TioDrakul

I don't know what to do anymore. I can't set the pump speed on my H115i to anything other than "silent" without causing an annoying noise in the pump (see the attached audio file, compressed in zip format) and it seems to me that even the silent setting still causes noise (just in a different pitch). I have tried moving the cooler in various ways to get possible air bubbles out and to no avail, as well as tried placing the pump in other positions on the CPU. Is my unit defective?audiosample.zip

I’m having a serious pump noise problem. It's like the impeller is scraping something or having a serious air bubble and/or cavitation. The noise reduces somewhat if the computer case is lying horizontally but the correct case position is vertical, and the noise only stops if I leave the pump speed in "silent". Given that I live in a tropical country and therefore I cannot be operating indefinitely in "silent" to avoid noise, what can I do about it? Or the unit would need to be replaced?

I also have an H115i and in the "extreme" configuration the pump makes a noise similar to what you described. The cooler works, but I are also worried if this would be a sign of trouble with the pump.

I just bought a K95 RGB platinum and I was wondering what I can do with the macro keys. Since I do not use key combinations in any of my applications or games, I would be able to assign other functions to them, such as open a specific application?

I am not liking this. While synthetic tests suggest that everything is working correctly, real-world tests such as copying a large file, creating a large file, or reading a large file show performance far below expectations. I know the application used in the test below is old but it gives me a good idea of the performance of the entire disk under normal use (even SSDs using the SATA interface), and at no time does the SSD get more than 2.6GB/s. Even more worrying is that performance drops to a maximum of 1.5GB/s in areas having written data (eg, used space), like the old Sandforce-driven SSDs.

It depends. In my case as I described my motherboard heatsink is pretty big so it provides sufficient cooling but may not be the case with your motherboard. My suggestion is to carefully remove the MP600's original heatsink, install it on your motherboard with your motherboard's heatsink and then carefully watch up with a monitoring application (HWinfo64 for example) its temperature while under operation. In my case the maximum I saw was 58ºC (under load and with a GPU above), what as far as I know is within acceptable limits.

I was in the exact same situation as you, an MP600 and a rubber block causing too much pressure for my taste. My motherboard have a pretty heavy heatsink covering the M.2 connection area, so I ended up carefully removing the original MP600 heatsink and am using the motherboard heatsink.

Maybe helps if you turn off pci-express energy management? (the default option is not off ("desligado" on my language)

Ok, I think I found a good compromise. Using a powered fan splitter, I plugged the splitter signal cable into the pump controller so the pump controller unknowingly is providing the pwm signal to all fans in the case and the power to them is now coming directly from the PSU in a separate cable (the splitter have two cables: One for control signal, another for power). So now the pump controller thinks it is only controlling one fan (the splitter) but in fact it is controlling all fans of the case, and now when the controller decides that it is necessary to increase the radiator fan speed all other fans follow and thus prevent heat from accumulating in the case. http://www.swiftech.org/images/products/detail/PWM-SPLITTER-SATAx600.jpg

Okay, new hardware up and running. I tested my MP600 unit and the result is below: At least in my case it seems to be working properly, but note that i turned off any kind of energy saving on the pci-e link just in case (custom power plan)

Hello, i installed the hardware and after some experimentation i finally understood why you said it would be better to monitor the coolant temperature... It's because I found out that Ryzen is wild, his temperature fluctuates like crazy! I even thought that the cooler was defective (how could it allow 10ºC increases in less than a second?), But after some research it seems to me that this behavior is normal for a Ryzen. iCue is working so far (I only enable it to set the pump and fan speed as you suggested), but I'm guessing I'll have to connect the other system fans to the same controller if I want them all to respond consistent with the cooler temperature (my splitter allows it, it distributes the PWM source control signal to all connected fans but the power comes directly from the PSU on a separate cable).

Hm... Ok, I was wondering this because with the CPU reaching 70ºC I was expecting the coolant temperature would rise to well above 30ºC, I was even worried that the cooler was not working properly (blockage maybe) or not seated correctly on the CPU (poor heat transfer). So, in short 70ºC under load for a Ryzen 3900X is normal, I understand correctly? As for the voltage I believe I got it right, it ranges from 1.3V ~ 1.47V under load to 0.8V ~ 0.9V at idle. (funny enough, when I am in the BIOS the voltage appears to be constant 1.47V and yet the temperature is below 34ºC, I have never seen it before)

Okay, now I have the new hardware working and because it is completely different from what I used to have before I would like, if possible, to have indications if the readings I am getting are as expected or if they indicate problems. Using a Ryzen 3900X with what I think is stock settings (2200MHz at idle, most cores at 4500MHz at full load and one or two at 4600MHz), I have three CPU temperature readings I think are the internal controller, CCX1 and CCX2 (in order): Idle: 40, 32, 32 Full load (some gaming): 72, 70, 68 These temperatures are much higher than I'm used to seeing on a desktop, are they normal? Everything seems to be working but I'm kind of worried about seeing 70ºC on something that isn't overclocked. As an experiment I have installed iCue and so far it is working (but I have taken some steps to ensure this), and its sensor indicates a temperature of 28ºC on the coolant after computer startup and 32ºC after a couple of hours playing. Is normal for the coolant temperature difference to be so small between idle and full load? Finally, when setting the pump speed to "extreme" I started to hear sometimes noise int the pump, as if I had a small object circling inside it. While if I put the pump in "quiet" the noise stops. Would this be normal or I really have some kind of object inside the loop?

Interesting. I have a Corsair H110 which I am going to retire it soon because it will not be compatible with the new machine, but as it still works I was wanting to avoid throwing it away (it is just a bit old, running for some years now). I've been wondering if I could make it last a few years more it by doing preventive maintenance and replacing the coolant and cleaning the unit like you did, did you have any trouble resealing the pump?

I am an old-school developer, I don't feel comfortable with a cabinet that looks like a carnival car as it seems to be the current fashion... My computer case isn't even on the desk or somewhere where you could see such light effects (it's too big to fit on the desk anyway). And as a developer I've learned to appreciate the wisdom of K.I.S.S., so I prefer the hardware to take care of itself rather than it relying on software that can crash when least expected and with unpredictable consequences.