XPS 13 9365 = Sleep and Hibernate problem after Fall creator, computer goes to AIRPLANE mode

"PS; for the one not able to start, simply go in the BIOS and setup the disk to AHCI back again.

By the way, who could make a list of the OPTIMAL BIOS PARAMETERS for the XPS 13 9365 ?

(this would be useful as this RTC proceudure by me, reset the full BIOS settings)"

Thank you Slaredo, I set it to AHCI and I can boot Windows again.

And yes if the 9365 BIOS has been reset it would be great to know how it should be set up, if not as default.

First, does anyone know how to do the equivalent of this RTC reset on a Precision Mobile 5510? I tried it and it doesn't work, which is not surprising since the article doesn't list the 5510 as supported (though the 5520 is).

Also, I find this problem disappears if I disable Hyper-V (via bcdedit); is everyone with this problem running Hyper-V? Unfortunately I need Hyper-V...

If anyone was wondering, the new Bios does not seem to cause any problems with the hibernation situation.  Modern Standby seems to be working better but sound and other hesitations continue.

I just did a clean install using the latest Win 10 build and it seems to be performing nicely.  The reason I did that was because I was having problems accessing my other systems and NAS in Explorer..

Hi everybody and thanks to Saltgrass for the information.

I saw th BIOS and was waiting for such a message before installation, I didn't want to start in another 3 or 4 months worries

All installed (BIOS and graphic update also) and all is still working fine = I start to enjoy again the machine, however will always stay a bad taste for the quality of service of DELL for this issue.

Thanks and let's stay tube in 2018, help on this forum is great !!

Hey @jsberg, I have the same issue as you do, and yes, I have hyper-v running (need it for docker).  Disabling hyper-v solves the issue.  Re-enabling it makes the problem come back.  I wonder if playing around with any related BIOS settings change the behavior...

I tried resetting the BIOS to factory defaults, and played with a couple settings, but there are really not too many things one can change. At this point I've rolled back to 1703 since that's the only solution I have at this point. There is a recent 1.6.1 BIOS that I've installed, but I'm going to wait at least until Patch Tuesday before trying 1709 again since the 1.6.1 release notes give no indication of addressing this issue, and I've vaguely hoping that somehow MS makes a change that fixes the issue. Staying at 1703 is not a good long-term solution unfortunately...

Thank you for trying the BIOS update; the update came out after I had rolled back and re-installing 1709 is enough work that I wanted to avoid it until I had some hope of the problem being fixed. So you saved me some work.

What's unfortunate is that this thread indicates there may be a fix for the issue, if we knew how to do the equivalent of the RTC reset procedure on a 5510. Removing the coin cell battery, which may or may not do the equivalent, is not just a disassemble-reassemble operation, it involves fan removal and reinstallation which means cleaning off and reapplying thermal paste, something for only the very confident to try.

I agree with them that it is a software issue, but I'd be a bit surprised if a full reinstall would solve the issue. It looks like something in Dell's BIOS is not getting along with something going on in 1709; if I had to guess from the symptoms I would speculate that it has to do with the new battery pass-through feature, but that's just a wild guess. Dell and MS would have to get together to fix that..

I just installed BIOS 1.6.1, and no joy.  Same behavior when trying to hiberate.  Looks like there are a bunch of other driver updates recently as well, so I might give those a shot as well.  Another option is removing the CMOS battery, to approximate an RTC reset, but you have to remove the entire motherboard to get at it, and I'm not sure that's a great route.

Also, my 5510 is covered under extended pro support, and that kind of intervention might void such a warranty, so I'm steering clear...  They're saying this issue is a software issue, not a hardware issue.  Their suggestion is to reinstall windows from scratch, turning off microsoft's automatic driver installation, using only dell drivers, and see if that works.  That's a major pain, and a major time investment, but not sure what the other routes are at this point.  This is definitely a problem, but doesn't seem to affect a vast number of users, so hard to imagine getting an engineer somewhere dedcating time to this...

I agree with them that it is a software issue, but I'd be a bit surprised if a full reinstall would solve the issue. It looks like something in Dell's BIOS is not getting along with something going on in 1709

It is only a software issue in that 1709 changed something which conflicted with some Bios function, probably related to power settings.

The RTC reset, according to the Dell page, not only resets the RTC but the entire Bios.  In addition it removes provisioning for the vPro processor.  But, as you mentioned, this thread is specifically for the 9365.  I would think starting a thread with 5510 as the subject would be better for anyone with that system.

Microsoft, after my explaining the problem in great detail, informed me my new Dell system was not compatible with 1709.  It turned out Microsoft was wrong.   

I have not dismissed the theory that Microsoft wants systems to stay in a low profile state so they can push what they want.  

I actually did start such a thread:

en.community.dell.com/.../20026787

I posted here because a) I didn't get any response to that thread; b) this thread describes exactly my symptoms; c) this thread has a solution and was hoping for someone knowing how to do the equivalent RTC reset on a 5510; and d) trying to find out if the Hyper-V correlation was there for the folks seeing this problem.

I'm speculating, possibly incorrectly, that the problem described here and my problem on the 5510 have a common cause and therefore a common solution.

Hi Saltgrass and others,

I've been wondering what your current modern sleep system current draw is? I've been floating around 480mw to 540mw. I have not had the troubles you guys have had with hibernation not working and have been running 1709 since it's release. So I don't know what made my system different? I've been trying desperately to reduce the power draw during the modern standby sleep session so that my sleep time can last longer then 2 days. I've had my system since Feb 2017 and in trying to trouble shoot this issue dell has replace my motherboard twice (that was back in August and October) though I think the second replacement wasn't needed, there was definitely something wrong with my first motherboard, the second I misunderstood the problem I believe now. didn't realize disabling connected standby wasn't recommended.) My guess is that my motherboard may have skipped a bad bios update so It didn't need an RTC reset to work with 1709?? I'm running 1.2.1 currently, but I've noticed they pulled that so only 1.1.0 is available for download. I currently have an active ticket with them, that's been open since August....

I've just reset my RTC and am about to put my system through a short sleep test see it's power usage while sleeping. My battery has cycled so many times now that I'm down to a health of about 60% and SupportAssist want's me to replace it but I don't yet until I've nailed down this 1 to 2% power draw per hour while sleeping.

At best currently I get about 2 days of sleep time. If the battery was fully charged and brand new I'd get about 4 days. Microsoft's documentation recommends a system get's about 9 days+ using modern standby, by my calculations that's a power draw of around 180mw. What are you guys seeing for your sleep study?

Thanks in advance. 

Here's a screen show of my most recent sleep session:

I can't see your sleep study attachment but the modern standby system may burn 2% of the battery per hour if left in modern standby.  There are a couple of things to check.

In Settings-System-Power & Sleep, see if the Network connection is set to disconnect - Always.  It will stay connected when plugged in but that will not drain the battery.

Check to see if your system is set to hibernate after an hour or so which should stop the battery drain at that time.  So it will drain 1% to 2% the first hour and then stop.

Hi Saltgrass and others,

I've reposted the picture to my website so you should hopefully be able to see it now.

I'm not interested in the hibernation workaround because I use my laptop too sporadically throughout the work day that it's too annoying to have that 30 second hibernation wake up hit going into a meeting. Also my network state is set to disconnect when sleeping that's one of the things I've validated with my Dell ticket.

I'd really like to know what your guys' average modern standby power draw is.  Is it around 500mw? are some burning a little higher? 700mw now? or are some of us actually lucky and are around 200 to 300mw?

Here are my sleep study and battery reports.

http://cpuchip.net/dell/sleepstudy-report.html

http://cpuchip.net/dell/battery-report.html

In my sleep study report on any given sleep session I've noticed 4 things off.

1) I have no hardware sleep. I noticed in one of your earlier screenshots Saltgrass that you had something actually listed there for one of your sessions? 74%?
2,3,4) Under SoC I've got these listed as issues:
+ NoHwDrips.PGD_XHCI_DOMAIN
+ NoHwDrips.PGD_PCIE_DOMAIN2
+ NoHwDrips.C10

Though in my research Intel C-state 10 is the deepest C-state and is for S0 low power idle. I don't know what hte NoHwDrips prefix means, besides the obvious No Hardware Deepest Runtime idle platform state.

A tool I've used to help get a better understanding of what's going on is WPA's wpr to record a power trace. It's very informative. I don't know if you guys have found that tool yet so I'm just posting it here.

https://docs.microsoft.com/en-us/windows-hardware/design/device-experiences/capture-and-view-a-wpa-trace-for-modern-standby-diagnostics

Cpuchip, the lowest I  have gotten my system to drain while on Modern Standby is 1% per hour with the hardware showing ~67% in Drips.

The hardware improvement and drain reduction may have been because I was able to disable the SATA controller in the Bios since I use my M.2 drive as PCIe and the NVMe controller.

Other things which might be involved, are Bluetooth, Audio and whatever the C states show as being used.

If you want a copy of my SleepStudy, let me know.  I have wondered why someone would want to leave a system in Modern Standby but I suppose I should have spoken to an enterprise person.  My system comes out of hibernation in 12 seconds, so that is fine for anything I do.

Saltgrass,

Did you have to have to reinstall windows after you disabled sata?

I wouldn't mind a copy of your sleep study for my comparison. I'm starting to come to the conclusion that this high sleep drain (500mw+) is industry wide. I've been tweeting anyone who's mentioned modern standby to get their story, but most have been successful at switching to S3 sleep. My guess is that Dell didn't implement the S3 object is firmware; which would mean that installing windows 7 is impossible... and that switching to s3 is impossible. Asus and some other's that haven't mentioned their hardware stated that the power draw under windows 8 wasn't so bad, 0.3 to 0.5% per hour. So maybe Windows 10 is at fault??? I still get the feeling that something is not switching off on our main boards and that's drawing near 300mw+

Anyway thanks for the help.