Related
I love AOSP ROMs. They are the bomb-diggity. I only have one problem with them:
GPS does not work for everybody.
Let me explain the problem. When doing a long navigation, after about 20 - 30 minutes, Maps will lose the GPS signal and stay at "Searching for GPS..." until toggling GPS OFF and then ON. After it loses the signal the first time, it loses it again and again in the same fashion in about 5 minutes intervals.
Now, this is not a problem for everybody. For some people it works fine, and for the rest, we seem to be stuck with this rather irritating bug.
I do have a theory, though - I believe that there are two (or more) hardware revisions for our phones. I believe that a lot of refurbs have an older hardware revision, that run into these problems, and the newer phones have a newer hardware revision that do NOT have these issues.
What I need is someone that has reported this issue to take a logcat during a navigation and to get the results onto this thread so we can get them to a dev that may be able to fix this issue. I would do this myself, but I am out of money and only have enough gas in my car to make it back home tonight after work.
So, please, if there is anyone running an AOSP ROM (CELB and WW both have the issue) on a refurb or near launch phone, if you have the time and the resources, take a logcat of a failed navigation. It may get us closer to a cure for this bug.
DEVS: If any of you read this thread, I have an idea - if it is possible, you may want to pull apart your kernels and take a look at the part that interacts with the GPS and compare that to another kernel like zanfur's or the HTC kernel, which both have nary a problem with the GPS.
I have already tried fussing with the *.so files that are for the GPS, and they either made the problem worse or completely broke GPS.
Let's see if we can find a cure.
-SSJ
I dont think its the hardware at all. I have run into this problems are certain versions of roms but once I flash another its fine. CLB2.7 gps worked great used it from Arkansas to Florida but the gps from the Cyanogen 1.1 the gps constantly lost signal.
edgeupgx said:
I dont think its the hardware at all. I have run into this problems are certain versions of roms but once I flash another its fine. CLB2.7 gps worked great used it from Arkansas to Florida but the gps from the Cyanogen 1.1 the gps constantly lost signal.
Click to expand...
Click to collapse
I don't think it's a hardware problem, I think it's a problem with the way that some kernels interact with different (read: older) hardware revisions.
I'm Hardware Revision 0002 according to Home > Menu > Settings > About Phone > Hardware Information.
I usually just use Google Maps to get the directions, then peek at them every once in a while to save battery.
Once I get a car mount for my phone (soon), I'm sure I'll use GPS more often though.
No troubles with any GPS I've ever done. It may be all Eris's, at random. Like when the 3g disappears and reappears and starts cycling (until you reboot the phone).
Same idea.
I have a refurb phone and while using CELB 2.7 I drop GPS signal every 20 minutes or so. When using Ivan's rom, I never drop signal.
I traded my g1 for an hd2. Gave it to my wife and she loves it. Only problem is the SOD. Until a solid kernel fix is made she figured this out and no SOD so far. She uses Aldiko a lot and you can adjust the contrast with a swipe down the left side of the screen. She turns it all the way down and its completely black. Swipe up the left side and screen is back on. Cheap fix I know but works for now till the real bug could be ironed out.
the only time i had an sod is when i tried 2.11 radio..
what radio you using?
I am using the tmobile radio 2.10.50. Or something like that.
if its related to radio, I have 2.08 and never had one sod
What rom are you using? Also which winmo rom? I know its not a radio problem.
So, could it be, that radio 2.11 and higher kill the screen?
Do you recommend to use only radio 2.08?
I am from germany and a week ago my touchscreen was defect and had to be changed for a new one. Could this happen because of the wrong radio?
SODs have nothing to do with the radio. it has to do with the SD card. i can confirm this by taking my sd card and putting it in 2 different hd2s with 2 different radios and i had no SOD what so ever but when i would use a different sdcard i would get SODs on both
I've heard that higher class SD cards have better results. But sods are typically causes by the g sensor bottle necking when the phone goes to sleep. Aldiko prevents phone sleep but you can shut the screen off by turning the contrast down all the way with it. Swipe down off. Swipe up on. Like I said cheap fix but works for the time being.
I use a class 2 sd card that came with the phone. the only form of SOD i experience is the screen powering on but it remaining black which is only temp, as a few turns on attempts anbd its fine. which i realy wouldnt consider screen of death since it never sticks like that permantly to requiire a reboot.
That's lucky for you its a common bug at the moment. I have to baby my wife's hd2. I'm a bit jealous. Lol
at0ms0ft.starkiller said:
I've heard that higher class SD cards have better results. But sods are typically causes by the g sensor bottle necking when the phone goes to sleep. Aldiko prevents phone sleep but you can shut the screen off by turning the contrast down all the way with it. Swipe down off. Swipe up on. Like I said cheap fix but works for the time being.
Click to expand...
Click to collapse
thats not true. i use the stock 16gb class 2 sd that came with the phone (TMOUS) HD2 and i have no SODs at all. I experienced SODs with a PNY 8gb Class 4 so higher class doesnt mean less SODs. it has to do with how the sd is formatted and if the sd has errors on it.
That's what I'm using the 16 gig that it came with. I'm just saying what I've read. Now that you said that I know I can save money and not have to but a expensive SD card lol.
Viewing the answers it seems to me that nobody really knows what is causing the SODs... someone said that could be the SD CARD, others that is on the ROM side, others Radio, others GSENSORS... in my opinion have to do with the early stage of Android development in HD2, maybe Cotulla will be able to release a new Kernel which fix the SODs and Freezes problems...
Use chuckydroidrom with radio 2.12.50
Sent from my HTC HD2 using XDA App
The Radio is the problem
ok Ive done this for 3 phones and I can tell u that SOD has only to do with the Radio. The only thing ive seen that the rom affects its the robot voice.
get this Radio Leo_RADIO_2.12.50.02_2
http://www.megaupload.com/?d=9GU2FF3X
And Im a 100 percent sure is gonna work and if u want get the chunkydroid just in case.
Also the 3 phones which ive put Android had been from tmobile USA.
Using chuckydroidrom, supherhd2 1.2, class 4 sd card -- I don't get SOD other than sometimes waking from sleep. Either it's the blank screen which just takes a few tries, or it shows up but is unresponsive, then the wifi reads "error."
Unfortunately that happens a few times a day.
[edit]
Also using radio 2.12.50.02
Great Success!!!
Guys,
I have had great success running Miri's rom / 2.11 radio / turned windows mobile sense off / running Nexus One build FRG33 9-8-10
I have had now screen of death, no audio issues, no bluetooth issues, camera & camcording work in 5mp, wifi working, wifi portable hotspot is working (I desperately need this because of my job), no freezing, all apps working, my NES/SNES emulator works like a charm, no freezing issues, Quandrant speeds range from 1550-1650 (low speeds when snes emulator running in background).
My only issue is that, when in sleep mode, the orange light stays on constantly, but even with this, I have had battery life of 36+ hours. This setup is ready for daily use.
If anyone can help me get that orange light out, I would be much obliged.
you would think after all the great development going on here, this issue (the biggest issue it seems) would have been the first to be sorted out. I for one am pulling my hair out over it. i get it constantly on all the builds I try except for darkstones which kills my battery in 2 hours. running stock rom, 2.08 radio, stock 16 gig sd card
i have tried radio 2.10.50.26 and 2.12.50.02.2
all have SOD. all rebooted when i tried to log in at the market place
i have tried nearly all the roms
it seems like only mattc is the most stable..
but he hasn't have any updates for the last 10 days.
thanks
jack
jack9999 said:
i have tried radio 2.10.50.26 and 2.12.50.02.2
all have SOD. all rebooted when i tried to log in at the market place
i have tried nearly all the roms
it seems like only mattc is the most stable..
but he hasn't have any updates for the last 10 days.
thanks
jack
Click to expand...
Click to collapse
Cant agree with you on that. I have also tried different roms, and stayed on Darkstone 2.1 from 18.8.10. I use 2.12.50.02.2, Elegancia v10, Class 2 16gb card which is almost full and I have only 2 issues:
1. SOD after you try to wake the phone up and immediately put it to sleep (Power button, and a second after again). It takes about 20 seconds until it starts working again. My suspicion, gsensor/sdcard, nothing to do with radio.
2. Bluetooth, A2DP controls do not work on my Bluetooth headset.
So, for me, this is the most stable.
BTW, I can run a battery for around 24 hours with gmail push, exchange push (through Touchdown, not native Android program) and average usage of calls, SMS, Wifi, Youtube and Market..
Whilst a full Android Build has some flavours thus making it taste-dependent, kernels are the core component of each meal here - and therefore I think it is a valid question - which is the best kernel so far.
There are several - each one with a problem
1. some freeze due to gsensor
2. some work with RMNET, some with PPP only
3. some have premature flash on camera
4. some have louder sound volumes
5. some have tethering
6. some cause random data disconnection ..
But it seems that the problems are different. It seems that if the best of all kernels is collected into one - then there will be a perfect kernel.
I have tried several but lost count and trace of which kernel exhibit which problems.
Important for me are 1, 4 and 6 to be working. The rest are bonuses. Have you tried a kernel that meets this?
If not - could you describe your kernel - what is working and what is not - taken from personal experience !
Hello,
I tried the froyo 2.2.
It freezes when I am calling someone (usually I can't do missed call because the OS freezes and I can't end the call before my friend opens on me the phone and answers :S)
This is the most annoying.
Also sometimes It looses the touch. Only buttons work
assaad33 said:
Hello,
I tried the froyo 2.2.
It freezes when I am calling someone (usually I can't do missed call because the OS freezes and I can't end the call before my friend opens on me the phone and answers :S)
This is the most annoying.
Also sometimes It looses the touch. Only buttons work
Click to expand...
Click to collapse
Yes that seems like a G-sensor problem - if you use GScript Lite and the two little scripts to kill the gsensor - freezes will go away. But you will loose the gsensor
iivuch said:
1. some freeze due to gsensor
2. some work with RMNET, some with PPP only
3. some have premature flash on camera, some will be good
4. some have louder sound volumes
5. some have tethering
6. some cause random data disconnection ..
[...]
Important for me are 1, 4 and 6 to be working. The rest are bonuses. Have you tried a kernel that meets this?
Click to expand...
Click to collapse
That's a general issue of actual kernel-developement. I think I have tried almost every kernel since the past weeks.
Whatever kernel you are using you always have one ore more issues.
The actual evo-kernel gereration resolved touchscreen freezes, has tethering, louder volumes but only works with ppp. There are (new) problems with camera (flash), music stuttering and random data disconnections.
The older desire-based kernels do not have random data disconnections (especially on rmnet), camera works imho better (but not perfect), but have random touchscreen freezes (if gsensor is turned on), no tethering and are more quiet in using the phone.
You see that there is always a compromise you have to face.
AFAIK the kernel developers have their focus on developing an EVO-based kernel and they are working hard to resolve the issues. The desire-based kernels don't have a great future because there seem to be general issues that cannot be resolved.
Not to forget the new Desire HD. I cannot estimate, what potential there is for using it on HD2.
IMHO it will take another 3-6 weeks to resolve the greates issues with EVO-based kernels. Maybe earlier...
Indeed!
Yes, the EVO kernels fix some problems probably because they are not very mature yet. GSensor still not working properly, and although freezes have gone, an annoying lag to wakeup is still there (not often - but still - noticeable)
I tried mattc 1.8 .. which is based on EVO, and although I liked the nice strength of the sound volume, it would drop data from time to time. It is great when it is working, but you can't see if it is working, unless you open something that needs data, because the indicator always stays - as if all is working well. And that of course means - all your background sync will fail from time to time.
I reverted back to mattc 1.7 RMNET which has both very fast and very stable data connection. I had to disable the GSensor - to get rid of the screen freezes.
Anyway - lets support the developers - they are doing a great job! I am sure that not before long, we will all have very stable Androids on our phones, and then we will be creating yet another partition on our SD cards to accommodate the new ports of WP7. On a completely different topic - the new HD7 - apart from having slightly more RAM - has the same hardware specs as our HD2. It seems that HD2 will keep its greatness for a bit longer And yet, its price is dropping on ebay, as it is now almost a year old model.
It seems that we all post our results on standby with certain thing enabled (GPS, WiFi, Data Link, Synchronization) but we keep on dancing around the subject of bluetooth. I've noticed that a lot of us have claimed that bluetooth enabled in standby is the most prevalent cause of the 60mA+ consumption so we just relegate to turning it off.
The thing is, turning off bluetooth while in standby defeats the purpose of "hands-free" usage. Turning off bluetooth means we have to answer the phone by getting the handset out of our holsters/pockets/coats/purses, whatever then manually enabling bluetooth, then, in some cases turning our headsets on again and waiting for them to pair, just to be hands free.
Of course, that all sounds like a hassle but that seems to be the reality of the situation now. The Devs (Chefs) and community have done well in finding solutions for various issues, but I haven't really seen an initiative towards resolving the bluetooth issue, other than "turning it off during standby".
So I guess, long story short... "Has anyone found a fix (confirmed) with the bluetooth standby power drain problem"
seems like no one has any ideas . On a good note, bluetooth power consumption (while in call) peaks at the beginning, but into the conversation, you can see the power draw get less and less. Bellow 200mA.
The drain issue doesn`t happen in Sense builds I`ve been using Gauners HD NAND build for the past week with BT on and I was getting 4-7mA drain in standby. I`ve now moved to MDJ`s CM7 NAND build now that BT, GPS etc are working. However the drain is 63-66mA with BT on and an average of 7mA with it turned off. Not as yet seen a fix for it.
Radio
I have the same issue.
Have you guys tried different radios with any varying success?
The problem seems to be in the stock kernels but I thought I'd see if anyone had luck with other radios?
Same for me.
Some time ago i played arround with a SD build to change the three bluetooth libs from a sense build in my stock build. Powerdrain was low and i was able to pair a bluetooth device but unfortunately i cant connect.
Maybe a developer can have a look on the BT powerdrain problem?
Best regards
trilu
Cheer!! I've got this problem too.
-67ma if on,standby
-6ma if off
Didn't found this problem on WMP.
anonuser1 said:
Have you guys tried different radios with any varying success?
Click to expand...
Click to collapse
Have tried a few different radios to no success.
Hopefully someone figures this this out!!!
It would be nice if they made a sticky for people that have tried different "fixes". I've noticed as well that this is not a sense build issue. I have tried different radios as well to no avail. thinking about resorting to a wired headset, lol. I think this has something to do with the Kernel. I've read, in some threads about a "toolchain" issue that has been mentioned to be the primary cause of this. We may all find temporary or lucky fixes but I believe the end-all fix will come when someone figures out how to find the cause in the kernels we use for our builds.
Thought I'd just share this. I had been stuck with this issue for a while; basically needing to fully recharge twice a day. I tried a lot of things recommended on here and elsewhere; all of them either made minor differences, none at all, and some made things worse. Things in that category are wifi, 3g, bluetooth, gps, setcpu, task-killers, screen-dimming, resetting battery stats, starting Haret with the key lights on, etc. All these things can potentially make a difference, but didn't help much. Plus, most of them involve de-smart-ing a smartphone, and what's the point of that? I tried changing SD card, as I have a 16GB Transend class 6 card which I believe draws a few more mW than e.g. Sandisk cards, and that didn't really help either.
As an indicative measure, I could charge to 100% (and green LED) at 11pm, put the phone into Airplane mode (GPS off as well), kill all tasks that are not needed, go to bed, and by 7am the battery was on around 50%. So that's with NO comms, NO screen, nothing. Android battery use showed most of this usage was the screen. Er, no. I installed CurrentWidgit (from App Market) and got it to log; during this overnight lack of activity it averaged around 65mW, and was never below 60mW.
So after all that rambling, the answer is: radio ROM version. When I installed Android, I installed the latest radio ROM available at the time (still the latest I think); i.e. 2.15.50.14, from http://forum.xda-developers.com/showthread.php?t=611787. After pulling my hair out trying all the above, I flashed the radio ROM with 2.12.50.02_2, and as if by magic, current draw under similar conditions to above is about 7mW; i.e. 10% of what it was, and an overnight period as above goes from 100% to 96%. Much better
I should point out a few extra things:
a) I arbitrarily chose that radio version. I don't really have time to try them all, so if anyone cares to test and post here, perhaps another version gets even better results. To make them meaningfully comparable, install CurrentWidget, get it to log every minute, put the phone into airplane mode, GPS off, kill any obvious large apps like Maps, camera, email polling, and don't run any "battery-saver"-category apps (I had just Messaging, Google Services and Android Keyboard under Settings > Applications > Running Services), turn the screen off for a few minutes (don't move the phone either), then turn it back on and look at the CurrentWidget log. If you post your results here, include which network you're on.
b) This "secret" is sort-of hidden in a few other threads about this, but only in passing and not really prominent. I thought it deserved its own thread so people can find it a little quicker.
c) I don't know if there are other dependencies at play here; e.g. perhaps 2.12.50.2_2 is best with O2 (which is who I'm with), and maybe others are better with other networks. I don't know enough about radio ROMs to know. If people post their results with their network we might figure it out.
d) Changing radio seemed to also cure another problem I was having, namely the data connection on both 2G and 3G regularly disappearing, even though I had a 2G/3G signal. (I used to fix this by switching between 2 virtually identical APNs, while cursing and muttering...)
e) I'm using AmericAndroid GB233 CyanogenMod7 on SD.
Very usefull info indeed! i use the latest radio too and i face the same battery issues. I've tried all kind of stuff (task killers etc) but with no luck, mb i"ll try a couple of radio vers. hope yours doesn't carry any robo voice incidents (I use the CoreDroid v0.8 one)
Stop using SD, flash android directly to your nand, my battery according to Battery Diviner, lasts 25hrs with a full charge not on airplane mode. I used to do SD but I got tired of the slowness and huge battery drain, NAND is the way to go.
I am thinking of going NAND now, but it wasn't really an option while I might have needed to go back to WM just to make it through a single day without access to a charger. What radio version are you using?
But if radio version is the key, i see no point why we should abandon the sd option & move to the nand one. And the 25h -no charge- nand duration isn't such a big deal, is it ?
I have to version 2.15 and I am on nand.... my battery drain is around 3-5ma standby.... which I think is pretty good.... so I think nand is the key..... not radio
Sent from my HTC HD2 using XDA App
It would be interesting if you now flashed to 2.15 again and record your consumption, you might find the early high consumption is not repeated, your original radio may have been very slighty corrupt? I'm using AmericAndroid on SD and get 3mA drain in standby with radio 2.15.50.14, battery widget predicts 12hrs use with 3g phone on and fairly frequent use or 4 days in standby (doubt I'll ever test that one but the 12hrs is about right).
My only temptation to go for NAND is speed, SD is very slow unless using a RAM build and there are too few of those. Can't quite leave WinMo just yet so sticking with RAM on SD.
Yea... even I'm curious... flash back to 2.15 and let us know your results..
Sent from my HTC HD2 using XDA App
I have also the 2.12 radio since january with different NAND roms (mostly sense) and had always 3-5mA in standby. WIFI and 3G off, GPS On/Off makes no difference. Had never the robot voice problems at the first call and no other problems that 2.15 users not have.
There was already a thread where some android nand users confirm the better battery life of 2.12, but not such a big difference like in your case.
Have found a similar thread of WP7 users with the same results.
Any sd android users around with positive battery results? my battery gets raped in almost 15h with normal use! And yeah my stock winmo 6.5 rom does pretty well, while my Leo stays alive & kickin for about 4 days. And just to get a bit off topic, neither Froyo/Gingerbread got me crazy as an OS, nor the "old ship" winmo 6.5 can keep anyone excited anymore. I hope Honeycomb port will be developed in a stable well working level for our Leos.
I think part of the issue here is different peoples' expectations are different. If you are already getting say 12-24+ hours with ANY type of usage, you are probably in a different category to the one I was addressing. The issue I was addressing in my OP is getting extremely poor battery life, of the order of perhaps 6-8 hours with light usage. Several people seem to experience this; i.e. charging required 2 or 3 times a day.
The WM thread mentioned a few posts back is interesting, as there seems to be a clear trend of 2.12 giving better battery life over 2.15 (some anomalies though - be nice if someone who really knows these things intimately could suggest what the dependencies might be).
When I have some time I'll try flashing a few different radios and measuring in an at-least semi-scientific manner. When I have a little more time I'll try NAND with different radios as well. (Anyone with time feel free to go ahead and post results in the meanwhile though!)
nand that killed battery life
I have been using nand roms for a few months now, ever since mdeejay blessed us with his gingerbread builds.
presently using ginger-honeycomb theme rom.
Battery has to be charged twice a day with minimal usage. 2 hours of kindle/music is enough to fry the battery. radio is 2.15.
the rom is superb in appearance and functionality but battery is a real eyesore.
suggestions appreciated.
@fx.dcny: have you tried 2.12 radio as I suggest in my OP?
Hi All,
I'm using Radio 2.15.50.14 and when I use WinMo (ROM Elegancia 2.0), my battery lasts longer than 24h, with a few phone calls and mails and sometimes use of GPS.
With Android on SD, I must charge my battery 2 or 3 times a day. Even in Airplane mode, drain is very high (losts 40-50% overnight). I'm currently with Rafdroid HD 4.2 SD build, but things are same with all other Android SD builds I've tried.
I've just installed current widget, and drain is 60 mA in standby, which seems very high ! and I frequently kill apps with Advanced Tasks Killer (apps like Maps, Soundhound that always reappear !) but maybe this Task Killer makes more bad that good...
As suggested by the OP (thanks for the tip!), I'll tried another radio when I have some time, I still have a backup of radio 2.10.50.28_2, and report to you...
Do someone know how to find the culprit app from Current Widget log? I just see a long list of running apps, but no clue about which one(s) is/are eating the battery.
@thivi: my experience was that the apps listed by CurrentWidget were not helpful (in this context), and like I say when I was getting ~50% battery loss in airplane mode overnight there were no unnecessary apps running, and Android was reporting the majority of the battery use as "Display". Something else must be happening under the radar of the various battery monitors (both native Android and not). I'm guessing (only guessing at this stage) that something in the radio ROM code causes it to draw more current.
Hi folks, this is a first post for me but battery use seems to be a never-ending holy grail to Android users!! I have a HD2 running HyperDroid CM7 V2.0 Fearless Falcon on NAND with radio 2.15.50.14. The phone is a branded TM UK handset and have to say that the biggest battery-eater I ever met was Google Maps. Once uninstalled (installed it on first setup), and with carefull widget management, I can just get 3 days between charges. I must point out though that I am not a heavy data user or anything, I just like to have a good functional handset that I know will do everything I need effortlesly. The HD2 is just the thing. Just as an edit, with Google maps being installed, even if force closed, I only ever got 6 to 8 hours from my battery.
I have found this out from master rom thread. Most of the build recommend 15 but i recommend 2.12.50.02_2 if you have battery drain issue on standby.
sellotape said:
Thought I'd just share this. I had been stuck with this issue for a while; basically needing to fully recharge twice a day. I tried a lot of things recommended on here and elsewhere; all of them either made minor differences, none at all, and some made things worse. Things in that category are wifi, 3g, bluetooth, gps, setcpu, task-killers, screen-dimming, resetting battery stats, starting Haret with the key lights on, etc. All these things can potentially make a difference, but didn't help much. Plus, most of them involve de-smart-ing a smartphone, and what's the point of that? I tried changing SD card, as I have a 16GB Transend class 6 card which I believe draws a few more mW than e.g. Sandisk cards, and that didn't really help either.
As an indicative measure, I could charge to 100% (and green LED) at 11pm, put the phone into Airplane mode (GPS off as well), kill all tasks that are not needed, go to bed, and by 7am the battery was on around 50%. So that's with NO comms, NO screen, nothing. Android battery use showed most of this usage was the screen. Er, no. I installed CurrentWidgit (from App Market) and got it to log; during this overnight lack of activity it averaged around 65mW, and was never below 60mW.
So after all that rambling, the answer is: radio ROM version. When I installed Android, I installed the latest radio ROM available at the time (still the latest I think); i.e. 2.15.50.14, from http://forum.xda-developers.com/showthread.php?t=611787. After pulling my hair out trying all the above, I flashed the radio ROM with 2.12.50.02_2, and as if by magic, current draw under similar conditions to above is about 7mW; i.e. 10% of what it was, and an overnight period as above goes from 100% to 96%. Much better
I should point out a few extra things:
a) I arbitrarily chose that radio version. I don't really have time to try them all, so if anyone cares to test and post here, perhaps another version gets even better results. To make them meaningfully comparable, install CurrentWidget, get it to log every minute, put the phone into airplane mode, GPS off, kill any obvious large apps like Maps, camera, email polling, and don't run any "battery-saver"-category apps (I had just Messaging, Google Services and Android Keyboard under Settings > Applications > Running Services), turn the screen off for a few minutes (don't move the phone either), then turn it back on and look at the CurrentWidget log. If you post your results here, include which network you're on.
b) This "secret" is sort-of hidden in a few other threads about this, but only in passing and not really prominent. I thought it deserved its own thread so people can find it a little quicker.
c) I don't know if there are other dependencies at play here; e.g. perhaps 2.12.50.2_2 is best with O2 (which is who I'm with), and maybe others are better with other networks. I don't know enough about radio ROMs to know. If people post their results with their network we might figure it out.
d) Changing radio seemed to also cure another problem I was having, namely the data connection on both 2G and 3G regularly disappearing, even though I had a 2G/3G signal. (I used to fix this by switching between 2 virtually identical APNs, while cursing and muttering...)
e) I'm using AmericAndroid GB233 CyanogenMod7 on SD.
Click to expand...
Click to collapse
sellotape, you're the boss !
now with radio 2.12.50.02_2 I have 4-7 mA drain in standby instead of 62-66 with radio 2.15.50.14... And last night I only lost 7% battery (and I was not in Airplane mode), it was 40-50% with radio 2.15.50.14 (and in Airplane mode !)
So this radio 2.15.50.14 seems to be the culprit for high battery drainage in Android, at least for some users...
I was happy with the RAFDROID HD 4.2 build but had 2 problems : 1.high battery drainage which is now solved thanks to sellotape thread, and 2.heavy lags that were caused by my SD-Card, problem detected thanks to this thread :
http://forum.xda-developers.com/showthread.php?t=1048649
so now I'm enjoying a very fine Android experience with my HD2, even with a SD build
@GPZ550: it's true that Google Maps is almost always in memory, even after killing it. But I'm a heavy user of it so I don't want to desinstall it. And with radio 2.12.50.02_2 it seems that I can have a decent battery life even with google Maps installed.
I don't want to add confusion here, but my experience as in the OP came to an end about 2 days ago, for reasons I couldn't really associate with anything in particular, except *perhaps* an app update from the market. So... what I did is flashed 2.15 again (to at the same time answer boomboomer and fahd_freak's questions), and I am now back to <10mA standby (down from ~60). I suspect if I'd flashed 2.12 again it would also have fixed it, but I'm out of experimental options in that direction for now. This leaves me rather confused. I would have thought that apps (even under SU) would almost certainly not have sufficient access to alter the Radio ROM.
As a dev myself I struggle with the idea of the ROM becoming "a bit corrupt". I would have thought *any* corruption would have been fairly catastrophic. And yet, mysteriously, re-flashing the ROM seems to be an answer (for now). Perhaps there is some state data that the Radio ROM stores that gets corrupted or invalid and this is overwritten with a re-flash?
Personally I'm not too fussed if I have to reflash the ROM every month or so for now, ridiculous as I find it in principle. It does perhaps mean it's not just radio version that's at issue, but perhaps the action of flashing itself.
I managed to get 5mA draw in standby mode on the NexusHD2 for SD builds (3G on but GPS and Wifi off), with some spikes (when data is being pushed). Able to last a whole day with moderate usage. (Radio version in my sig.)
But really, the NAND builds are much better overall. I managed to get a very similar battery usage pattern as above with the latest NexusHD2 Gingerbread ROM V2.7, 3G, GPS and Wifi all on.