Jump to content

marteee

Members
  • Posts

    11
  • Joined

  • Last visited

Reputation

0 Neutral
  1. Pressure wise I hand tightened in the usual pattern (Top left, bottom right, bottom left, top right loose, then tightening gradually in the same pattern), then tightened off with a screwdriver today to be sure (maybe a quarter turn at the most), I think any more pressure would be excessive from my understanding? AIO is brand new out of the box, so the TIM on the cold plate is fresh, as for the TIM under the CPU's IHS - obviously several years old now (and it was a 24/7 machine for the first two years of its life!) - so could well be dust! Your conclusion is pretty much my own (along with the VCore at AUTO over-feeding it!), but I was hoping to be wrong! I'm kindof resigned to it just being "the way it is" - it isnt throttling, maintains the all core boost 100% of the time and as I mentioned the 100% all core load is totally synthetic and not typical of anything my daughter will do with it (Minecraft, Fortnite, Roblox - the usual kid stuff!) Im still all ears for anything else to try though if anyone has any thoughts!
  2. Pressure wise I hand tightened in the usual pattern (Top left, bottom right, bottom left, top right loose, then tightening gradually in the same pattern), then tightened off with a screwdriver today to be sure (maybe a quarter turn at the most), I think any more pressure would be excessive from my understanding? AIO is brand new out of the box, so the TIM on the cold plate is fresh, as for the TIM under the IHS - obviously several years old now (and it was a 24/7 machine for the first two years of its life!). Your conclusion is pretty much my own (along with the VCore at AUTO over-feeding it!), but I was hoping to be wrong!
  3. "Always a bit counterintuitive since, the AIO actively cools the water, but only indirectly the CPU" That was really useful by the way, should have been obvious but yeah makes total sense! I guess this also answers why the temperature ramps straight up to the same point regardless of fan state. Thanks for the tip!
  4. Thinking more about mounting pressure, would I be able to get the idle temps down into mid 30s if it wasnt mounted properly? Wouldn't the idle temp but much higher?
  5. Isn't it strange though that seemingly no matter what the fans are doing (anywhere from 400 to 2400), the upper temps are basically always the same with the all core 100% load? (90) - its like the cooler just cannot get the CPU temp lower, regardless of what it does (coolant temps do rise from an idle of about 29 up to low 30s after a long soak) Is that again pointing towards either too much voltage or bad mounting (the latter I just cant see how is possible though)
  6. Thanks LeDoyen. It isn't overclocked at all (board doesn't support overclocking anyway with that chipset I don't think), its all stock. It is however all on auto, perhaps its being fed too much voltage? What would a good manual dial-in for voltage be? I tightened everything down again as I did think about the bad mount but its stock thermal interface material as it came (1st install from new) so I would've thought pretty fool proof? Wouldn't have a clue about delidding it (Ive read a little around it but would never go there!) and no interest in overclocking it, just would like to see it running a little cooler (albeit with an artificial load). Ran the Unigine Heaven test a few times and gets to around the 80 mark
  7. Last thing Ive tried is unlatching the speed from coolant temp, which has what I would consider the normal effect on noise: Fans immediately ramp when hit with load and drop back to silent when idle, using this curve, temps are actually (weirdly) a tad lower at around 87% at 100% all core
  8. A fixed fan speed of 1100RPM yields the same performance too, around 91-92 within about 30 seconds and stays there
  9. Without adding too much rant, the Extreme profile is the same, activating it the fans immediately ramp up to 2400RPM with the system IDLE - here is a screenshot without any load, about 5% utilisation, CPU temp at around 35: Whats also interesting is at this point if I hit the all core 100% load again, the temps are almost identical to the Quiet profile, with only about 5-6 degress difference :
  10. Seeing temps of ~90 under all core 100% load, doesnt seem to matter which fan profile (Quiet/Balanced/Extreme) I set, the temps settle in at around 90 degrees celcius. The all core boost of ~4.4Ghz never drops, so it isnt throttling but is it not uncomfortably hot? Selecting the Balanced profile, the base speed of both fans immediately jumps to about 1800RPM, with the system idle (at around 35 degrees celcius), and again the speeds trickle down VERY, VERY, slowly to a base speed of about 1400RPM (which is still quite loud) but as you can see, this is about 8 minutes of slowing down!.... I then set both to Quiet, and it immediately drops, as I wouldve expected the curve to do?: Using this profile the fans seems to max out at about 1050 RPM (both) after running an all core 100% load test for about 15 minutes, averaging 92-93 - but it literally takes about 5 MINUTES to get from 400 to 1050 - why?!. On stopping the load, instant silence from case fans (220T via a DC curve as theyre 3pin), AIO ramps VERY, VERY slowly back down The below shows just how slow it is, thats about 5-6 minutes to get up to 1000RPM I think the problem appears to be everything is tied to coolant temp, which makes sense but my main concern is the peformance - why are these temps so high? Any advice appreciated (and Im hopefully being stupid!)
×
×
  • Create New...