Jump to content
Corsair Community

Is my h150i Pro XT 360mm working?


BlackDragonFun
 Share

Recommended Posts

I recently bought the Corsair h150i Pro XT 360mm AIO to cool a Ryzen 3700X running at stock. It's running inside a Lian Li PC-011 Dynamic XL with 6 intake fans (QL120), with the AIO being the exhaust.

 

Under load (OCCT, Linpack), it stays around 69C according to Ryzen Master. Cores seem to be running at 3.9GHz. Voltage around 1.27v.

 

On idle, it seems to be stuck at 47-49C. Idle voltage seems a bit high at 1.4v. So temps seem a bit high. It seems like iCue keeps some activity on the CPU when I think it is idling, as temperature is around 42-46C when it is closed.

 

AIO fans are at 1000RPM and pump is on balanced at 2326RPM.

 

I'm trying to get an idea of whether the cooler is performing acceptably? Not sure what my ambient temperature is, but I suspect around mid to high 20s.

Edited by BlackDragonFun
Link to comment
Share on other sites

It's working. The Linpack run would be unsustainable if were not. You can measure cooler efficiency and it's impact on CPU temp by looking at the "H150i Temp" (coolant temperature) in iCUE. The change from idle to load is a direct reflection of how much heat goes into the cooler vs how much is released. It is affected by changes in internal case temperature, just like everything else. I would expect a +6C or so rise in coolant temp after about 10 min of the OCCT Linpack.

 

Other Ryzen users have had similar experiences with the idle behavior and there was a lot of discussion at AMD launch. You can quit iCUE and see what effect it has on the idle voltage. Also, if you don't have other Corsair gear, you may not need to run it at all. The pump LED ring can be set with its "hardware lighting" that shows anytime CUE is not running. The fans on the XT's controller will follow their curves as long as the control variable is H150i XT Temp. That applies to the 3 presets as well. It is easy to set this one up to run without iCUE, only launching it to change lighting schemes.

Link to comment
Share on other sites

Awesome, didn't know you could run without iCue :) I'll give it a go. As i said, I do notice a drop in idle temps when it is not running. It goes down to almost 40c.

 

I ran the Linpack test for 10 minutes to see how it compares to what you described. My coolant temperature went from 30.5c to 32.5c. CPU temps according to Ryzen Master are between 58c and 68c.

 

It seems like quite a difference between the CPU and coolant temperature, so is the heat being transferred to the coolant effectively?

 

Also my sensor shows 'Emergency Shutdown' which I can't find any explanation as to what it means. I've not changed the default notificaiton, which apparently is set to trigger at 70c.

n4x3KE2.png

 

Edit: 20mins of Linpack, coolant is at 32.9c. CPU temps around the same range.

Edited by BlackDragonFun
Link to comment
Share on other sites

I ran without iCue for a bit, idle temps are around 42C. Ambient temp inside the case is about 36c.

 

I suppose idle temps are OK, but I'm still questioning the load temps. I see reports of people getting temps in high 50s or low 60s.

 

I ran OCCT Linpack for an hour, my CPU temp was between 67c to 72c. Coolant temp was around 36c.

Link to comment
Share on other sites

I may have to start adjusting my predicted coolant temp increase for the 7nm Ryzens. That really small rise is a reflection of everything following proper directives and only outputting something close to the 65W TDP on the box. It may be more like 90W, but that is still a really small number. When it comes to coolant rise in a perfect environment, it's literally as simple as watts in vs watts out. With the power safeties removed in the BIOS, it is likely you can draw more watts and would see more coolant rise, if you were curious. However, where you will probably see the most coolant rise in mixed use or heavier GPU loads (gaming/renders). Most people will have an internal case temp rise as large as their coolant temp rise from CPU heat. There really isn't much you can do about that, but that invisible baseline that affects everything is now easily visible as the coolant temp changes.

 

I can't peg the exact spot where a 3700X should be in terms of CPU temp under Linpack and as a non-AMD owner, I don't follow it as deeply. I do thing your temps are slightly better than what I see from other AMD users, but their tests may be different. One thing you have noticed is the difference between the coolant temperature and the actual CPU package and core temps under load. This is normal. The coolant temperature is the baseline or lowest possible CPU temp with 0 volts. While the coolant carries the heat away from the cold plate, it also transmits it back across. Meanwhile most of CPU heat comes from the voltage on the socket/pin side. That has to travel through the CPU to be conducted elsewhere. There is no way around that and it is the physical limitation that keeps us from running 8GHz clocks or whatever else. At some voltage, the CPU will become too hot before it can conduct the heat away. The extreme overclocking results you see are done by lowering the base temperature of the CPU to negative numbers, effectively making the CPU a baseline -80C to offset the massive CPU temp differentials created at those voltage levels. We are all voltage limited, even if you cooler is the size of a wall.

 

Most CPUs of the last 5 years (AMD or Intel) will have a coolant to CPU temp differential of +35-50C. This number is predominately voltage level based as well as CPU construction, both model type and lucky silicon. Two 3700X at identical settings on identical boards may not yield the same temperature. Most CPUs at original settings are going to be closer to +35C. The more voltage you add on, the higher the differential will become. It is a good number to know, because it sets your maximum coolant temperature. If my personal temp limit is 80C and I know my CPU has a +40C differential, then my maximum coolant temp I want to stay under is 40C ensuring I can't break 80C even in the worst circumstances. This can be helpful for setting your own fans curve and in the above scenario, you would set that 100% fan blast at 40C to "get your attention".

Edited by c-attack
Link to comment
Share on other sites

Thanks for the explanation :) Good to know that the number sound reasonable.

 

I did also re-apply the thermal paste and tightened things cooler down more than I did last time. I noticed my CPU temperature in the BIOS go down from 42c to 39c/40c. I was a little anxious of overtightening the first time around, as I did that with a h100i in the past and broke the screw off resulting in the baseplate being unremovable!

Link to comment
Share on other sites

 Share

×
×
  • Create New...