7700k OC temps with medium tier setup

[michal8998]

Hi guys.
I can't get stable (without over voltage ) my 7700k when using offset no mater how I will set LLC.
I found that my CPU is 100% stable and coldest when I have static voltage of 1.29v with LLC High (auto,standard,high,extreme) then temps stay at no more than 80*C in IBT max and PC is inaudible at all, in games I have max of 75*C.
When I have offsets then PC goes off at idle or low load, only one way to avoid this is to give more voltage than it need at load , high LLC with offset makes temps too high at 90*C rank so I don't want to keep it like that.
CPU is delided second time in last 3 years, ambient temps is around 24C now, idle temps are at 33*C.
Max of this CPU is 5.3GHZ 1.4V but temps are about 90*C.

I am using it as daily PC for gaming and some amateur music production and question is are this temps normal for this setup? With previous Cooler Master Evo 212 I had same temps but with waaaaaay louder fans, now I really can't hear PC at all in any way with any task.
Second question is that possible for more massive heat sink to let go down with voltages on CPU? I think that what happen in my case, with Evo 212 as I remember I had to keep higher voltage around of 1.38v to keep it stable but temps were same...any help please? Cooler is mounted properly and with Mx2 paste.
Thank you

Spec:
Dark Base 900 3 BQ! fans
Dark Rock 4 Pro 2 BQ! fans
7700k delided with grizzly conductonaut + Mx2 at top of IHS
Fractal Design Edison M 650w Gold (Seasonic G)
Gigabyte z270 Gaming K3
Corsair LPX DDR 2133 CL13 1.2v OC'ed to DDR 3000 CL 15 1.35v (rock stable)
MSI GTX 1080 Gaming X 2100/5500 with max of 75C* (rock stable)
NVME M.2 Corsair MP510 500GB for OS
2x Samsung Evo 860 500GB RAID 0 for data

LAST photo is from few months ago with Evo 212 on it..what is going on? Ambiant temps seems to be similar.

 

[EarthDog]

Temps seem normal, but I don't understand what this means...

is that possible for more massive heat sink to let go down with voltages on CPU?

If I understand correctly.....

The difference in temps between heatsinks won't allow much of a voltage change at all. For example, if you needed 1.3V to be stable on a potato, then you should still need 1.3V on a better potato.

Also, use the regular .exe of realtemp. The GT is for hex cores, hence why you are missing info.

 

[michal8998]

I am using GT from the time when I had evo 212 so I don't want to change it for now,
from my testings... Evo 212 is better than Dark Rock 4 Pro.. I am not sure what is going on

 

[EarthDog]

Your choice if you want to see missing columns there... (sorry, my OCD was kicking in, LOL). Otherwise, it will read the same, just without the two extra blank columns.

The DRP4 should be better, but only by a couple of C according to some reviews...

I'm not sure either, but it seems perhaps a mount, or thermal paste application, something is different if that is what you are seeing. That said, you mentioned similar temps but way quieter... which is the point of the cooler in the first place. Am I missing something? The explanation isn't clear to me (getting coffee now.....).

 

[michal8998]

I think I will go back to Evo... I really don't know what is going on, redelided just now again and same thing. Evo was louder but way better with cooling down cpu afer OC....no matter what ;/

edit: I just looked on temps today and at the day when I did tests on Evo 212 in june and then it was 12*C, today I have 26 in my city (...)

 

[EarthDog]

I think I will go back to Evo... I really don't know what is going on, redelided just now again and same thing. Evo was louder but way better with cooling down cpu afer OC....no matter what ;/

edit: I just looked on temps today and at the day when I did tests on Evo 212 in june and then it was 12*C, today I have 26 in my city (...)

redlided? Is your CPU delided or are you trying to say you remounted the heatsink again?

So, when you last tested it was 12C in your room and now its 26C?? 12C is cold as hell for inside a house.....It is only the temperature of the room that matters, not the outside.



EDIT: I can see you have used different applications to confirm voltage from the two pics... that could be one reason why you see it is different. You need to compare like apps. That said, the difference is negligible anyway...

The DRP4 should cool better. Regardless your temps are OK... not sure why you would go back to the louder solution. It doesn't make sense at all how you are working this out.......I don't see a problem here......so confused.........

 

[michal8998]

Same readings on both apps and still I have them both. I just did deliding third time with liquid metal to be sure about how it was applied, also new thermal paste on cpu ihs/ heatsink. Same results, I think ambient temps have a lot to do here ..

- - - Auto-Merged Double Post - - -

On Evo212 I had similar temps on 5.2GHz 1.4v as on DR4P 5GHz 1.3v. This is confusing mate :/

 

[EarthDog]

Same readings on both apps and still I have them both. I just did deliding third time with liquid metal to be sure about how it was applied, also new thermal paste on cpu ihs/ heatsink. Same results, I think ambient temps have a lot to do here ..

/

Ok my friend, let's be crystal clear because it feels like you are skipping over questions and necessary information and just trudging forward aimlessly...

1. You have delidded the CPU, twice, and doing it a third time? Correct? (I don't see why you would do this again)
2. 12C ambient room temp is ~54F... that is COLD for a room. Was that your ROOM temperature or OUTSIDE temperature? What matters is the ROOM temperature.
3. Voltage appears to be the same on each of the images. Look at Vcore in Hwmonitor and the voltage in CPUz.
4. The temperature difference is literally 4C. Your load temperature will change 1:1 with ROOM ambient temps. It's clear that one of the tests the idle was a lot lower than the other...be it from ambient or whatever other reason. If your room was actually that cold with the 212, then it makes complete sense the results you are seeing. That said, I wouldn't go back to the louder, inferior CPU heatsink.

Good Luck!

EDIT: Please stop deleting your posts... it makes things incredibly confusing!!!!!!!!

On Evo212 I had similar temps on 5.2GHz 1.4v as on DR4P 5GHz 1.3v. This is confusing mate :

It really isn't... I don't see that voltage difference in the screenshots provided if you are looking in the right places. I see the CPUz screen at 1.296, and the other screen also shows a peak of 1.296. Both images are at 5 Ghz as well. I don't think you are making much sense as, again, you are reading the wrong thing on the other piece of software.

 

[michal8998]

1: Yes, delided once more to be sure.
2: Yes I had 12C outisde and windows open at home all the time, I like when is cold so probably it was at home about 16C
3: Yes at this occasion it was the same but Evo 212 is on offset and DR4P not offset but manualy 1.29v because it can't handle voltage sipkes to 1.4v with offset and Evo212 can (...)
4: Very possible.

Readings are same on both apps

 

[EarthDog]

Well, if your room was 16C and now its 26C... the results make complete sense.

That finally cleared up......

There is no reason to swap coolers. Enjoy the rig!

 

[michal8998]

Well, if your room was 16C and now its 26C... the results make complete sense.

That finally cleared up......

There is no reason to swap coolers. Enjoy the rig!

yeah ambient temps are now way higher and with open case it is still better than Evo in colder day. I will wait for cold day to test with closed case but it seems to be not bad at all, also I will have to set up my own curve of case fans in bios because I forgot to say about that I have added also PSU cover... and that changec air flow in a PC ... heh
EDIT: With closed case and fans on 100% there is no more than 70*C on CPU... that was the case and ambient temps. Thank you

 

[EarthDog]

LOL... ok! Enjoy!

 

[Zerileous]

...
When I have offsets then PC goes off at idle or low load, only one way to avoid this is to give more voltage than it need at load , high LLC with offset makes temps too high at 90*C rank so I don't want to keep it like that.....

Many motherboard the High LLC setting actually increases voltage under load, while the stock LLC behavior is a decrease in voltage. This is due to the way VRM works when a load is applied to them. If you're working with an offset voltage that is also decreasing the voltage under low load or idle conditions, it makes sense that this could induce instability. When you're using a set voltage, yes you want LLC to limit or eliminate voltage droop or decrease under load (recognizing that the CPU will see higher voltage spikes because of it, OCing isn't without risk ). When you're using an offset voltage, you can almost look at it like higher LLC is lowering your idle voltage (because the CPU will adjust the VRM until the required amount of voltage is achieved under load). Hope this is making sense, I haven't had my coffee. Maybe let this guy explain it

yeah ambient temps are now way higher and with open case it is still better than Evo in colder day. I will wait for cold day to test with closed case but it seems to be not bad at all, also I will have to set up my own curve of case fans in bios because I forgot to say about that I have added also PSU cover... and that changec air flow in a PC ... heh

Even with fans maxed it should be a pretty quiet case. I wonder if it's possible to easily mod the case so that both fans can blow into the main chamber, instead of part of the bottom fan blowing into the PSU cover. This appears to be a similar case, and after removing the font cover (which I don't anticipate you appreciating as a solution), they found that an inverted setup provided the best CPU temps. Is that an option in your case? You could also try modding the PSU shroud with a diverter to direct all of the flow from the front fan into the main part of the case (I'm imagining cutting the top with a dremel from the fan back about 6 inches, then bending that down to make a little ramp). If you mess up I believe replacement shrouds are available.

I agree with EarthDog that the biggest explanation is ambient temps. When PC components are tested thermally, the results are usually reported as deltaT over ambient. In other words, the ambient temperature is subtracted from the result, as it has been shown to essentially have a 1:1 impact on component temperatures. In other words you would expect your CPU temp to be 10C higher with an ambient (*room) temperature of 26C than it was at 16C. Furthermore if your CPU temps drop by about 10C with the side panel off, then you know the best way to improve CPU temps is to improve case ventilation (this isn't unusual btw).

You already have your cables well managed, which is awesome. You already mentioned setting the fan curves for the case fans. TBH first thing I would do is max them out and see if it bothers you. The fans you have are quiet even at full RPM, and your case can absorb a lot of that sound. If it does bother you it will also help you decide at what point you you're willing to be that bothered for better temps / a higher OC. When setting the fan curve, I would avoid using CPU temperature. This will cause fan speed to fluctuate up and down frequently, which most people find more annoying than a constant sound. If your motherboard reports some sort of socket temperature that increases and decreases more gradually with CPU temp then you could use that. I also wouldn't necessarily use the "motherboard" or "case" temperatures reported, as these will change substantially with ambient, and also may not be near the CPU, so they could take some time to respond to an increased thermal load in the case, thus not benefiting you a whole lot. The advantage would be, that those sensors will also account for GPU load. It's just something that will take time and experimentation to really perfect. You might also find a set RPM that is quiet enough for your needs but also improves airflow enough.

 

[michal8998]

I have now case closed, three case fans at 100% and two for cpu at auto. Temps are 75C° at most in IBT Max with higher ambient temps than usual. I expected more from that cooler vs Evo212 but it is fine and Quiet

 

[EarthDog]

Your expectations are off is the point.

If you compared with like environments, you would see the DRP4 is better and quieter, but since you had around a 10C difference of room temps, the 212 showed better performance. Compare like with like and THEN come to a conclusion.

 

[pgdeaner]

Also, you could probably shave a couple of degrees by replacing that Mx2 with a better paste like TG Kryonaught.