NVME high temps and fans control

Houston, we have a problem!
I get new NVME to add in my Dell Inspirion 15" 7510+ recently, so after installation my device is filled with one stock drive 512GB on M.2, PCIe 3.0 lane (swapped from master to slave PCIe bus) and one new Crucial P5 1TB drive on PCIe 4.0 master lane.

Unluckily the nvme’s drives among generate much heat than expected, around 65 C° and this ends with high FAN speed usage, indeed the BIOS triggered to cool down automatically when nvme sensors is above 60 C° I see, or something like “combined sensor values” get high temperature and so, weird frame layout of this laptop not allow a great cooling. :rage:

I’m on a big issue here, just not want to hear fans at full speed everytime. :face_with_diagonal_mouth:

Now I’m on Cassini EnOS live system, and PCIe both lane is set disabled from BIOS “Storage” setup. :innocent:

No nvme’s active no issue, fan control is fine. :exploding_head:

I think to apply on PCIe M.2 area some kind of metallic tape for cooling aid, like on socket base pcb and some over bottom laptop cover, aligned to M.2 top heatsink
HeatSink for this M.2 isn’t good enough, of course.
Also none of all software I try for control fan speed look effective on my weird laptop. I’m read all topics related to fancontrol, one interesting is intel-undervoltbut this isn’t CPU or GPU issue, I’m very hungry.
Now I take out bottom cover and make to work, I will post some pics of hardware layout… see you later folks!

Oh! the question is it possible ignore M.2 temps from fancontrol, if exist a way to go, I try @Echoa script Fan Control for dumb Dell Laptops , but no luck.

Thanks for any reply in advance. :green_heart:

Well usually people use some of those thin SSD copper heatsinks for laptop.

Yes, you are right, great idea thanks, I think to buy copper heatsink like this

But need to check internal laptop room, my laptop is very slim damn :persevere:

1 Like

Yeah, there are many different models, some are thin like paper and still better than nothing!
So just make sure to find one that won’t touch something it shouldn’t and you’re good to go :wink:

1 Like

Maybe you can use the thin copper ones that have a nano thermal pad if the others are too thick.

Example:

1 Like

you need to disable the motherboards fan controller in order to not have bios automatically ramp the fans/try to control them, this should handle it assuming dell hasnt changed anything
https://aur.archlinux.org/packages/dell-bios-fan-control-git

I need to update the readme on that tool so people know that my bad

1 Like

Thats a strange thing for the laptop to do, 65c really isnt that hot for an NVME drive :thinking:

1 Like

Yes I take some pics of intel - undervolt measure and sensors, from TTY laptop idle, not running DE,
I start from cold and after a while around 5 minute, the temperature growt and fans start on value you see here.





Not on my watch! :cold_face:

honka_memes-128px-39

My nvme temps (desktop)

nvme-pci-0100
Adapter: PCI adapter
Composite:    +39.9°C  (low  =  -5.2°C, high = +79.8°C)
                       (crit = +84.8°C)

nvme-pci-2200
Adapter: PCI adapter
Composite:    +26.9°C  (low  =  -5.2°C, high = +79.8°C)
                       (crit = +84.8°C)
1 Like

Yes your temp look “normal”, my nvmes temps looks higher by +15 C° on EnOS (liveuser) and I’m now on kubuntu with nvmes re-enabled, temperature is fine like below:

nvme-pci-e100
Adapter: PCI adapter
Composite: +42.9°C (low = -0.1°C, high = +79.8°C)

nvme-pci-e200
Adapter: PCI adapter
Composite: +39.9°C (low = -0.1°C, high = +82.8°C)
(crit = +84.8°C)
Sensor 1: +38.9°C (low = -273.1°C, high = +65261.8°C)
Sensor 2: +42.9°C (low = -273.1°C, high = +65261.8°C)

This mean the problem is under EnOS :sob: :sob: :sob:
But what is I don’t know…

@Echoa hello, and thanks for reply! I try to follow your directive after installed ‘dell-bios-fan-control-git’ and ‘i8kutils’ , just started services after ‘sudo modprobe i8k force=1’ and no luck, i8kctl print some errors and do not sows correct status for sensors and fans, return all -1 -1 -1 -1
Maybe this time is hard solving scenario ahead you think?

Maybe is a regression (?) on latest kernel upgrade, really what, some kernel will pump up my nvmes up to 55 or more C° in idle so, I don’t know.
This is very strange and happens on EnOS live user with dell-bios-fan-control-git installed in live and manual modprobe i8k module just not work on this laptop, maybe someone others guys can confirm this, is soo much welcome… :call_me_hand:

Anyway…
not sure how can fix it… Now I go to restart and boot Arch and see what sensors said… Just I keep updating,because I made good job today on my laptop… Well well thank you all guys for feedback, I love this. Next post with some pict from heatsink on nvme, bye bye…

All sounds very strange to me. Mine runs at a steady 42, with brief spikes to 50 under heavy use. Might do even better with a heatsink :grin:

1 Like

Did you try switching to LTS kernel? Also try enabling power save options for your deivces: Install powertop, run it with sudo, go to “devices” tab and set everything to powersave.

1 Like

I highly doubt that :enos: is a problem (unless something writes to it constantly like crazy), different SSDs produce very different temps.

1 Like

Yes indeed, is Arch system not EnOS.
Today start to load live Arch and playng around nvme-cli but I’m under learning sruff about this weird temperature increase… now try powertop as @mrvictory advice me…


Rhis is temperature on Arch live.

1 Like

The problem isn’t EOS. nvme drive normal operating temps are between 40 - 65 C with critical being about 85 C. They could run at lower temps too. My current motherboard has Pci-e 3.0 drives in it. The Pci-e 4.0 drives run a lot hotter and they are twice as fast or more. The Western Digital Black SN 850X was the drive I’m getting for a new build and it was faster than the Samsung 980 Pro but now the Samsung 990 Pro is faster. They run hot! I run a heat sink on the one drive that is the Pci-e 4.0 slot even though it only has a Pci-e 3.0 drive because the board came with it. You may have a particular package on the laptop that is causing the increased temps.

I would also caution that it may not be running as hot as it is reporting. These figures aren’t always accurate.

My motherboard also has a chipset fan that activates when needed. Laptops are a totally different animal.

2 Likes

Yeah man, now I’d change bad tunables from powertop Arch live, and waiting for nvme ccooling down but appear temperature still steady around 50C° and fans still running not at high speed seen on my EnOS, but still running and laptop leave idle.






Sorry I know for take pict isn’t good to read but is my only way on live system.

I also newd to learn how to commit these changes under installed system.
On powertop Overview i see a drop of summary wakeups/second now around 41.

The initial status you can see on pict, wakeups is very high for idle laptop!

That’s not a hard rule, in terms of temperature it still depends on drive itself for the most part.

In general Pci-e run hot! Or hotter! That’s all i know. I’m not concerned about temps that much because if it’s getting too hot you are going to know it. Drives running that hot would cause problems.

Edit: I also think a lot of linux users with laptops install every dam package they find trying to manipulate stuff that doesn’t even need it! So they end up with packages that are causing more issues than they are doing good. Just my opinion. :man_shrugging:

1 Like

Here we can see my work for improve cooling on weird Dell laptop, I apply some kind of alu-tape.

Internal room is very low, heatsink maybe fit if like stock heatsink or 1 mm more height.

And remember to all, on 'buntuOS my nvme temps look great around 42C° in idle and fans is quiet and off. No need to deal with powermanagement with 'buntu … this is another weird thing ahaha.