[SOLVED] CPU and cooler seem scratched, one core running hot

[Most Worrisome]

Sorry for the long ramble, I'm a bit spooked right now.

A few days ago I replaced my CPU with a Ryzen 3600 and hearing that the stock fan ran a bit hot and loud, I got a Dark Rock Slim for it. I used the included BeQuiet thermal paste, which was maybe a mistake. It was miserably hard to get it out of the tube, but it looked fine once out . Powering up the system, I noticed that 3 of the cores were going up a to 80 degrees celsius under load, which I thought was a little much, given the type of cooler I had. Plus I got a Kernel Power 41 reboot, which I'm still not sure what to make of.

So I got some MX-4 and decided to re-apply the cooler. But once I took it off and started cleaning off the thermal paste, I noticed that both the CPU and the cooler are scratched up. Most notably, there is an alarmingly deep, 2-4mm long scratch on both, with a coppery color. At first I thought only the CPU was scratched, and there was something sticking out from the cooler surface that did the scratching. But looking at it more closely, I think both might be scratches? I could have sworn the cooler surface was clean when I first installed it. There are also a few much smaller and lighter scratches.

I have a really hard time making out WTF is going on, even with a magnifier. Here is the best pictures I managed:

https://imgur.com/a/cWIN3sM

View: https://imgur.com/a/cWIN3sM


Anyway, in retrospect this was maybe really dumb, but I just applied the MX-4 (maybe not enough of it?) and put the cooler back on. I've powered on the PC, and it's not on fire, but one of the cores hit 92° on startup and is still idling around 75. The rest seem fine, idling between 31-36 (with occasional flashes of 41) but given how hot that one core is, I'm not doing any stress testing.

So, what did I Removedup and how badly? What should I do, take off the Dark Rock once again, put on some more MX-4 and hope it goes away? Should I break out the Ryzen's stock cooler?

 

[RodroX]

Im assuming you update it to the lastest BIOS your motherboard support and that you also have the lastest chipset drivers installed from AMD web page (https://www.amd.com/en/support) right ?

Also guessing you have "Ryzen Balanced Power Plan" selected with "5%" on the "Processor Power Management" ----> "Minimum Processor State" on the advanced tab of the energy settings.

(Some people have it like this https://i.redd.it/at54nbjii2n31.png, if tthats your case change the Minimum to 5%, accept the chnages and see if that fix it)

As of how did you scratched the heatsink and the CPU that is out of my knowladge. Im guessing the CPU was brand new and didn't have any scratches before you tried to mount the heatsink right?

 

[Most Worrisome]

Thanks for the power info, I'll look into it.

As for the CPU, that's correct, it was brand new and unscratched. How is it even possible for two flat pieces of metal pressing together to get scratched that deeply?

Or more importanly, what should I do here? That core that is running hotter than the rest has calmed down a few degrees, but that is still way too much for idle. Will I keep damaging my CPU if I re-apply the cooler again?

 

[RodroX]

Ryzen CPU have a top operating temp up to 95°C. So you will be fine.

I will first try the chipset and power plan thingy (keep in mind that the right power plan for 3rd gen ryzen was implemented on the lastest chipset), so check if you have the lastest drivers for your chipset, and after that try the power plan thingy.

As of idle I have the same CPU, and not even with the stock cooler it reached 75°C, even less 92°C on idle. So really look into the bios and chipset drivers update, and then make sure the 5% is set on the ryzen balanced power plan.

 

[Most Worrisome]

I did have the latest BIOS and the latest chipset drivers from ASUS (clean installed for the new HW), but now I've installed the newest chipset package from AMD's site on top of that and changed the minimum CPU power to 5% (was 99%).

This hasn't really changed temperatures, though. And of course that one core wasn't running at 75° idle before I took off and re-seated the cooler, either.

At this point my biggest concern is, should I just take off the cooler again and redo the thermal paste again (which will hopefully fill in the scratches)? Or am I going to damage my CPU every time I take off and replace that cooler?

 

[RodroX]

I did have the latest BIOS and the latest chipset drivers from ASUS (clean installed for the new HW), but now I've installed the newest chipset package from AMD's site on top of that and changed the minimum CPU power to 5% (was 99%).

This hasn't really changed temperatures, though. And of course that one core wasn't running at 75° idle before I took off and re-seated the cooler, either.

At this point my biggest concern is, should I just take off the cooler again and redo the thermal paste again (which will hopefully fill in the scratches)? Or am I going to damage my CPU every time I take off and replace that cooler?

The thermal paste you applied should still be pretty "liquid", since is just a few days since you put it onto the cpu.

If you wana be on the "safe side", just warm up your cpu a bit before you take out the heatsink (turn on the pc for like 10 mins) shut it down and then gently remove the be Quiest cooler. I would not add more thermal paste, instead I would clean it all off, and start over from zero.

And no, you should not damage your CPU just for changing the thermal paste one mroe time. Just be sure to not drop anything heavy on the CPU and the contact surface of the Cooler.

Cheers!

 

[Most Worrisome]

Okay, I just might be extremely stupid. Does the TMPIN4 sensor in HWMonitor actually have anything to do with the CPU? I just.. kind of assumed it did.

I went and re-applied the thermal paste and thankfully there were no new scratches visible. Yet when I started the PC back up, TMPIN4 was again sitting at over 80 degrees. But then I tried CoreTemp ("maybe is HWMonitor is just mistaken"), it tells me that all cores are idling between a much more reasonable 35-42 degrees.

So, uh, what exactly is TMPIN4? So far I've only found a lot of old info and a bunch of "check your mobo manual", but my manual doesn't seem to have anything like that.

 

[RodroX]

hwmonitor kinda sucks for Ryzen 3rd gen, better use hwinfo portable - 64 bits - Sensors Only option (https://www.hwinfo.com/download/), or AMD Ryzen Master.

I like hwinfo more cause it shows lost info and you don't need to install it.

 

[stonecarver]

what exactly is TMPIN4

Just some reading.

View: https://www.reddit.com/r/overclocking/comments/agp40x/what_does_tmpin4_represent_in_hwmonitor_and_why/

 

[Most Worrisome]

hwmonitor kinda sucks for Ryzen 3rd gen, better use hwinfo portable - 64 bits - Sensors Only option (https://www.hwinfo.com/download/), or AMD Ryzen Master.

I like hwinfo more cause it shows lost info and you don't need to install it.

I did not know that about HWMonitor, thank you! HWiNFO64 also shows two readings constantly at 71-77°, but they're clearly motherboard readings (helpfully labeled "Temp2" and "Temp5") not CPU. They also don't seem to go up much under load, which is a bit of a relief.

Just some reading.

View: https://www.reddit.com/r/overclocking/comments/agp40x/what_does_tmpin4_represent_in_hwmonitor_and_why/

Thanks, I hadn't seen that specific thread yet. There are a few different answers floating around about what TMPIN4 is. Some, like that thread, say its just a disconnected sensor giving random or maxed out readings. But my readings are clearly not maxed and behave pretty consistently for random.

Others suggest it might be a VRM reading. Which is... um, would that be an alright reading for a VRM sensor?

Anyway, a lot of the info online seems to predate Ryzen itself, let alone X570 boards. Maybe I should try asking ASUS.

 

[Karadjgne]

Contact beQuiet. In order to damage the heatsink base to that degree, it would have taken serious abuse on your part to gouge it hard enough to leave an external nodule. Which leaves a factory quality control issue/manufacture assembly damage as the most likely subject. This damage has impacted the cpu IHS surface, and imho has rendered it Destroyed. It would need to be resurfaced to repair that damage and any further installs of the damaged cooler will just re-damage the IHS.

If you are absolutely positive there was no prior damage to base or IHS, then the only place that could have occurred was during repaste using the supplied beQuiet paste, leaving the paste having a physical contaminant as the culprit.

Either way, I'd contact beQuiet and see what they say as to any replacement equipment.

 

[Most Worrisome]

Oh. Well, maybe I got a bit luckly, but I've managed to re-intall the cooler without any apparent new damage. Maybe I'm laying the thermal paste really thick.

I'm not sure what state the cooler base was in originally. Didn't really occur to me to look at it that closely. I do kind of wonder about the beQuiet paste I wiped off the CPU just before discovering the damage, it felt kind of chunky compared to the MX4. Unfortunately it occured to me too late to examine it. I could look at the leftovers in the tube I guess.

I did contact beQuiet when this first started, but they haven't got back to me yet. I'll definitely be taking what you've said into account when they do. Thank you.

 

[Karadjgne]

Oh, it's paste. It's supposed to be a creamy grease, definitely not supposed to be 'chunky'. Even Diamond IC which has a diamond dust component is so superfine it doesn't even feel gritty.

 

[Most Worrisome]

Okay chunky may be the wrong word, I didn't actually check it for solid bits. It was just kind of bunching up into smaller blobs as I wiped it off while the other stuff came off much more easily and evenly.

I tried touching a bit of the leftover beQuiet paste and it doesn't feel gritty at all. But when I was first using it, I had to seriously strain to make anything come out of the tube, to the point I was afraid the syringe would break. Now it comes out easily. Could there have been something mixed into the tip of the tube that's gone now?

 

[Karadjgne]

Maybe some dried out paste, or contaminants of some sort, no telling now.