i know why battery life sucks so much!

[tmz]

Here's a review on the Diamond that talks about the battery life and the phone having a little more heat then others.

Thanks T M Z


http://www.pcmag.com/article2/0,2817,2331270,00.asp

[tg1.0]

I believe the discharge current is calculated by determining what hardware is on and how much the CPU is being used. It is just too consistent to be a value that is physically measured/sensed.

I have a simple utility (attached) that can display the discharge current when you ask for it, as opposed to continuously. With the radios off and the backlight on the dimmest setting, my battery current is always 57 mA. One increment in brightness adds 9 mA. With the nuePower/Backlight CPL utility, however, the battery current becomes exactly 100 mA higher because of the continuous polling.

If you take your battery's mAh rating (900 for the GSM version, 1340 for the CDMA version) and divide it by the disharge current, you should obtain your runtime in hours. I've tested my battery several times and found that it actually lasts a little longer than it should, even though I'm not particularly happy with my battery life.

I think I may have found another reason for the poor battery life. Please see "Issue #2" in this thread of mine.

[the7thson]

Originally Posted by tg1.0

I believe the discharge current is calculated by determining what hardware is on and how much the CPU is being used. It is just too consistent to be a value that is physically measured/sensed.

I have a simple utility (attached) that can display the discharge current when you ask for it, as opposed to continuously. With the radios off and the backlight on the dimmest setting, my battery current is always 57 mA. One increment in brightness adds 9 mA. With the nuePower/Backlight CPL utility, however, the battery current becomes exactly 100 mA higher because of the continuous polling.

If you take your battery's mAh rating (900 for the GSM version, 1340 for the CDMA version) and divide it by the disharge current, you should obtain your runtime in hours. I've tested my battery several times and found that it actually lasts a little longer than it should, even though I'm not particularly happy with my battery life.

I think I may have found another reason for the poor battery life. Please see "Issue #2" in this thread of mine.

I agree with just about everything you're saying, and have my own results to report as well. Let me start off my saying I have 2 identical Sprint Diamonds in front of me. I ran nuepower on both of them as well as your program on both.

Both programs ALWAYS report the exact same numbers on both phones under the same conditions (which radios are on etc) so I think both programs calculate the amount of power being used - the only difference is that one is on demand and one is constantly polling. I ran the same test you did - all radios off and lowest brightness. Both phones reported 57mA with your program. Then I ran nuetools and both phones started off reporting 57mA and then in a second jumped to 157 due to polling like you mentioned.

With all radios off and the power setting set to auto adjust (which is 3 notches over lowest in my current lighting), both phones read 75mA, which is 57 + (9*3) or 57 + 18, again consistent with what you found.

With all radios off and ONLY bluetooth on (just on, not paired to anything), both phones report 61mA with your program. Again, with nuepower, both phones jump up by 100 to 161mA.


CONCLUSIONS:

1. Both programs - systeminfo and nuepower work the same way. They most likely calculate power consumption instead of actually measure it in real time. The only difference is that one is on demand and the other one is constantly polling and polling seems to consume 100mA extra.

2. Battery life sucks, the best we can do for now is put in some so-called battery tweaks etc to manage power better. With JUST THE PHONE RADIO on and all other radios off, if you use NO DATA, NO WIFI, NO GPS, NO BT and MAKE NO CALLS, the standy time is only 1340/75 = 17.8 hours. This is just simply phone standy time, with absolutely zero phone calls, data, bluetooth etc. Of course, this is with the screen on and standy time would be significantly higher with the screen off (will do some testing on this and report back). This is still pretty poor imo. Seems like I need at least a 3000mA battery for this phone to be practical

3. More than likely all phones work the same or are at least not wildly different in power consumption (at least stock). I think everyone in this thread should run all the same tests I did and report back what find so we can all compare. This could all be meaningless though if both the programs we're using to measure/calculate power are essentially useless.

Whew!

[cappy]

Good Posts mabmarc, the7thson, tg1.0. The numbers and analyses that you provided have moved this discussion forward.

Comparing apples with apples is important. Many of the numbers that all three of you have provided are identical, once equalized for the tools used. My phone reports similar numbers.

Following up on the7thson's suggestion, I have listed the available reported numbers for all 4 phones, adjusted by -100 for numbers reported from NuePower and assuming lowest backlight. Perhaps it will provide more clues. Please correct any errors.

All radios off
mabmarc - 57Ma (includes -100 adj. for NuePower)
tg1.0 - 57Ma
the7thson - 57Ma
Cappy - 57Ma

Phone radio on, others off
mabmarc - 57Ma (includes -100 adj. for NuePower)
tg1.0 - 57Ma
the7thson - 57Ma (or 218Ma?)
Cappy - 57Ma

Phone, Bluetooth On
mabmarc -
tg1.0 -
the7thson - 61Ma
cappy - 61Ma to 125Ma (see comments below)

Phone, Wifi On
mabmarc -
tg1.0 -
the7thson -
cappy - 283Ma to 301Ma (see comments below)

Phone, Bluetooth, Wifi On
mabmarc - 198Ma (includes - 100 adj. for NuePower)
tg1.0 -
the7thson -
cappy - 287Ma to 369Ma (see comments below)

COMMENTS and OBSERVATIONS
I have the Airave running, so the signal strength is optimal. NuePower is installed, but the numbers that I report are from a plugin tool called HomeScreen Plus. It always reports 100Ma less than NuePower. This plugin is installed in conjunction with the SecondToday screen, both by the same vendor. If you are interested, the link to the site is http://www.chi-tai.info/cs_wm_tools_im_cs.htm. I find this combination of the SecondToday and Homescreen Plus convenient and Homescreen reports CPU usage, battery status, and discharge rate, among other things, on one easily accessible screen.

Phone, Bluetooth On - the7thson reports 61Ma. My phone varies from 61Ma to 125Ma and tends towards holding at the 125Ma.

Phone, Wifi On and Phone, Wifi, Bluetooth On
When Wifi is On, the discharge swings often and does not go below 283Ma. It goes as high as 400Ma+. Also, I noticed that the backlight dims, but does not go off as quickly. In fact, it did not go off at all for the 3+ minutes that I watched it before turning it off. Without Wifi On, my backlight goes off after about a minute. I aslo noted that CPU usage goes up considerably. With Wifi Off, CPU usage hangs at about 1% at rest. With Wifi On, usage goes up to as high as 20%, but it is not constant. Note that I have not actually connected to my Network.

Miscellaneous 1 - tg1.0, your Password observations in the thread that you linked to are interesting. When I have time, I will try to duplicate the test.

Miscellaneous 2 - The way I have my phone set up (no more energy to go into detail), the Second Today screen briefly appears when turning it on from standby. During this brief period the discharge reads 13Ma. Perhaps, for my phone this is the discharge rate with the backlight off. I cannot be sure, but if it is, that indicates 104 hours of standby time can be expected. Just an observation.

Finally, moving away from the details, I have been monitoring the battery performance for 2 days. In the last 49 hours (6:40am on 10/2/08 to 7:40am today). After 1 hour of talk/data time, several weather checks, turning the phone on/off 30-40 times to check the clock, and switching to the settings screen a few times, my battery reading was 47%. During the last hour or more of switching screens and changing settings to report the findings above, the battery dropped to 39%. The phone has been on the whole time with the phone radio on and other radios off.

NOTE: I just turned the phone off and on again to test someone's observation in another thread that the battery reading fluctuates and cannot be relied upon. Upon turning the phone on again, the battery reading did drop from 39% to 27%. Turning off and on a second time it remains at 27%. I don't know what to conclude from this. I suppose it is subject of another thread.

Quoting the7thson - "Whew" !!!

[tg1.0]

What we should do is test our actual running times with different hardware usage. I've already done three tests. I don't have a lot of time right now, so I will just post a quick summary.

Full backlight + WiFi, 258 mA: Down to 20% in 5:39.
Full backlight + WiFi + GPS, average of 464.33 mA: Down to 20% in 2:59.
Full backlight + All radios + GPS, 535 mA: Down to 20% in 2:30.

(All other hardware, including the cellular radio, was off unless I stated otherwise.)

If you do the math, you'll see that my battery actually lasts a little longer than it should! However, this is assuming that the calculated (not sensed) discharge current values are accurate.

[cappy]

Originally Posted by tg1.0

What we should do is test our actual running times with different hardware usage. I've already done three tests. I don't have a lot of time right now, so I will just post a quick summary.

Full backlight + WiFi, 258 mA: Down to 20% in 5:39.
Full backlight + WiFi + GPS, average of 464.33 mA: Down to 20% in 2:59.
Full backlight + All radios + GPS, 535 mA: Down to 20% in 2:30.

(All other hardware, including the cellular radio, was off unless I stated otherwise.)

If you do the math, you'll see that my battery actually lasts a little longer than it should! However, this is assuming that the calculated (not sensed) discharge current values are accurate.

A good suggestion, but I will find it hard to allocate time to do all three. I have just embarked upon a different test, in hopes of pinning down why my Bluetooth connection draws so much power. The Ma readings tell me that I should not lose 10% per hour when Bluetooth is On. So my plan is to turn all radios off except Phone and Bluetooth. I will not pair the Bluetooth during the test and I will monitor the discharge rate periodically by turning the screen on. If I receive any calls during the test period I can adjust for that because I have learned to expect a discharge due to phone use of approximately 20% per hour. At the end of an 8 hour period, I will have learn something, I hope. The phone is charging now.

[jmz768]

Does TF3D use a significant amount of battery power compared to using the regular look of WM6?

[cappy]

Originally Posted by jmz768

Does TF3D use a significant amount of battery power compared to using the regular look of WM6?

I cannot answer that, but I can say that all of the discharge numbers reported by me in this thread are with TF3D running. I have never turned it off. I aslo run Second Today (see one of my posts above for a link), which allows for the display of a Second Today screen that is similar to pre-TF3D. Both run simultaneously and I can switch between the two with no lag time.

[marcjen]

Today's test will be, unplugging from power and letting sit (turned off and face down) for the entire day.

I'll see what the batter life is and report if I received texts/emails or phone calls.

I did not do the BT test the other day, I needed to actually use my phone. I've ordered a new battery and cradle/charge kit, but I'd still like to help figure out if the battery life can be improved in some manner.

[snovvman]

Charging/battery is still an issue for me. I had the phone on charger showing 100%, LED ring solid. As soon as I disconnected, it went to 90% (w/ the 1% driver), within 15 minutes, it went down to 81%.