Abnormal spikes in temp

Okay so I was running Prime95 on my new cheap custom loop, and on a 8150 @ 3.9Ghz I hit a max of 47.5C but the weird part is as soon as I stopped prime, my temps just spiked up to 55C and fluctuated between 36 and 55 (and slowing decreasing as the stored heat goes) this is all in idle. my ambient temps are 26C

I was wondering what would cause that AND anything to control my fan speeds (all are connected to mobo) doesn't work, Speedfan, AMD Overdrive. Any idea on what I can use to become their masters? I was going to buy a fan controller, but then I though I'm not spending $60 on something I didn't really need.

Umm could you tell us what's in your loop? Fan speed won't cause instant spikes, temps will change slowly.

Tell us your loop please................

And $60? Wow........

Good quality cheap high wattage controller. It's gotta be somewhere down under too.

http://www.jab-tech.com/Sunbeam-5-1-4-Rheobus-Kit-BLACK-pr-2530.html

I dont care whats in his loop.. spikes like that have nothing to do with the loop. Im sure its just an anomoly of exiting Prime or something or a faulty sensor even.

Ohh good point.

Wonder what he's using to monitor temp? Since it's an AMD Coretemp should be used?

If it's some mobo software etc, we know how they can be wonky.

I was also thinkig he's got pump control on the mobo?

+1 To Coretemp. Its weird behavior. I havent seen that happen when exiting P95 before. But even if it is, its just an instant spike at least, I cant imagine their being a shortcoming in the loop to warrant that spike... outside of the pump block drying out for a second ro two but thats just not possible really (air bubble?)

Did you actually stop prime95? It doesn't die when you alt+f4 it if memory serves right.

Well I'm using Coretemp and at the time I also had AMD Overdrive open and they both showed the same temps and spike. Possible airbubble, but if so then moving my case around isn't doing anything to get it out.
I actually stopped with though the menu and the core usage did drop to sub 10%

I wanted the Lamptron FC5 V2 Fan Controller.

sounds like a software glitch, core temp on my system will leap and hop about a little and aod's temp readings don't get accurate on mine till core approaches 50c backed up with my ir temp probe. they both show my idle temps below 20c

still really weird.

CPUID Hardware Monitor, seems to give a completely different temp and something that's believable at idle (~10%) It's hanging around DeltaT 1C and at load it rises 13C.

the dT we use is the difference between ambient temp and water temp, is this what you meant? Or did you mean the chip-air delta?

how are you powering your water pump?

please make sure the it is plugged into the correct cpu fan port if it is powered off the motherboard.

Pump uses molex and it's the difference between chip and ambient.

but it says the min was 6C last night, but the minimum for the ambient was 15. HWmonitor's wrong

There is many ways to measure Dt. It's a difference between two temps. In watercooling, Dt is what Mor means.

I think we need to work on whats really hapenning here.

Again, WHAT'S IN YOUR WC LOOP?

XSPC Rasa 750 RS240
and just the CPU in the loop

Ok, here's the deal: you're assigning WAY too much importance to idle temps. There is a KNOWN issue with AMD temp sensors that they are STUPID inaccurate and that the absolute numbers are not to be trusted at ALL, but merely taken as relative numbers. Don't think of it as degrees Celsius, think of it as degrees AMD (it scales closely to celsius, but not identically). Also, the temp sensors are not designed for you to monitor your temps, their primary purpose is to prevent the chip from blowing itself up (an issue AMD actually had an issue with, see video at bottom of post), so their resolution is much higher as they approach TJ_max (and, inversely, lower the further you are from it).

Also, dT = water to air (when you're on this board). I'm aware that technically nothing is specified as to what it is, but here, if you say dT and don't specify anything, EVERYONE is going to assume water-to-air...so if you're talking about chip-to-air dT, then say chip-to-air dT, not just dT. It will prevent confusion and allow you to get quicker, more accurate help.

Okay Thank you