Jump to content
Corsair Community

H100i V2


BuzzScream

Recommended Posts

Hey guys,

 

Just installed my H100i v2 into my new system, see specs below.

On idle my CPU is sitting at 50 c, which is crap. Have installed correctly and turned the Qfan controller in my Bios which I've heard causes issues.

Loading into windows temps stay around 50-55 but I know that as soon as I game those will jump 20-30 c.

 

Ive loaded the corsair link software and can see that fans (1400ish rpm)and pump (2800-2950rpm) are running. I had the same cpu in my old build and temps were 45 at load with H100i!

 

What am i missing or is the quality of these coolers turning to crap? :(

Link to comment
Share on other sites

Need to know H100i v2 Temp (coolant temp) from the main page in Link. Is this value steady? Or is it slowly creeping up the longer you run?

 

Make sure you have not reduced the voltage to the CPU fan header (or wherever it is connected). However, since your pump is at full speed, that is not likely to be the issue.

Link to comment
Share on other sites

Oh sorry I should have put that info in. The coolant is running at 35-40c but the cpu is still a good 10-15 c higher. Admittedly id love for someone who has the same Mobo as me to tell me the settings they have in their bios! But that may not be a possibility.
Link to comment
Share on other sites

Just a quick update on temps, using corsair link i can force the performance mode but then my system is super loud, however with these settings i can see that pump and fan run around 2800 rpm, the coolant temp is 29 c while my processor is around 45-53 c at idle. This is with the computer on for about 1 hour.
Link to comment
Share on other sites

Based on that information, I don't think you are looking a cooler issue. It seems more likely this is a BIOS, OS, or processor setting. When the cooler has problems, the coolant temp will start off cold and then continually rise every minute you are up and running. You might see coolant temps in the 50-60C+ range. If your CPU core temps are +15-20C over the coolant temp, then the processor is doing something that entire time.

 

Just so I can get a perspective on the expected temperatures, are you running the Performance (100% up-time) or Balanced Windows power plan? Do you have C-states or EIST enabled to any degree? Neither one would account for 15-20C, but it does help set a baseline.

 

What voltage and frequency is the 5820K? Stock?

 

Approximate room temperature?

 

 

Before we get too deep into this, open up your task manager and see how much of your CPU power is being used. Any other monitoring program that shows load or core activity would also work.

Link to comment
Share on other sites

ok then, I wont lie, my BIOS competency levels are below average haha. I have been running the computer at idle for a while now just to see what happens, temps are just hovering mid 40s then peaked at 63c! It just feels way to high for idle. In terms of the coolant can confirm it is hovering really low 28-30c was the max. I have been using HWmonitor to check these temps fyi. So maybe it is the BIOS making the processor do more.

Oh i didnt even know to check those power plans in windows unless it was a laptop. I have set corsair link to run in performance mode so fan and pump are running max. I dont know what exactly C-states or EIST enabled is? :(

Yes processor is running at stock, the only overclocking I've ever done is change the multiplier haha. Room would be maybe 25c at most if not cooler the past couple of days/ now.

My loads seem to jump all around the place while it was idling. No real processors running in the background other than HWmonitor.

Thanks for helping man :)

Link to comment
Share on other sites

That's OK, I know yours and have been running on 5820K/5930K for a few years, so lots of data about expected temperatures.

 

In a room in the low 20's, your 5820K should top out around 50C at default settings on any normal stress test (Prime AVX excluded). Obviously you are near that already and that is why I inquired about background programs eating resources. Windows in particular can be very pesky and just last night it was trying to do something while I was playing a game that needed 4 of my 6 cores at full tilt. It caused similar behavior to what your are describing and I would not have realized what was happening if not for my AIDA64 frequency graph showing the selected cores maxed out. I am still wondering if Windows or something else is trying to do work and causing this for you.

 

EIST (speedstep) and C-states are enabled on your board right now. That is fine and you should leave it. They will be in Auto state, with C1 and C6 enabled. Your core frequency should drop to 1200 MHz or so when the cores are not active and your idle core temps should be in the mid 20's. Ignore "CPU temp" and "CPU package Temp" readings. They are misleading and behave very strangely on HW-E. My package temp sits at 38C while every core temp is in the low 20's. CPU temp will run +10C over the cores at idle, then lag 10C under the cores when fully stressed. Focus on individual core temperatures.

 

Heat is being transferred into the cooling stream (the coolant temp goes up). If you don't get the mount right, you get very high CPU temps, but low coolant temps with little change. The 2011 socket is just a screw down, so I don't think this is the issue.

 

That leaves the TIM application. I am assuming you used the included TIM that was already on the cooler. Good. That is what you should do the first time. It is very unlikely there is a problem there, but when there is the CPU core temps will bounce around a lot. It won't look like loading activity, but often random or even nonsensical data (17-81-65-45-88-31C etc).

 

If you are sure the OS isn't trying to reorganize everything behind you back, it might be worth taking the pump head off to see what has happened. The TIM should have spread out evenly across the CPU without spillage over the sides. If you see a large block of unspread TIM, you didn't get it screwed down all the way. Be aware, normally you would want to clean and reapply TIM any time you take the block off. However, if you want to take a quick peak and remount with the stock thermal paste, I think it will be fine. If it gets worse, then we learned something as well.

Link to comment
Share on other sites

Yeah around that 50c is where im thinking it should be staying as well. Just played some rocket league on the computer for an hour and a bit and the max reading was 75c which look, isnt exactly great but at least isnt pushing 85, however with this cooler i expect it to run like its on water not on a stock cooler haha.

 

Ill get some thermal paste and reseat it. Im not sure if this helps but i dont have the CPU header plugged into the CPU WITH PUMP header cause it keeps on throwing out a weird error. But when I plug into the normal CPU header there isnt an issue.

Link to comment
Share on other sites

Hey man sorry, havnt been able to get to my computer in a couple of days. Yeah its always SVC **** running either my network or cpu usage. though its really just spikes to 30 something % so nothing to crazy. No difference when I restart the computer. Tomorrow ill finally be able to re seat the thermal paste and see if that helps.
Link to comment
Share on other sites

I don't know what it is either, but I also had the same problem earlier this week after the Tuesday Win 10 update. I was talking with another HW-E owner and she had the same issue at the same juncture. The CPU model is probably just coincidence and what the OS is trying to do may be lawful. It's hard to tell what those SVC host processes are really doing. You can use the drop down menu to see more, but it isn't all that helpful. I thought my Windows update might have been stuck, so I uninstalled the last update, and did it again. I has been quiet since. I don't know if that is your issue, but playing the odds, I wonder if you found yourself in similar circumstances.
Link to comment
Share on other sites

So today I have reapplied with to much and the right amount. Temps have come down a little. Interestingly enough when I took the cooler off first only the bottom right had taken some of the paste on the CPU, the rest was pretty much just like it was on cooler block when I bought it.

I have screwed in the stand offs and with my particular mobo there is no need for the back plate you just screw the 4 screws in and put the cooler on top of the CPU. problem is now this is screwed right down.. I cant make a firmer contact with the CPU and the block. The thermal paste is spreading out but maybe there is still not full contact?

Also using Aida64 to run a stress test, temp readings of 38 c on idle which still isnt great, given my other computer is running a crappy coolermaster system, yet idle is 28 c. Not sure where to go from here..

Link to comment
Share on other sites

It seems pretty clear the cold plate was not making full contact. The standoffs you need for 2011/3 are the uneven ones, like this below.

 

http://i.imgur.com/aFJRTCFl.jpg

 

The short, stubby end goes into the board. Rest the pump head bracket on top and tighten the thumbscrews down. If you have used the proper posts, make sure they are inserted fully and DO NOT force them. It should easily go in and out.

Link to comment
Share on other sites

Yeah that's pretty much it, so annoying that I threw them away. I opened up a ticket to see if I could get some sent out but havnt heard back yet. :( I looked on my old computer since I ran the h100i in that. Top those same 4 stand offs out and went cool I can just use these. However when I put them in they were too short there was no thread exposed for me to use a thumb screw on. The mounting plates are different, so that would be the reason for it not fitting the v2 setup, but I just want to confirm they are a different size??
Link to comment
Share on other sites

I think the problem there is the old H100i was CoolIT and the current H100i GTX/v2 is Asetek. Different mount systems and brackets.

 

There is an Intel mounting kit, but the details are a little fuzzy on the webstore. It does not list any of the current generation of coolers. Without having the kit in my hand, I am reluctant to declare it is the right one for you.

Link to comment
Share on other sites

Hey dude thank you so much, I think you're exactly right there, just wasn't sure if you had the measurement of the old standoffs (H100i) vs the new ones (100i v2). I have put in a ticket to corsair but still have not heard anything.. :( I wish they would tell me so i can get these standoffs haha
Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...