a better idea if there is something wrong with that MCR320. With the parts in my sig, and using Martins flow rate estimator, I have a GPM of somewhere around 1.5, so it's not my flow rate limiting me. Ambient is rather warm, and it's not the actual temp # that I am worried about. It's the delta from air, and delta from idle, which should remain similar at any given ambient (within 5-10c).

I'll be able to test my system later with a higher load and test the deltas.

Thanks for you help.

[eXa]

What excactly is ur temps? (water\air++)

[WoZZeR999]

Not sure of ambient, probably somewhere around 23-25c. Water at idle can be anywhere from 0-2 degrees from that. Cores are at 28, cpu @ 32, gpu @ 32 as well. Orthos load at 1 min 48c, cpu 50c. Ati-tool scan @ 1 min, 33-34.

[eXa]

water under load then? (30min+)

[ElEctric_EyE]

I start to see :banana::banana::banana::banana: happening when my WC loop temp approaches 10*C above ambient. and yes it rises very quickly (which means good flow). About .2*C/sec. GPU in loop is not good IMO.

[WoZZeR999]

at 30+ mins orthos alone my temp's don't really increase at all. I had 51 @ load, but it's more with I have gfx running and cpu at the same time that I see it increase. I will test how much the water increases at long full runs with orthos + atitool and have my temps max temps tomorrow morning after I let it run all night.

[NaeKuh]

I start to see :banana::banana::banana::banana: happening when my WC loop temp approaches 10*C above ambient. and yes it rises very quickly (which means good flow). About .2*C/sec. GPU in loop is not good IMO.

It can also mean overloading radiator :yepp:

[septim]

too high an cpu overclock? together with gpu overclock...

maybe its time to think of going dual loops?...

[WoZZeR999]

I'm going to run a lightsmark loop all night, we'll see tomorrow morning if it evens out and ends up working fine. I have a P182 case, so getting 2 loops on the case is going to be a bit tricky.

[Cronos]

MCR320, PA120.3, GTX360 -they all have very similar performance in actually interesting fan rpm range, difference is practically negligible.

In this case, 10C increase in water temps is rather high and simply indicates low air flow through the radiator.
Here is simple formula to estimate delta air in / out:
1.6* (Power in W)/ (total air flow in cfm).

From here, the airflow can be estimated, if we know approximate difference in power consumption between idle/load and corresponding change in water temp (which will follow closely the change in air in/out delta)

[WoZZeR999]

I let lightsmark loop all night, and it seems like my rad wasn't overloaded at all. In fact, it didn't seem to even overload my rad when my YL SH's were set to 35% with automatic speed change disabled... (opps). My room did get cold at night, water temp was at 20c when I woke up, 17c after 20 mins of idle afterwards.

[Aerou]

cmon, its a dual-core and a G92! any 360 rad can handle that easily, hell, even a 240 with higher rpm
when you do full load (cpu-gpu), does "hot" air come out of the radiator? if not, then there is a problem with flow
20min load is enough
do the work:
monitor temps under idle and load, that way you'll see if the loop is performing as it should and the problem is the probe, or if it is otherwise

[WoZZeR999]

cmon, its a dual-core and a G92! any 360 rad can handle that easily, hell, even a 240 with higher rpm
when you do full load (cpu-gpu), does "hot" air come out of the radiator? if not, then there is a problem with flow
20min load is enough
do the work:
monitor temps under idle and load, that way you'll see if the loop is performing as it should and the problem is the probe, or if it is otherwise

I figured out it can handle it fine, I had a temp probe in the res, so it was mixing warm and cool water, and I had never let the water come to a high point while watching it.