Start a Conversation

Unsolved

A

28 Posts

77345

January 19th, 2019 10:00

XPS-15-9570 (Bios 1.7) - Fan 3sec On-Off Issue on Idle dGPU

My XPS 15 9570 is from October 2018 and has all the latest Updates.

The fans are constantly running on AC-charging. Thats fine. I understand, that the XPS gets a little hotter whle charging, so I am Ok with the fan noise.
While on AC without charging (I limit the charge to 90% in the Power Manager and start only if below 75%) the fans kick in sometimes on little load. And if the load stays, the fans are also constantly on. Thats also fine.

But the really annoying thing is when the fans go on and off in an approximately 3 second cycle.

After BIOS 1.7 I noticed that the fans kick in on Battery when the dGPU is activated. So for example I start Lightroom and the nvidia card activates.-> The fans go on and off in an never ending cycle. (They end, when I exit lightroom).

CPU Temps: 42-43°C  Idle with usage below 10%

GPU Temps: 40°C        Idle with 0% usage

Screenshot_for_FanNoise.png

I tried the Temperature/Fan-Control in the Power Manager. There is no difference between Optimized-Mode, Quiet Mode or the Ultra Performance Mode.

Before BIOS 1.7 I didn't notice such a behaviour. It is like the fan curve was equipped with an extra setting to start the fans whenever the dGPU is on without looking at the temperature.

So they are going like crazy "The dGPU is on, we must turn" ....3sec later.... "But we are cool now. Turn off the fans" ....repeat_from_beginning.

Is it possible to let the fans kick in only if the CPU or GPU is over 60°C and let them run until they are below 50°C. I can't think of how a short burst of fan power could make a positive impact on cooling the device. Sometimes the CPU is just used for a second to turn on an application and therefore goes just above ~50°C for a very short time. The fans start to spin a burst. Then the CPU goes Idle and and the temps are again below ~50°C. And they would reached it even without fans. But thats another issue.

 

EDIT: I posted this video of the described issue on the second page. Just want to add it here in the first post for completion.

https://www.youtube.com/watch?v=7_w4lRP7rCo

4 Operator

 • 

3.4K Posts

March 21st, 2019 23:00

@A_Kri,

 

I have escalated this to the product team & have shared the links you have posted. However, I do not have an ETA on this.

 

 

28 Posts

March 23rd, 2019 09:00

Here are some more users that complain about the fan issues.

 

https://www.reddit.com/r/Dell/comments/b4abhh/dell_9570_fan_constantly_switches_on_and_off/

https://www.reddit.com/r/Dell/comments/b4iiwr/xps_9570_fan_noise/

https://www.reddit.com/r/Dell/comments/avbp8d/dell_xps15_9570_181_system_bios/ehnciyp/

https://www.reddit.com/r/Dell/comments/aeu2xx/xps_9570_bios_17_is_here/edwv7bi/

https://www.reddit.com/r/Dell/comments/aeu2xx/xps_9570_bios_17_is_here/edtfy08/

https://www.reddit.com/r/Dell/comments/b2flqg/dell_9570_freezing_and_fan_noise/

https://www.reddit.com/r/Dell/comments/akglnm/dell_xps_15_9570_fans_on_at_low_temps/

 

 

And it seems that some dell representitives try to help with the same suggestions over and over again. Try reinstalling IDPTF and so on.

Is this normal that dell representatives are not aware of this issue?

Not one time I saw a post like:

"Yeah we are currently looking into this issue. It was mentioned by others before"

 

I would guess, this work of collecting information about a specific issue from as many different users as possible, should be done by dell representatives?

 

Is there one specific person from dell assigned to this issue?

Otherwise the mods waste the time of the users and their own.

@Dell-SreejithR @Dell-Alan D @robert p  (I saw you commenting on fan issues before)

 

 

--------------------------------------------------------------------------------------------------------------------

For example in the case of the DPC issue there are:

- https://twitter.com/AzorFrank

- https://twitter.com/maknelly90

They collect information and report the progress.

28 Posts

April 4th, 2019 10:00

Status Update:


After the issue got escalated, I got this message in return:
"We had escalated this to the product team & we were not able to replicate the issue anywhere. Since the issue you have is that there is a fan blast every few seconds (which is not loud) & is not affecting the performance.

If you still observe the issue, we can have the system picked up for service & ensure that this is resolved at the earliest."

I am not willing to send my notebook in. I don't have the time to be some weeks without any Notebook.

Dell can't replicate the issue or does not want to recognize it.

There are many other users with the same issue and I collected countless posts here in this topic

 

I asked @DCFAdmin @SreejithR if he can reach out to those other users. Unfortuantely he can't:
"Unlike earlier, we do not have an option to reach out to a customer directly on the forum as this is assigned to us through a tool."

 

If somebody with the same issue is able to send his XPS to Dell for further research, please contact @DCFAdmin or @SreejithR 

 

-------------------------------------------------------------------------------------

I am not optimistic anymore that this will be resolved.

Next time buying a device or recommending it, I will definitely avoid DELL.

"Ohh whats this nice device that you have?" - "Forget it, its not nice, buy something else!!!"

And this goes for everything:

Dockstation? Rather not DELL

Monitor? Rather not DELL

...

Because, if they have such issues in one of their top product line, how is the situation for customers with low- or mid-tier devices. Pathetic.

And I will make more research in the forums of other brands to avoid a similar situation.

123 Posts

April 5th, 2019 05:00

This is just a JOKE... just a JOKE... READ THIS **bleep** POST:

https://www.reddit.com/r/Dell/comments/akngpy/incompetent_dell_devs_answer_to_bios_170_fan/

STOP LYING. STOP BEING INCOMPETENT. EVERY 9570 HAS THIS ISSUE WITH/SINCE BIOS 1.7.0 BECAUSE... IT IS SIMPLE. READ THE **bleep** POST. STOP LYING.

You changed the **bleep** bios with 1.7.0 so that any dGPU activity, even if it is just a milisecond, causes a 9 second fan hysteria, ignoring the dGPU temperatures. Reverting back to bios 1.6.0 solves the issue. Flashing back to >= 1.7.0 shows the problem again.

@A_Kri I told you btw this would happen, and they will lie to you about this that they cant reproduce this or have never got any other ticket about this.

@DELL-Cares

@SreejithR

66 Posts

April 16th, 2019 03:00

Hello A_Kri
I also have this fan behavior in my XPS 9570 from bios 1.7.0. But occasionally I noticed a kind of "lag" just before the fans start. That is, the system crashes for half a second and unlocks when the fans start.
Have you noticed this too?

28 Posts

April 16th, 2019 05:00


@Renx-78  schrieb:

Hello A_Kri
I also have this fan behavior in my XPS 9570 from bios 1.7.0. But occasionally I noticed a kind of "lag" just before the fans start. That is, the system crashes for half a second and unlocks when the fans start.
Have you noticed this too?


I didn't notice such a behaviour. But i will pay more attention to it in the future.

Do you have a specific example where you can reproduce this 'lag'. I would then try the same.

 

 

---------------------------------------------------------------------------------------------------------------------------------------

In the meantime, nothing has changed. On Frank Azors twitter account in the comments under the audio issue post, somebody asked about the fan issue and pointed to this forum post here. They seem to be aware of it, but just go on with the same reaction.

"I understand your call out and I'm sorry for the inconvenience caused. I need the system service tag or any Dell reference number so that I can review the details and assist you further. -Prashanth"

https://twitter.com/Dell/status/1114972832373592068

 

There is no point in asking other users for their service tags. You will waste their time, and get the same information that I provided here. And I am sure there are many others that provided Dell with some similar information about this issue. But everything is decentralized in some PM messages with different dell representatives.

In the end Dell will always say something among the following lines and close the issue:

-"Did you try quiet mode, and can you look at the FAN FAQ on this website"

-"Nahhh we can't reproduce it"

-"Your system in particular is acting weird"

-"Send it to us"

-"We will repair it"

66 Posts

April 16th, 2019 07:00

I can not understand if it happens by doing a certain thing, sometimes it does so even simply browsing the internet with a few tabs open and without other applications in use, once it happened to me while I was drawing notes with the image capture tool. I only know that among the many times the fans turn on and off they can occur. Besides, just before the fans start, I hear a kind of rapid hum like a vibration always coming from inside the laptop. I always check the Task Manager but there are no heavy loads of applications or services on cpu and other components.

123 Posts

April 16th, 2019 16:00

I have enough of Dell's lies over and over and over again. I got the same answer, "cant reproduce" or even this

"Good day to you! I have got an update from the Product Engineering team. They have suggested to perform a few additional troubleshooting steps. Please let me know when I can take access to your computer and perform it."

I have spoken to over 10 owner of the 9570 and EVERYONE can reproduce this in a matter of seconds/minutes.

Enough is enough. I will do everything in my power to make this public, and make bad publicity for Dell.

28 Posts

April 18th, 2019 03:00


@Renx-78  schrieb:
I can not understand if it happens by doing a certain thing, sometimes it does so even simply browsing the internet with a few tabs open and without other applications in use, once it happened to me while I was drawing notes with the image capture tool. I only know that among the many times the fans turn on and off they can occur. Besides, just before the fans start, I hear a kind of rapid hum like a vibration always coming from inside the laptop. I always check the Task Manager but there are no heavy loads of applications or services on cpu and other components.

 

The sound you described from the inside of the laptop is maybe the ram or the ssd.

Sometimes I hear it when I start a larger application (Simulation in Matlab). But I can't reproduce it at all times. Sometimes the sound also appears when I rsume from hibernate mode.

If you say that you can't see the loads in Task Manager, then I woul guess, it is the ram. A load on the ssd you woul easily see. But what the ram does you don't know. But that's just a guess from me.

 

I didn't noticed any lag in the past days. Maybe this is an additional issue? I wouldn't be surprised knowing the QC issues in Dell products.

123 Posts

April 24th, 2019 22:00

@Renx-78 What Windows version are you running? Type in winver in search/cortana. The freeze/lag you described "before fans turning on" is actually the typical famous Optimus (the automatic switch between the integrated Intel to the Nvidia GPU) freeze/lag, which was "fixed" some time ago actually "officially". It was never really fixed, but MS reduced the lag, so you wont notice it anymore that much, it is still present though. Read about it all here: https://answers.microsoft.com/en-us/windows/forum/windows_10-performance/partially-fixed-intel-nvidia-laptop-freeze-problem/93e7004a-62b1-4211-8e37-4c136608865e

The fan issue described in this thread basically is connected to the Optimus switch bug, but Dell is incompetent and doesnt seem to understand, what is going on here, where it is crystal clear and simple.

The incompetent Dell devs changed the fan logic with 1.7.0, that the moment the dGPU gets triggerd, the fans turn on. This was already kinda happening before 1.7.0  randomly, but with a way lower trigger chance. I think, the EC logic before 1.7.0 worked like, that the EC was checking for dGPU activity every x seconds, maybe every 2000ms or something. So that manifest into the fans randomly going on too before 1.7.0, but way way less. With 1.7.0 they seemed to have lowered the check rate dramatically, to something like miliseconds, or maybe 100ms or something. So it checks every 100ms, if the dGPU is "on".

The problem is now the Optimus bug, which is still present in Windows 10. You can actually see it happening, if you open HwInfo and double click on the dGPU temp sensor reading and just let it log in the background. If you then just go around and use Windows, the dGPU gets activated for just a few milliseconds randomly. It then looks like this:

https://i.imgur.com/pC2LHoI.png

And THIS is now triggering the fans up from 1.7.0 all the time.

Dell is incompetent. The Optimus bug can easily be reproduced on all kinds of Nvidia laptops. I am actually speaking to Nvidia support right now, if there is some hope they could fix this, but they dont seem to be interested in looking into it.

There are at least three different bugs with Optimus as I see it.

- the right click menu DLL hook is bugged, and triggers the dGPU randomly. deactivating the "show run as GPU processor" hook in the Nvidia panel, deactivates this trigger actually at first, but it will come back (I dont know why but it does, even with the hook off)

- the Optimus switch is bugged, at least on the XPS 15. it triggers / polls the dGPU for no reason randomly all the time, maybe just on the 4k model because of a too low trigger value of GPU activity, and Optimus tries to shift over to the dGPU.

- the Optimus Override is bugged and doesnt work properly.

Dell has to revert back the change they secretly made with bios 1.7.0.

I dont care they acknowledge their mistake. For me, they can lie as usual and deny there is an issue, but secretly fix/revert it back with an upcoming bios. But knowing Dell, they dont care and wont revert a mistake they did.

17 Posts

May 3rd, 2019 02:00

Mine with this configuration working on normal tasks I have no Fan or very low ;)

Current Status:

-BIOS 1.8.1
-All Drivers up to date
-Windows 10 1809 everything up to date
-Dell Power Manager on "Quiet Mode" for fan control
-Nvidia control settings for energy mode are set to "adaptive"
-Running on Battery power cable unplugged

 

28 Posts

May 3rd, 2019 03:00


@mhbuzios  schrieb:

Mine with this configuration working on normal tasks I have no Fan or very low ;)

Current Status:

-BIOS 1.8.1
-All Drivers up to date
-Windows 10 1809 everything up to date
-Dell Power Manager on "Quiet Mode" for fan control
-Nvidia control settings for energy mode are set to "adaptive"
-Running on Battery power cable unplugged

 


What are normal tasks for you?

My config is now:

-Bios 1.10.1

-All drivers up to date

-Windows 10 1809 everything up to date

-Dell Power Manager on "Ultra Performance" for fan control

-Nvidia control settings for energy mode are set to "adaptive"

-Running on Battery power cable unplugged

 

As you can see, our configs are pretty much the same. Just a newer Bios version and another setting in Dell Power Manager. The setting in DPM has no impact on the described issue. And I would suggest to use the ultra perfomance setting because the fans will no spin with max speed in the quiet mode. So you would have to change the setting every time you have a more demanding task like a render or a game.

The Bios version also doesn't play a role. Basically every Bios since 1.7 has the issue.

 

Currently I am browsing with a lot of tabs open and even downloading some files (3 downloads a 1.8GB), and the fans are off. The CPU is ideling at 44°C.

Works perfect?

But wait till I open an application which activates the nvidia gpu. In my case its Matlab, Lightroom, Photoshop, Davinci Resolve, ...

I will sometimes just leave them open in the background for a break to work later with them. They are ideling and do not demand any processing power. The temps are not rising.

And at this point I have issues where the Fans go berzerk. Every few seconds the spin on, and shortly after go off.

 

Especially with Matlab I just have to look at a static model and write a report at the same time. The fans make me crazy. 

I once even made some screenshots of various parts of the model in matlab, and then closed the application. But this workaround isn't staisfying as i have to reopen the application if I missed some part to screenshot.

 

And that is a more annyoing issue with DELL. As they suggest all kinds of tweaks where the users have to cripple their notebooks just to work somehow.

For me its: Just don't use any applications. Let your XPS sit there and look good. If you don't click anything, nothing will trigger the fans. Deal with it.

For the DPC issue: Put it in maximum performance, disable WiFi, disable C states, .... Really????? I have never heard of a PC that has to be modified that much to be able to fulfill one specific task.

There are a lot of more examples in other topics with other issues. Just look at all the suggestions. Please make a fresh restart, put it in this mode, and then in this mode, and later make this. But don't open anything else. Do you see how the super XPS is performing? Perfect. like in our Labs. 

Is it crazy to expect a bit more from their "top of the line" product? A bit more then ideling in a lab environment?

3 Posts

May 3rd, 2019 14:00

Hi i have the same -Bios 1.10.1 -All drivers up to date fans keep on going on and off regularly with just light internet browsing, and as I type this message, ac plugged or not, the problem is same. Also, fans keep doing this on sleep mode...

28 Posts

May 21st, 2019 05:00

@SreejithR 

Now the Dell Engineers acknowledge a change of the fan behaviour with BIOS 1.7.

Some weeks ago they couldn't even replicate this fan behaviour???

And now it is an intended change of the fan functionality. 

This obviously means an impact on the system functionality.

 

May I ask you @SreejithR , if the fan behaviour is intended with Bios 1.7, how the Dell Engineers couldn't replicate this fan behaviour?

 

 

http://forum.notebookreview.com/threads/xps-15-9570-owners-thread.817008/page-274#post-10913299

T1Z0Byi.png

No Events found!

Top