Default brightness won't stick

ADDENDUM: this behavior is intermittent. One instance I can reliably reproduce the backlight issue....and then it won't do it.

You need to downgrade the Intel DPTF drivers from your current ones to this one: www.dell.com/.../DriversDetails

Just install the driver from that link over the one you currently have. Rest the brightness and reboot and your settings will stick. These 5175s are full of bugs. Terrible machines especially with the CS drain issue mentioned in my other thread.

Any drawbacks to using the older DPTF?

Yes, the same issue on my 5175, it  returns 40% brightness spontaneously. Very inconvenient to use at night time.

None that I have observed through 2 months of use. In fact, the new DPTF seems to drain battery faster in use, though this may be a perception thing. I have not tested for this specifically.

@felixdd: Did this work for you? How is the sleep testing going with your 5175?

I haven't changed the DPTF yet; I wanted to finish the sleep testing first.

I did one sleep test on the weekend but work has taken over again. My most recent sleep test was 873 mW/h. I'll do one more 2h sleep before disabling the driver.

Ok good to know. 873 mW/h is about 2 - 2.5% battery loss per hour I guess. Not great. In your sleep report are all boxes in the "Top Offenders" section green?

Yep. Drilling down to the maximum drain. I've popped up sleepstudy up here: app.box.com/.../alq4v2atedq58hnsyts14q84roxyngmy

As you can see, most of the energy is being drawn by UNKNOWN process according to the SRUM. there's also a MBN.WCM.RME start/stop even that I don't know what to make of.

I'm doing to do a pair of studies now with the driver disabled. Then I'm going to consider pulling it out although, I like having the flexibility of having a GPS chip....

A couple of things. Unknown process I think is an amalgamation of all the energy use. It is not a particular process. The srumutil tool allows you to download a more detailed energy study and if you look on that, the Unknown process is attributed to screen on use.

MBN should be your mobile broadband. It looks like you have a driver/firmware issue there. Maybe you installed the wrong driver or you uninstalled it? (or you are disabling it and enabling it)

Looking at your sleepstudy, you are consistently averaging about 2.5% loss per hour. I am not including your battery slice from your keyboard in that estimation. That figure is not good but it seems like you are allowing apps to connect in the background during CS. I'm not 100% sure this is the case - just guessing from your WIFI activity. If you turn background apps off I think your sleep will improve. In anycase, your sleep study at least makes more sense than mine as you do have some drain offenders dotted around the report. Mine are all green yet drain is still always above 2% per hour, regardless of flight mode or disabling in BIOS etc.

Dell have agreed an exchange so I will report back once I get a hold of the new tablet and test.

app.box.com/.../52995wsnre6impekdprlnf25dj9sop35

All sleep sessions should be in airplane mode EXCEPT #5 (made a mistake)

Last 3 sleep session are with the WWAN card driver disabled.

As you can see, I hover around 659-924 mW/h

To rule out background apps, I'm running one more without background apps. The only things I have that can background are People, Mail, Calculator, FlipBoard, and a game. Since I'm on airplane mode I don't think that would've made a difference.

As you say -- not great. My previous tablet ran 200-300 mW/h I think.

EDIT: I did one more with the background apps disabled and a fresh reboot. 1060 mW/h. Airplane mode, driver disabled, broadband adapter active 97% of the time. So yeah, as you say airplane mode doesn't seem to make a difference. 

Now I'll have to think about whether to take out the card or not. I'll struggle with that decision a bit.

Good luck if you try the card extraction. Just a piece of advice. When you crack open the back, unscrew the tiny screw inside the sim/sd card flap. Otherwise the back won't come off. From what I understand, the back comes off relatively easily.

PS - disabling the WWAN is not a good idea - as you see from the sleepstudy, the system will throw a fit and treat it as a misconfigured driver and drain all your battery. Even without disabling the WWAN, airplane mode doesnt make a difference on battery life. It mystifying. My current theory is that there is nothing wrong with the WWAN, but rather there is a power leak somewhere in the system that just cannot be addressed by software. An Core M processor should not drain anywhere near this much power in standby.

The card is out. It wasn't terrlble to remove, and the clips actually seem fairly well built from some other electronics that I've disassembled. And I saw your thread in another forum and someone'd mentioned the screw so I knew to take out out.

In my trials at least, disabling the driver did not cause any significant difference in sleep draw.

I'll run a few trials with the card out. I do wonder whether the card has something to do with it since looking thru my sleep study, a few occasions the WWAN was active throughout much of the CS. period.

app.box.com/.../alq4v2atedq58hnsyts14q84roxyngmy

Well this is interesting.

The most recent 3 sleeps are with the card out, airplane mode, a handful of background apps enabled.

The numerous two-hour sleeps before are all with the card in, +/- driver disabling, and +/- background apps.

1) If you look at the prior ones, there is a high degree of variability between the drain -- anywhere from 659-1086mW/h; the variability does not appear to be due to disabling drivers (did not lower power draw) or disabling background apps (if anything, increased power draw).

2) Once the card was out, that variability disappeared. The three sleep sessions were 794, 794, and 798 mW/h.

3) Even with the card out, I'm getting ~1.5% per hour in sleep. Not good.

So conclusion:

1) the WWAN card is affecting sleep; usually negatively

2) Even with it out, the sleep is not much better.

After much testing on both a WWAN and non-WWAN equipped unit, I think I may have found at least one part of the cause of high sleep drain.

On both units I did a fresh windows install - not from the system recovery image from dell - but from directly from MS.

Then I loaded all my normal apps (Office and a few other tools) and observed drain overnight. The WWAN unit was losing anything from 700 - 1200 mW per hour and the non-WWAN unit was losing between 5-700 mW. So felixdd, you're right, the WWAN does add additional drain to these units. Not surprising I guess, but adding 3-500 mW is excessive and really not within MS spec for low powered parts. Anyway...

I happened to try one of the units without the SD card plugged in and left it overnight. It happened to be the non-WWAN unit. I registered about 150 mW per hour. Phenomenal! That's what these units should do. So I took out the SD card from the WWAN unit too and let it sleep overnight. Result was around low 400 mW per hour. Not fantastic but far better than before.

There may be something wrong with the Realtek card reader drivers in putting the part into deep sleep - I can't say for sure. But without an SD card plugged in, there is definitely less power drain.

I may continue to test for other factors but I think this may be the largest contributor. I think the other discrepancies in power drain may be down to the make of our SSDs and the firmware they are on. So be sure to have the most up to date FW for your particular SSD.

Hope this is helpful to 5175 owners struggling with sleep drain issues.