06-23-2016 01:07 PM
As for power & battery tweeks - here's the big fish that had to go:
1 - Search - turn it off unless you really can't find things, it just burns power to make a table of words found in your files - it also degrades performance when running.
2 - Cortana - probably great on a desktop with AC power to run it, but again - it uses battery power to do essentially nothing most of the time. I don't miss that feature.
3 - McAffe (sic?) is not the most efficient antivirus with several processes running in the background, and often in the top 10 for cpu usage. Norton has fewer processes - and they don't eat processor time nearly as bad - thus lower power use. That bought me an extra 30 minutes of runtime.
The higher the CPU load, the faster the battery drains. Most of these changes were done to keep the CPU load around 3-5 % when I'm not running any program but task manager.
4 - User training. This is where the person owning the system needs to change habbits from a desktop AC powered system. I frequently had multiple browser windows each with multiple tabs open - each one uses a little power and cpu time. SOME webpages (with animated ads) can draw 30% of the CPU attention. I've learned to recognize those and close them. Now, I try to keep it around 3-6 tabs open for email and I don't leave programs (word/photoshop/etc) running when I am done using them.
5 - "become one with the system" - by that I mean sense when it isn't normal. Normal for me is no fan running, and very little heat under the keyboard. If the fan powers on or I feel it getting warm, then I open the task manager to find out why. Usually the cause is something in the background I don't need or a webpage burning cpu cycles for no good reason.
6 - last you can do some tweeks in the OS configuration for minimum / maximum clock rate on battery in settings. There are some gotchas here - say you limit the cpu to 50% of its speed on battery. If you have a complex job to do which would use 100% without the limit - it's now going to take twice as long (in theory) - but while running for twice as long, the display will be on for twice as long drawing backlight power, so in a perfect world you actually use a little more power. The number that I found most beneficial for saving power with cpu clock speed is the minimum (idle) clock rate. Windows 10 will throttle back the cpu when there is nothing to do. ~400 mhz is more than adequate for the minimum.
There is another 'battery saver' feature in windows 10 - it's intended to cut your display brightness & performance when the battery is low to squeeze the last bit out. You control what charge level this kicks in - usually 20 to 25 %. I set mine at 99% so that this feature is on all the time rather than just 'at the end' of the charge. One (good) consequence of this is power intensive tasks (updates, virus database download, etc) are pushed out until AC power is available.
HP suggests turning off wi-fi claiming it uses power. It does, but the benefit of powering off wifi & bluetooth is so small I couldn't measure it. The alternative of using a USB wired ethernet is likely going to draw as much power if not more. If you always use the system without a network connection - then the airplane mode is easy to turn net on/off.
In the system bios, there is a setting for the USB jacks to have 5V power 24/7 - regardless of if the system is really ON or OFF - even if no AC is present. The intent of this is to use your system to charge a phone - which is a very expensive power bank. I turned that off in the bios settings. Also - I had used some of these really tiny USB 3 ramdisks and found they get warm even when not being used, so I switched to the SD card for removable storage with a lower hit on battery life.
If you find other ways to extend the battery - please post.
07-21-2016 11:38 AM
I've had my HP Spectre X360 for about 9 months now, the Display driver crashing was a constant problem for the first few months until there was finally a graphics driver update that seemed to reduce the problem down from daily to occassionaly.
To be fair I have seen exactly the same problem on other makes of machine such as Dell and just as frequent so I think it is less to to with the hardware model and more to do with the Intel driver and Windows 10.
Recently the problem has returned to daily, I note that there is a new version of the Intel(R) HD Graphics 520 driver availble so I am in the process of updating that from
Intel HD Graphics 520
Driver dated: 12/31/15
Intel HD Graphics 520
Driver dated: 25/01/16
I regularly run the HP Support Assist and Check for updates and messages. It has sometimes picked up the driver updates but in this case it hasn't and it seems that only by going into Device Manager and checking for updates will you be aware that a new version is available.
The problem may not be of HP's making (although it might be) but considering the number of people who seem to have the problem it really is in HP's interest to help people resolve these types of problems and however HP Support Assist determines what updates are available they need to keep this driver in particular up todate until it is no longer a problem.
07-21-2016 02:09 PM
I started having this (EXACT same issue) in November 2015 when I bought my X360. I use my system heavily - it's on 24x7 and doing real work not just surfing the net. I found that what activity you do on the system affects the likelyhood of the display driver crash & possible restart - i.e. watching netflix or youtube/amazon seldom (if ever) had a problem, but number crunching could crash it every hour (website flightradar24.com was the worst but even typing a manuscript in MS-Word would cause the problem on my machine).
Around Jan 2016, the system updates really fixed the display issue well, BUT there were some new added bugs with the synaptics tracpad that were not there before. Also, the system started doing a blue screen of death about 3 times per week instead of display driver failure "warnings". A couple more updates [6 weeks] from HP and all problems went away. I used to get 3 to 4 display driver fails every 8 hours - sometimes 3 fails per hour if I was typing heavily (I type 85 wpm - slowing down seemed to make it better)
Finally around March 2016 the problem went away, and the BSOD went away as well. I could keep my system running for over a month, and I was really satisfied with the architecture at that point.
Around June 2016, I got just ONE display driver fail message - but no others. It was not the royal pain like it was in November. I also checked on other support websites and found other brands (dell, lenovo, etc) had the exact same problem. My HUNCH (and that's all it is) is that the combined on-processor graphics from Intel weren't working - perhaps the wrong memory page was getting written over. My HUNCH was based on the 'odd' patterns that showed up on the screen - every other line was majenta in a box of the display panel about 4" x 4". Usually when the video gets gibberish it is due to a corruption or overwritten data in the display memory contents (which are shared with main memory by the processor) - There was yet another patch rolled recently; I don't know if that really fixed the June problem or if it was just a one time fluke. I saw your post which makes me think this bug is resurfacing again (albeit not as bad) since I'm not the only one who recently noticed it.
In any case, it is much better now than before. When the display driver was semi-fixed around Jan 2016, the BSOD became the replacement problem - which made me wish I could roll back to the pre-fix with just a display driver issue box in the lower right corner of the screen. That took a good 2 weeks to patch.
Again, my HUNCH is that in the effort to reduce parts count by using the processor to interpret the touchpad, do the graphics on chip, and keep track of what memory is allocated to which of those tasks is the underlying cause of this. The only evidence I had to support this was corrupted lines in the display indicative of video memory overwrite. Intel rolled out patches even on Christmas day 2015 to address this issue (a couple days earlier or later for different hardware vendors). The Christmas patch only modified the problem and was followed up with 3 patches at the same time. The problem is not found in "reference design" pcs - but since each company does something special, custom patches must be written and tweeked for Lenovo, HP, dell, etc machines.
Getting this stuff resolved is easiest when we can demonstrate more than one system suddenly started having an issue after an update was automatically applied. I encourage anyone else reading this thread to please post their experiences with as much detail as possible - especially what you were (or were not) doing when it happened. There are some valuable tools from download.cnet.com for reading system logfiles and crash dumps. Now might be a good time to install those, just in case. The output from their interpretation of crash dumps clearly pointed to the offending system drivers - even audio.
My past experience indicated it takes 2-4 weeks for one resolution to be rolled out - longer if 2 drivers combine to cause the end result (which they did in 2015). About all we can do is document it, give the documentation to HP, be the squeaky wheel, and wait while Intel / Microsoft figure out the how & why plus release a patch specific to the unique hardware HP includes.
Hoping for the best - If this is happening again, I have all my notes and contacts from last year. Everett
07-21-2016 05:01 PM
Em seguida mudei a configuração do plano de energia equilibrado, na opção "Alterar configurações de energia avançado".
Na opção PCI Express > Gerenciamento de Energia deixei as duas opções (Bateria e Na Tomada) como "Economia de energia moderada".
Resolvi meu problema. Computador não está mais travando em BSOD (Tela Azul da Morte) e estou usando a bateria normalmente. Até que enfim uma solução que acho que nem a HP tinha... kkkk Valeu a todos e espero que aproveitem a dica, pois demorou quase uma semana intensa de diagnósticos para conseguir resolver meu problema por conta própria, após muita observação e testes.
07-21-2016 05:04 PM
Thank you for visiting our English HP Support Forum. We are only able to reply to posts written in English. To insure a quick response it would be advisable to post your question in English. The following links are here to assist you if you prefer to post in the following Language Forum.
Thank you for your understanding
07-21-2016 05:07 PM
Then I changed the balanced power plan configuration, the "Change advanced power settings."
In PCI Express option> Power Management left the two options (battery and the socket) as "moderate energy saving."
I solved my problem. Computer is not locking in BSOD (Blue Screen of Death) and am using the battery normally. At last a solution that I think neither HP had kkkk ... Thanks to everyone and I hope you enjoy the tip, it took almost an intense week of diagnostics to solve my problem on their own, after much observation and tests.
07-22-2016 05:22 AM
Well just following on from my previous posts and those that have replied to them. Despite having updated the Intel HD Graphics 520 to the latest driver (184.108.40.20677), and the BIOS, although I think the changes to the BIOS were superficial updates, the problem is still the same and with the same frequency.
I am currently experincing the problem when using two screens with the Dell U2414H external monitor running a udemy video tutorial.
The external monitor screen keeps going blank every now and then, gradually with more and more frequencty before the message to say that the driver has stopped and recovered. At which point the video tutorial has to be refreshed and restarted.
My experiences have been much the same as efbasham - the problem seemed to have been fixed - but now it seems to be back with a vengence although I am running things that I haven't previously run so it could be the old problem.
I am going to download the external monitor driver and see if that makes any difference as the problem so far does seem to only be a problem when I am connected to the external monitor.
07-22-2016 06:50 PM
In my (somewhat exhaustive) search for problems with video, I ran across a documented issue at Intel that *might* apply to your external monitor situation. Since the main processor does double duty for system tasks and video, there are thermal issues where complex graphics will contribute to the heat generated in the processor. It's been several months since I read it, but with a single (integrated) display panel, this was not an issue up to 4KUHD, but with 2 monitors there is a thermal derating curve which specified how many total pixels could be driven based on temperature and cooling effectiveness (combined display panel and external display). I seem to recall this info was at intel's website as it applied to integrated graphics on the i7.
A couple easy tests I can think of to determine if temp/pixel count is the controlling issue are:
1) Lower the resolution on both panels and see if the problem goes away at some point
2) raise the resolution to the most you can support, but put the laptop inside the freezer - see if the problem goes away - and if it does, take the laptop out and let it warm up to see if the problem returns
Intel stressed the actual limit on total number of pixels displayed on 2 panels was limited by thermal issues - I've only run an external HDTV display for watching video. At least that will identify if you have a thermal issue or a newfound driver problem. I'm also getting 'forced updates' today on the HP app - I've been postponing them to get some work done, but I'll install that update next.
Good luck - please keep us posted!