Sanity check for new Hydro X Install

Hey guys, I got my Hydro X equipment installed and running. I just wanted to run things by the community to make sure everything is normal. Here are my specs:

 

CPU: Ryzen 3700x

GPU: RTX 2080 Super

Rads: 360mm x2

 

Here are some observations I've seen while gaming:

 

1. GPU temps gets up to 52 at an absolute maximum, typically hang in the mid 40s

2. CPU temps get up to about 70 but usually hang around the mid 60s

3. Fluid temps get up to about 38

 

Idle temps are obviously lower. GPU hangs at around 32c and CPU temps hang around 45c.

 

I still need to tweak the fan curves and such to optimize, but just wanted to make sure these temps seem within the range of normal. The CPU seems hot but from what I'm reading, Ryzen likes to run hot since it wants to boost until it's thermally throttled. Not sure if that's true so wanted to check with you guys.

 

The other thing I'm wondering is on the D5 pump speed. Using the flow indicator as a metric, I've noticed that it will not spin reliably when RPMs dip below 2000. It starts stuttering and will stop altogether some points. Is this consistent with what you guys see? I've just been keeping it steady at 2500 RPM.

 

Also, I've noticed some gunk has accumulated on the fins in the GPU block. I cleaned out the rads pretty thoroughly. I think it might be bits of acrylic. I did make it point to wash the tubes after cutting and deburring but it looks like some bits escaped the wash and have gotten stuck in there (see attachment). Do I need to worry about this? I know I'll eventually get around to breaking everything down to clean the system as part of annual maintenance, but I'm wondering if I should do it sooner than later because of this? I really don't want to have to crack open the GPU block. Has anyone done that and recovered from the experience?

 

Thanks in advance for your thoughts!

Edited April 29, 2020 by nerdballer
forgot to add attachment

1 & 2 - CPU and GPU temps are both on target for your max coolant temp. There is no problem with blocks or contact or those would be higher.

 

3 - Coolant temp 38C needs a baseline. If you start off at 18C and go +20C, then we need to look at some things. If you idle around 27C coolant and got +11C max on some torture load, then you are in good shape.

 

The one thing I can't tell from the 2nd picture is which direction your front fans are blowing. Depending on the answer, there might be one thing you can do to increase efficiency.

Thanks for your response! Yeah my fluid baseline is about 30C. I just ran the Aida64 CPU stress test and the Heaven benchmark at the same time for about an hour. Here were the maximum observations.

 

Ambient Temperature in the room: 24 C (warm, I know!)

GPU Temp: 54 C

CPU Temp: 72 C

Fluid Temp: 41 C

 

The fans ramped up based on a curve so my front fans were running at 1800 RPM while my top fans were running at 1500 RPM.

 

The front fans are in a push configuration, in-taking cool air from the front into the case. The top fans are exhausting through a push configuration.

 

So it sounds like these numbers are at least acceptable?

 

Thanks again!

For a 30C basis, +8C is pretty normal. In my dual 360 I will be at around +10C over the base when running 500-550W with fans 1100-1300 rpm.

 

However, if you are looking to pull back another degree or two, the one are might be the front fans. In this configuration, you are dumping the waste heat from the front radiator into the top radiator. It does have an effect on the top radiators efficiency. You might see better results by flipping the front fans to exhaust so all waste heat is dumped directly out of the case. Technically, you wouldn't even need the rear fan, but you would turn it around to intake if you wanted a little more direct cooling on the general CPU socket area.

That's an interesting idea. I would think that would create a lot of negative pressure though, pulling unfiltered air in and causing dust build-up. I guess, like everything else, it's a tradeoff. It's easy enough to test with though. Maybe i'll switch them around this weekend and run the same tests just to see what happens.

Yes, it would be all an all exhaust, passive intake build. People have taken this only semi-serious notion of positive vs negative pressure and made it one of the core build factors. It's really rather regrettable. If a particle of dust comes near the front intake of your case, it is coming into the case. If it hits a charged surface while passing through, it will stay there. Your fan balance has nothing to do with that.

 

People tried to build this up in the days of heavily vented cases as a way to keep dust from coming in all those open slots in back. That is true and it also may have had some benefit for an air cooled GPU where you want the heat going out the back in one way or another and away from everything else. However, we are now in the era of the glass case and the same logic may not apply. Some glass cases are open everywhere with panels "hanging" on the outside of the case. It may help. Others are sealed and it makes no difference. But once you water cool the GPU and take that piece of of play, you are really down to setting up your build based on whether you want to dust every 2 months or every 3. Seems silly.

 

I ran a O11 Dynamic with top/side radiator exhaust, no active intake. All the air came in from the back and bottom. No dust filters. It was the cleanest, most dust free build I have ever had, but it is mostly because of the materials involved and not the airflow patterns. I have a 740 Air that is heavily biased toward positive pressure, fully covered with custom dust filters for every piece of mesh, and it is always dusty. Black plastic. It never had a chance.

 

If you have a particularly difficult environment (desert dust bowl, cats, etc.), then perhaps you do employ every possible method of prevention. But most people should set the case up in the way the makes the most sense for the hardware, then actually dust every 3 months. 4 times a year is not really so bad and that is a more aggressive schedule.