Surface Pro 4 - Screen won't wake from sleep

I have a feeling that most of the problems are caused by the 6th generation Intel CPU.

Maybe this new CPU is not mature.

All the previous models do not have such problems discussed here.

Does anyone have the same feeling?

Thanks dkb1892. You’re burning about the same rate I was after the 2/17 update. Which is under 1% per hour. We can deduce this by determining your battery capacity which is around 38,000 mWh (you can do a ‘/Batterystudy’ to confirm yours) so 380mW would equate to 1% per hour.  At least there’s some consistency with a new system and a post updated one.  Running this test helps avoid realized observations over timed intervals.

I noticed that crazy 1,147mW value on line 1. I get those on occasion where we see a fast burn rate (and likely warmer sleeping machine).  I don’t know exactly why that happens. It’s rare and if I had to guess, maybe some web site in my Browser may have caused it.  I suspect we all experience that occasional burn rate and that’s when having your system go to hibernate after some specified time will help limit that instance of sleep battery drain.

You can approximately confirm my <1% per hour usage by observing row 6 which says you used 341mW per hour and it slept for 2 hours and 5 seconds before going to hibernate.  Your total energy drain was 682mWh’s (or 2.000.. times 341mW) and it rounded that burn rate to 2% of battery usage (really 1.8%). 

My pre 2/17 update was in the 400-500mW range. After, I was in 100-300mW. A slight improvement but then the costly NoWake.   I changed out my graphics driver and that mostly fixed the issue except after a full charge.   Then after the 4/19 update, I had similar observations of NoWake with 100% battery except it happened with less regularity.  

The key thing that I noticed, that may be contributing to my NoWake after full charge, is that I get very low mW levels – in the range of 0-20mW after my 1^st^ sleep.  It also affected my timing to hibernate.  It seemed my SP4 was expecting a higher mW and so it took much longer to jump to hibernate than specified.  For example, if I set my sleep to hibernate for 30 minutes, it would remain in sleep for over 2 hours and not push to hibernate – I think because there is some connection to mW’s and sleep to hibernate.  You can observe this just in your 2 hours and 5 seconds.  I suspect your setting are set for 2 hours. Why the extra 5 seconds?  I would like to see if a new system also observes this phenomena, of super low power consumption after full charge, as it is maybe a related to a specific NoWake scenario.

If you do that full battery test, then immediate sleep on 100% battery and wake after 10 minutes, please share your results.  Thanks.

That 1,147mW was me leaving the Microsoft Store with the brand new machine after I updated the machine to the latest 10586.208 build, outside of firmware (wasn't yet at 40% charge with that update required). So I was still on an older firmware and clean machine until the 5th sleep

It is interesting I was getting some better numbers because I undated the firmware and installed Office, Acrobat. and downloading OneDrive documents. Then installing Quicken 2015 later on...I'd be willing to be Quicken has something to do with the higher burn rates, as it's ancient software that hasn't been brought forward well.

It is interesting that none of those burn rates get a green highlight, which is good news to me. It means Microsoft wants to do much better.

I just did a full charge, I have to look at what you want me to try out again?!

Try reenabling USB Power Management on the 2520, maybe a Windows 8.1 Update fixed that. I'd make sure ALL updates are installed, even the recommended ones.

That and the Surface 2 are great tablets, it really pissed me off that Microsoft didn't bring Windows RT forward. Especially since they were still selling the Surface 2 in January 2015

I rechecked the USB hub box and left my 2520 overnight after intermittent use and it didn't have any FTW problems. At some point over the years they must have fixed it. Maybe there's also a software solution the SP4 and SB waking problems? I also was one of the few people who actually liked Windows RT, the 2520 was really an awesome tablet. I remember presenting a PowerPoint slide show with it using an HDMI cable while an external HDD and mouse were attached through the type cover and someone goes "How is that even possible, my iPad can't even do that!" 

Another note, I set "hibernate after" to never, never, on my SB and left it unplugged for 9.5 hours again, this time I lost an average of 0.8% an hour. This is with wifi enabled and using the USB hub workaround. I realize that this isn't as precise a test as you guys are going, but it seems acceptable to me.