**** BEGIN LOGGING AT Sat Jul 14 02:59:59 2012 Jul 14 03:11:27 louisdk, if such is the case, I must admit that your perception is fairly diminuitive. Considering the fact that: 1) the way nokia handled the microUSB fault on N900 was really poor, 2) the way n9 was to be sold (i.e. in specific countries only), 3) the fact that n9 is also removed from the shelves because of poor lumia sales Jul 14 03:17:07 psycho_oreos, Well Nokia is still giving N9's away to seminars like KDE Academy and has helped Jolla to start their company. So it strange that they are both reacting and supporting the platform at the same time. Jul 14 03:19:04 sleeptime Jul 14 03:19:13 bbml Jul 14 09:28:24 vi__: Pong? Jul 14 09:28:25 23 hours late Jul 14 10:21:16 * DocScrutinizer05 hands MohammadAG a glass of tea Jul 14 10:21:33 O_O Jul 14 10:22:59 pronto: what? if you want one as well, just ask! Jul 14 10:23:03 :D Jul 14 10:23:06 Jul 14 10:23:09 o_o Jul 14 10:23:32 i had this really weith thai iced tea thing yesterday Jul 14 10:24:08 yes, you need a nice glass of tea as well. :-D Jul 14 10:24:19 weith= weird Jul 14 10:26:02 * DocScrutinizer05 prepares another glass of sweet moroccan nana-tea for pronto Jul 14 10:26:11 never had that Jul 14 10:26:14 * pronto drinks Jul 14 10:26:28 (i mainly drink yerba mate) :p Jul 14 10:31:55 I hope sun dried one Jul 14 10:33:11 is meego.com down ? Jul 14 10:34:53 http://www.downforeveryoneorjustme.com/meego.com Jul 14 10:36:02 up and dead ... Jul 14 10:36:08 well Jul 14 10:37:38 :nod: Jul 14 10:39:04 probably moving to maemo.org infra ;-P Jul 14 10:40:10 back up again, funny enough under https Jul 14 10:41:22 "thank you for attending Meego SF2011" WTF?! Jul 14 10:42:06 What's Next for MeeGo Jul 14 10:42:07 Submitted by Imad Sousou on 27 September, 2011 - 20:01 Jul 14 10:42:16 ohwell Jul 14 10:44:14 MeeGo planet Latest posts from the community MWKN Weekly News for Monday, 13 Feb 2012 21 weeks 4 days ago Jul 14 10:47:57 I'd say meego is "down" since Feb. 11. Jul 14 11:16:18 hey guys.i think i m having trouble setting up scratchbox for harmattan development Jul 14 11:16:38 basically my Xephyr's display seems to be black Jul 14 11:16:54 when i do a meego-sb-session start Jul 14 11:17:36 http://paste.kde.org/517496/ Jul 14 11:17:42 thats my scratchbox output Jul 14 11:18:15 I hate my N900 right now I Can't get U-boot to load omap1 kernel'n or kernel power 50 but every thing is right in the settings for U-boot what I can see Jul 14 11:18:22 http://paste.kde.org/517502/ Jul 14 11:18:31 and this is my Xephyr's output Jul 14 11:19:39 ooooookay Jul 14 11:19:41 nevermind Jul 14 11:19:43 found the solution Jul 14 11:19:52 didnt think the nokia docs are outdated Jul 14 11:20:00 http://forum.meego.com/showthread.php?t=3613 Jul 14 11:22:41 Somebody that think they have a solution for my problem? Jul 14 11:26:01 Dynamit: sure, don't use uboot Jul 14 11:26:13 haha Jul 14 11:26:38 I have U-boot to dual-boot Jul 14 11:27:10 daa Jul 14 11:29:11 okay this is weird... the Xerphyr's window size is bigger than my laptop's height and i cant seem to view half the applications :\ Jul 14 11:30:10 any work around for this guys? Jul 14 11:30:38 (apart from getting a bigger external display) Jul 14 11:31:45 change the display when writing in the terminal Jul 14 11:31:58 stupid explanation on that Jul 14 11:33:00 i'm searching for the qt4.8 on maemo Jul 14 11:33:09 for my n900 Jul 14 11:34:03 saidinesh5: -screen 854x480x16 reduce that to maybe ½ of that Jul 14 11:34:46 Dynamit: i changed the 480x854x16 to 854x480x16 Jul 14 11:34:58 assuming it would take it as a horizontal device Jul 14 11:35:04 but thats even worse Jul 14 11:35:23 it looks like the window manager in N9 has a hardcoded screensize or something Jul 14 11:36:50 * saidinesh5 still gives the halving thing a try Jul 14 11:37:03 nope Jul 14 11:37:05 it doesnt work Jul 14 11:37:16 it only clips the view Jul 14 11:37:21 it doesnt scale anything Jul 14 11:44:10 saidinesh5: /topic Jul 14 11:44:32 DocScrutinizer05: ? Jul 14 11:45:43 saidinesh5: you know, this chan has a topic, visible in usual IRC client at least via cmd /topic Jul 14 11:48:02 marbru: i'm not aware of qt4.8 for maemo - might be me Jul 14 11:48:20 oh my bad.. didnt notice that... Jul 14 12:26:02 http://talk.maemo.org/showthread.php?p=1237403#post1237403 I connected some dots, this is how you can get microb to speak with socks proxies... Jul 14 12:50:10 ~utg Jul 14 12:50:13 ~ugt Jul 14 12:50:14 hmm... ugt is Universal Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. http://www.total-knowledge.com/~ilya/mips/ugt.html Jul 14 12:57:23 * DocScrutinizer05 ugts teotwaki Jul 14 12:57:51 teotwaki: merry christmas Jul 14 12:58:49 happy easter kerio Jul 14 13:01:10 ~google google teotwaki Jul 14 13:02:08 I guess there's a good reason for chanset _default doesn't include google Jul 14 13:04:09 ~+chanset google Jul 14 13:04:09 Showing google values on all channels... Jul 14 13:04:09 google = NOT-SET(155): # ##bspress ##bz-inc ##debian-br ##ducleague ##essy ##guleague ##icf ##kierra ##ols ##pxe ##stouteriks ##t42 #aegis #android #arl #arm-netbook #asterisk #asterisk-bugs #asterisk-dev #asterisk-doc #asterisk-scf #asterisk-scf-dev #botpark #brits #brlcad #buglabs #byumug #bz-inc #bzflag #bzflag-chat #bzleague #bzmods #casualti #catroid #crosstool-ng #curseforge #cyrannus #dbgc #debian #debian-bots #debian-br #debian-france ... Jul 14 13:04:09 End of list. Jul 14 13:35:18 spamtastic Jul 14 14:13:01 ~ping Jul 14 14:13:02 ~pong Jul 14 14:38:50 so Jul 14 14:39:04 how can i get qt4.8 on maemo Jul 14 14:39:09 for my n900? Jul 14 14:42:09 what for? Jul 14 14:46:16 curiousity killed the cat Jul 14 14:48:05 the latest version of qt seems 4.7 Jul 14 14:48:17 there is possibilty to reach 4.8? Jul 14 14:58:44 what for? Jul 14 15:00:48 i've a n900 with maemo Jul 14 15:01:10 yeah, that's what most people here have Jul 14 15:01:38 ok Jul 14 15:01:55 i want to know even for future Jul 14 15:02:09 when qt5 is ready Jul 14 15:02:15 how procede? Jul 14 15:02:27 4.8 and qt5 only works on Mer/Nemo for n900 Jul 14 15:03:22 mmm Jul 14 15:03:45 mer and nemo are different from maemo 5? Jul 14 15:04:41 very Jul 14 15:04:51 in which? Jul 14 15:27:12 the first two are meego, the second is maemo Jul 14 15:28:11 er third Jul 14 15:28:13 my math is good Jul 14 15:31:36 :) Jul 14 15:44:16 yo Jul 14 15:47:13 hmm, i think i finally got to bork my N900 Jul 14 15:47:41 what is the /media/mmc1 mount point for? Jul 14 15:49:14 sdcard? Jul 14 15:49:20 i mistakenly did a 'mount /dev/mmcblk1p1 /media/mmc1' and copied about 1.5GB of stuff to /media/mmc1 Jul 14 15:49:28 lol Jul 14 15:50:37 now, i notice rootfs is 100% full (my N900 still boots fine, lol). and there are no other partitions in df -h output, eg, /home, /home/user/MyDocs are missing Jul 14 15:52:22 what's interesting is that all these partitions are still intact. i suppose they got mounted under / Jul 14 15:52:28 as for why, i don't know Jul 14 15:53:10 if emmc can't be mounted, a new blank /home is created on rootfs Jul 14 15:55:18 makes sense. but how is it that i can still access everything, eg Documents dir. Jul 14 15:55:45 and i didn't mount anything manually now Jul 14 15:58:27 hmm, so let me check syslog for an answer Jul 14 16:15:58 nothing informative there Jul 14 16:16:45 the generated fstab is fine, and has all the partitions Jul 14 16:17:03 * sicelo|n900 considers a reboot + fsck of emmc Jul 14 16:18:30 hm, what generates /etc/fstab at boot, and how can you change it? Jul 14 16:19:43 some scripts.. Jul 14 16:19:55 fstab is generated after things have been mounted Jul 14 16:20:04 ...i don't even Jul 14 16:20:07 so its contents is rather pointless Jul 14 16:20:11 iirc Jul 14 16:20:16 i could be wrong :P Jul 14 16:21:27 yeah, it is autogenerated. so you mean it will generate crap, lol ;) Jul 14 16:26:34 i'm relieved that at least it boots Jul 14 16:27:33 jeez! now absolutely no output for mount and df commands Jul 14 16:31:51 eew, and Pali isn't here Jul 14 16:38:22 well, /proc/mounts seems to show i have a healthy system Jul 14 16:38:24 hmm Jul 14 16:39:24 question is.. why would 'mount' and 'df' not show that Jul 14 16:39:47 and N900 still behaves as normal when i connect to usb Jul 14 16:40:50 can one of you please paste me their /proc/mounts Jul 14 16:44:14 Sicelo: because of mtab silliness? Jul 14 16:44:57 my /etc/mtab is blank. what about yours? Jul 14 16:45:39 1000+ bytes Jul 14 16:46:05 please paste Jul 14 16:46:07 :- Jul 14 16:51:41 http://pastebin.com/fzcJaZ7V Jul 14 16:52:19 thanks Jul 14 16:52:43 Sicelo: just ln -s /proc/mounts /etc/mtab Jul 14 16:52:58 btw, i have NO IDEA of what's going to happen if you do that Jul 14 16:53:05 heh Jul 14 16:53:46 i mean, it *should* work Jul 14 16:53:47 but it mightn't Jul 14 16:54:08 well, i noticed that mtab doesn't have the 'continue' options that are in /proc/mounts Jul 14 17:01:21 kerio, ShadowJK, please check what's in /media/mmc1 Jul 14 17:01:38 ...none of your business Jul 14 17:01:58 it's my sd Jul 14 17:02:38 ah, thaks. Jul 14 17:03:02 don't mention it :) Jul 14 17:03:37 * Sicelo does a nice 'rm -rf /media/mmc1' Jul 14 17:04:02 reboot time, again Jul 14 17:11:30 great! this the the most useful 'rm -rf' out there Jul 14 17:11:31 all is fine now Jul 14 17:12:20 now i know what happened. my earlier mounting mistake caused me to now i know what happened. Jul 14 17:13:35 /media/mmc1 was not mounted, but i copied to it about 1GB of stuff (actually an image of debian), in effect, saving to / Jul 14 17:14:30 now, 71% free on rootfs Jul 14 17:16:44 Sicelo: ubifs is weird as shit Jul 14 17:17:27 agreed. Jul 14 17:47:21 Sicelo: LOL Jul 14 17:47:29 age old pun Jul 14 17:48:27 [general advice] NEVER write anything to MyDocs or uSD from device, as ölong as those are actually mass-storage mounted Jul 14 17:49:02 DocScrutinizer05: wat Jul 14 17:49:21 btw I wonder if you could umount MyDocs; chmod a-w MyDocs; mount MyDocs Jul 14 17:49:45 oic Jul 14 18:03:08 i actually had done a mount /dev/mmcblk1p1 /media/mmc1 but i guess it didn't succeed. either way, it was irresponsible of me not to look carefully at output of that mount Jul 14 18:30:33 df -h /media/mmc1 Jul 14 18:32:55 run for shelter! spam homing in Jul 14 18:32:59 t900:~# df -h /media/mmc1 Jul 14 18:33:01 Filesystem Size Used Available Use% Mounted on Jul 14 18:33:02 /dev/mmcblk1p1 483.6M 219.8M 263.8M 45% /media/mmc1 Jul 14 18:33:04 t900:~# df -h /media/mmc1 Jul 14 18:33:05 Filesystem Size Used Available Use% Mounted on Jul 14 18:33:07 rootfs 227.9M 218.5M 5.2M 98% / Jul 14 18:33:08 t900:~# ls -l /media/mmc1 Jul 14 18:33:10 -rw-r--r-- 1 root root 15 Apr 9 16:17 this-will-vanish-but-eat-your-stoage-nevertheless Jul 14 18:33:11 t900:~# Jul 14 18:33:16 ;-) Jul 14 18:33:17 SPAM Jul 14 18:33:18 :P Jul 14 18:35:43 >:-( unix-file-permissions Jul 14 18:36:15 t900:~# stat /media/mmc1/ -> Access: (0555/dr-xr-xr-x) Uid: ( 0/ root) Gid: ( 0/ root) Jul 14 18:38:17 still writes to /media/mmc1/foo happily Jul 14 18:38:41 even creates /media/mmc1/foo happily Jul 14 18:39:08 WTF root can't forbid root to write to any dir? Jul 14 18:40:32 whatever, make sure you write to mounted media, not to the mountpoint dir! Jul 14 18:44:23 yeah Jul 14 18:44:51 * DocScrutinizer05 idly wonders if friggin ke-recv (or whatever the "automounter" on the particular distro) couldn't also _clean_ _up_ on umounting a volume - I.I. delete the mountpoint dir Jul 14 18:44:58 at that time there was zero percent free space. not even to edit the mtab file Jul 14 18:45:02 mount as ro Jul 14 18:45:30 jacekowski: you missed the point Jul 14 18:45:34 ahm Jul 14 18:45:45 root is supposed to be able to do everything Jul 14 18:45:45 I wnna mount rw on a ro-dir Jul 14 18:46:02 including writing to unwriteable file Jul 14 18:46:06 and folders Jul 14 18:46:39 which actually seems no problem, modulo the ro-dir doesn't guarantee nobody writes to that rootfs-dir-mountpoint Jul 14 18:47:33 if ke-recv would nuke the mountpoint dir in /media on umounting sdcard, you had a trouble less as user Jul 14 18:50:46 jacekowski: I'd expect root to be alowed everything on chmod(1,2) but not on fopen(2) Jul 14 18:51:26 well, ambivalent Jul 14 18:52:37 probably I meant: root supposed to obey unset write perm bit of owner when owner is root Jul 14 18:54:06 sure root should override all other user's perms, but when a file is root root 444 then I'd not expect root to be allowed to write to it Jul 14 18:54:34 blah Jul 14 18:54:39 try it. Jul 14 18:54:49 my tyan board has bad caps too Jul 14 18:54:55 no wonder it acts flakey Jul 14 18:54:58 sigh Jul 14 18:54:59 hmm, then that file would almost be a ghost, i suppose. because no one has higher privileges than root, right? Jul 14 18:54:59 NIN101: I did :-( Jul 14 18:55:28 Sicelo: see my comment about chattr/chmod above Jul 14 18:55:48 nobody except root could set a file to root root 444 Jul 14 18:56:03 so root could also set it to root root 777 Jul 14 18:56:32 but as long as it's 444, root shouldn't be allowed to cheat on own decision Jul 14 18:58:38 I.E a file is owned by root. Root says: chmod u-w. So after that user is supposed to not write to that file, no? Jul 14 18:59:12 even when that user is root Jul 14 18:59:50 IOW to root, root is a normal user Jul 14 19:00:17 meh Jul 14 19:00:26 I won't fix unix, I guess ;-P Jul 14 19:02:12 anyway for a file owner=root the first 4 bit of perms are useless (actually 2nd 4 bit) Jul 14 19:02:45 except for sticky bit, so in fact the rwx 3 bits are useless Jul 14 19:07:00 * DocScrutinizer05 feels like it's time for a re-read of manpages and docs about unix file permissions and ACLs Jul 14 19:58:58 Hi. I have promised a friend to setup his newlybrought used N900. Without checking for updates or SSU-version I just installed CSSU which failed. The SSU version is 1.1.1. Should I try to updates packages via apt or just reflash? Jul 14 20:00:32 I don't know about ssu, but normally you never want to use apt Jul 14 20:00:35 ~cssu Jul 14 20:00:35 [cssu] http://wiki.maemo.org/Community_SSU Jul 14 20:01:12 If you haven't done yet (unlikely): Upgrade to at least PR1.3, Nokia's last official major Maemo 5 update. PR1.3.1 is fine as well. If you're unsure if you have, Settings > About product should show Version beginning with 20.2010.36(PR1.3) or 21.2011.38(PR1.3.1). Jul 14 20:05:27 I've 3.2010.02-8 which seems to be PR1.1.1. In the update manager I have an updated called Maemo 5 Community SSU Package 21.2011.38-1Smaemo3 which fails. Jul 14 20:07:44 Well. Tries to upgrade though apt and see for fun what happens. Just reflash it if it goes wrong. Jul 14 20:08:21 apt isn't even designed to work Jul 14 20:10:55 Hmm. Maybe I should try to uninstall CSSU and then upgrade to PR1.3(.1). Jul 14 20:12:22 Well. I'll try "apt-get install mp-fremantle-community-pr". It was written at the CSSU faq Jul 14 20:14:31 if you still have PR1.1 you should update as soon as possible :) Jul 14 20:14:56 but reading your initial question Jul 14 20:15:13 you should simply flash PR1.3.1 (latest) and go ahead with CSSU Jul 14 20:16:51 Off-topic. You installed sudser via root though SSH and got this message "Please read and close the popup dialog". LOL. Jul 14 20:17:02 lol Jul 14 20:18:38 Running apt-get install mp-fremantle-community-pr and x-ing my fingers. Jul 14 20:24:52 Yes. My device now reboots in loop. Think it's time for reflashing. Jul 14 20:28:16 louisdk: uninstalling CSSU is not supported. You should flash right away Jul 14 20:31:32 louisdk: instructions on http://wiki.maemo.org/Community_SSU are no kind suggestions, and are meant to get followed sequentially from start to end Jul 14 20:32:35 DocScrutinizer05, I'm already on it. Have setup/flashed a frew N900's over the years ;) Jul 14 20:34:02 DocScrutinizer05, I mean I've tried some things at this page: http://wiki.maemo.org/Community_SSU/Installation_FAQ but now I'm reflashing Jul 14 20:36:17 installation-faq are not a generic installation instruction - that's http://wiki.maemo.org/Community_SSU Jul 14 20:36:41 http://wiki.maemo.org/Community_SSU#Installation actually Jul 14 20:37:12 only if you got >>Problems? Check the Installation FAQ<< Jul 14 20:38:37 so obviously in the Installation FAQ there won't be any instructions or problem workarounds that are already clearly explained in main page http://wiki.maemo.org/Community_SSU#Installation Jul 14 20:47:58 louisdk: simply make sure you flash the latest PR1.3.1 Jul 14 20:50:46 and while installing CSSU you need some patience as it's sometimes not clear if/what it's doing Jul 14 21:01:32 ( Off-topic. You installed sudser via root though SSH and got this message "Please read and close the popup dialog". LOL.) a) who's "you"? b) wtf is sudser? c) "though SSH" sounds like insttalation by apt-get, so the msg is more than sane as sudser then obviosuly is one of the pkgs that aren't meant to get installed via apt-get since they open a requester on display:0.0, a "normal" thing for pkg install in HAM Jul 14 21:01:44 NIN101: ping Jul 14 21:04:55 Sicelo: pong Jul 14 21:06:38 hmm, now i think i pinged you for nothing. i needed a reminder for rescueOS, namely, how do i login to the system via usb after i have usb-networking up Jul 14 21:06:50 i just remembered that i have to use telnet Jul 14 21:07:10 okay :-) Jul 14 21:07:37 this evening at the restaurant i learned the hard way that backupmenu2 doesn't backup the kernel and that not having the correct kernel modules is a Bad Idea Jul 14 21:07:52 :P Jul 14 21:08:02 it was especially sad because i really wanted to "audit" a wep wifi network that was there Jul 14 21:08:05 this means today is a bad day for N900. Jul 14 21:08:10 indeed Jul 14 21:08:20 i've borked mine the second time already :P Jul 14 21:08:31 that's why i'm on rescueOS as we speak Jul 14 21:08:48 right now i have the kp50 modules and a kp51 kernel Jul 14 21:08:57 i suppose i'll just flash the stock nokia kernel Jul 14 21:09:21 (i made a full backup right before the problem, and it has kp51 modules and omap1 modules) Jul 14 21:10:25 Sicelo: just make backupmenu work regardless of stuff Jul 14 21:10:39 DocScrutinizer05: Sudser allows user to run root commands while typing sudo in front of them. This oage does not say using apt is a bad thing: http://wiki.maemo.org/Terminal#Debian_package_management expect for autoremove which will remove things you want to keep. Jul 14 21:10:44 where's bootmenu installed? Jul 14 21:10:57 louisdk: [citation needed], actually Jul 14 21:11:00 autoremove works perfectly here Jul 14 21:11:03 no "lost" packages Jul 14 21:11:05 (latest cssu-t) Jul 14 21:11:08 ... Jul 14 21:11:25 fine down, not tomorrow. Jul 14 21:11:27 *fine today Jul 14 21:12:02 DocScrutinizer05: package install scripts that open windows is a GIANT wtf Jul 14 21:12:04 kerio: Thx for info Jul 14 21:12:15 HAM should properly support debconf, and it should use *that* instead Jul 14 21:12:26 as always, fuckin' nokia Jul 14 21:12:43 kerio: +1 Jul 14 21:12:52 kerio: ever tried rescueOS? Jul 14 21:12:56 Sicelo: nope Jul 14 21:13:08 where is it installed? sd? Jul 14 21:14:08 nowhere, it's an initrd. Jul 14 21:15:27 hmm Jul 14 21:15:34 does backupmenu work with the default kernel Jul 14 21:15:35 ? Jul 14 21:16:18 NIN101: rescueOS isn't able to mount MyDocs for me, is it my system only? Jul 14 21:16:41 no idea. Jul 14 21:17:20 ~$ mount /dev/mmcblk1p1 /mnt/free1/ Jul 14 21:17:21 mount: mounting /dev/mmcblk1p1 on /mnt/free1/ failed: Input/output error Jul 14 21:17:25 kerio: suggest a better 3way! Jul 14 21:17:42 what does dmesg say? Jul 14 21:18:15 also I have no clue, maybe you nedd to say -t vfat or whatever. Jul 14 21:18:25 i support any and every 3way Jul 14 21:19:17 pling Jul 14 21:19:28 lol. yeah, needed the vfat thingy Jul 14 21:19:41 yeah the busybox mount has no autodetect, or at least wasn't compiled Jul 14 21:21:10 anyway, have fun with your workshop here Jul 14 21:21:54 DocScrutinizer05: ok, wat do Jul 14 21:22:04 if you finally find out you can actually cook eggs in your dishwasher, please don't let me know Jul 14 21:22:26 DocScrutinizer05: installed system with kp50 modules, kp51 flashed, i want to restore a backup that was made with kp51 installed Jul 14 21:22:35 does backupmenu work with the stock kernel? Jul 14 21:22:39 or does it need a framebuffer? Jul 14 21:22:54 what should i do? Jul 14 21:23:40 please could anyone paste `mount` result Jul 14 21:23:52 kerio: "sorry, your draft limit for today been exceeded. Contact your bank consultant" Jul 14 21:24:12 really helpful, thanks Jul 14 21:25:15 sorry if it's less helpful than my warning about your token-jar depleting by asking nonsense questions Jul 14 21:25:30 specifically, i would like to know how /opt is actually mounted Jul 14 21:25:42 as well as MyDocs Jul 14 21:25:51 /opt/ is a bind mount Jul 14 21:26:13 something like mount --bind /home/opt/ /opt/ Jul 14 21:27:19 thanks Jul 14 21:30:11 restoring Jul 14 21:31:32 Sicelo: does restoreOS have a web page of some sorts? Jul 14 21:32:25 http://206.253.166.96/N900/rescueOS/ Jul 14 21:34:57 NIN101: neat! Jul 14 21:35:07 nothing that runs on the n900 itself though Jul 14 21:35:16 unless you're willing to kinda dick around with uboot, i s'pose Jul 14 21:35:41 ? Jul 14 21:35:59 hmm, i think i still need a full 'mount' output Jul 14 21:36:17 NIN101: it requires a computer to work Jul 14 21:36:19 (rightfully so) Jul 14 21:36:36 kerio: it's the best thing since sliced bread, imo Jul 14 21:36:37 :P Jul 14 21:36:44 no, that's backupmenu Jul 14 21:37:14 Sicelo Jul 14 21:37:15 hmm, i just found a (huge?) problem with cssu-thumb Jul 14 21:37:24 this is what maemo does during boot: cat /proc/mounts > /etc/mtab Jul 14 21:37:37 meh, i suppose it's not a big deal if you can just flash the correct kernel with flasher Jul 14 21:37:42 nvm Jul 14 21:38:06 could give you my mount output, it is full of dm-crypt stuff etc. so wouldn't probably help Jul 14 21:38:16 :P Jul 14 21:38:23 needed a basic one Jul 14 21:38:25 Sicelo: /home/opt on /opt type none (bind) Jul 14 21:39:17 and MyDocs? Jul 14 21:40:05 /dev/mmcblk0p1 on /home/user/MyDocs type vfat (rw,noauto,nodev,noexec,nosuid,noatime,nodiratime,utf8,uid=29999,shortname=mixed,dmask=000,fmask=0133,rodir) Jul 14 21:41:18 Sicelo: why would you need to mount /opt on a rescue system? Jul 14 21:41:59 thanks. well, i don't need /opt. i need MyDocs to be mounted alongside maemo rootfs though Jul 14 21:42:16 to run python scripts, of course Jul 14 21:42:40 about that... why does python on n900 only have about half of the python stdlib? :/ Jul 14 21:54:19 A quick question, I tried to google for it but no luck. Is it really normal that apt-get will say that some packages cannot be authenticated when trying to install scratchbox? (I'm trying to follow the instructions on http://wiki.maemo.org/Documentation/Maemo5_Final_Installation for a text based installation Jul 14 21:54:23 (The GUI based installation didn't work.) Jul 14 21:54:57 yes Jul 14 21:55:12 as you don't have signing key installed into your keyring Jul 14 22:03:18 Hmm, any idea where the signing key can be found? Jul 14 22:03:33 on keyserver Jul 14 22:07:17 oh wow, just got a segfault in busybox Jul 14 22:07:23 with rm Jul 14 22:07:25 i don't even Jul 14 22:07:56 morning, guys Jul 14 22:09:16 ShadowJK, I've tested charging with 0x71... BTW, it turned out, that whole "life" I was using version that charges via 0x50, = 1050 mA, without any ill effects :P Jul 14 22:09:42 BTW, are there any other improvements in latter versions of charge.sh? or was it just tuning this value? Jul 14 22:10:26 anyway, charging 3A battery with 0x71 works fine, peak temperature 46 C degrees Jul 14 22:11:30 hoever, for some reason, I got feeling that it have cap to 1050 anyway. Even with everything disabled and screen off, I never saw more than ~1040 current reported Jul 14 22:12:24 as for cable, it's very good one (had to pay 2 usd for it, after all :P ) - no problem with putting 3A through it, @5V, without negative temp effects Jul 14 22:13:16 I'm also perfectly sure, that it touches port in N900 okey'ish, it doesn't look like problem with too much current for small pins Jul 14 22:13:20 any idea? Jul 14 22:13:42 BTW, thanks ShadowJK and DocScrutinizer for explaining me, patiently, how it works ;) Jul 14 22:14:17 probably would only pull 1.2A when empty Jul 14 22:14:24 it seems, that DocScrutinizer was right about heat being generated by single-cell itself, if threaten with 1050, not charging chip's step-down converter Jul 14 22:14:45 ShadowJK, tested it when almost empty, charging status reported 0x10 Jul 14 22:15:26 400 mAh left out of 3000 mAh (ok, maybe it isn't almost empty, after all, but shouldn't pull whole current anyway?) Jul 14 22:16:48 also, there is no way to force more than 1250, by any humanly means? I wouldn't mind charging this 3A battery with 2A, it would be still less than 950 mA vs standard battery Jul 14 22:17:05 (my travel charger supply even little more than 2A) Jul 14 22:18:11 bq24150 can't do more than 1250 Jul 14 22:18:23 not to mention, that in home, as part of my DIY "ultimate docking station for everything" :P I have modified ATX power supply, that can push 40A @ 5V - wouldn't mind trying to charge prototype of 6,8A battery with, lets say, 4A, observing temp Jul 14 22:18:27 pity. Jul 14 22:18:41 with 1250 being absolute maximum Jul 14 22:18:51 so you are risking killing bq1250 Jul 14 22:18:56 BTW, 1250 isn't stressing step-down converter? Jul 14 22:18:57 and ending up with no charging at all Jul 14 22:19:05 hah, it was my next question :) Jul 14 22:19:45 with no ill effects re chip's reported temperature, should I still be afraid of killing it? Jul 14 22:20:01 hmm, it looks like bq24150 has thermal throttling builtin Jul 14 22:20:10 Of course, it would be much more sane to do 1150 or even 1050, in case of *real* risk Jul 14 22:20:21 throotling, or just cut-out? Jul 14 22:20:29 throttling and trip Jul 14 22:20:34 throttling from 120C Jul 14 22:20:37 and trip at 165C Jul 14 22:20:52 http://maemo.jacekowski.org/docs/bq24150.pdf Jul 14 22:20:53 maybe throotling is reason, why I feel caped to 1050, after all? but throotling on 46 C degrees? 0_o Jul 14 22:21:04 I see Jul 14 22:21:16 i think it's measuring current going to battery Jul 14 22:21:23 so 200mA is going to rest of phone Jul 14 22:21:31 but chip itself can only deliver 1250 Jul 14 22:21:36 thanks a lot. As said, peak reported temp is 46C. Jul 14 22:21:49 negative, I did it with disabled activity, nothing at all enabled Jul 14 22:22:00 power draw at this state is 4-8 mA, in my case Jul 14 22:22:21 charging with BME result in real ~940 being reported Jul 14 22:22:41 and it's caped to 950 Jul 14 22:23:03 so I would expect ~1240 or ~1200at least, when charging with max Jul 14 22:23:08 that puzzles me Jul 14 22:23:16 Hmm, now the GUI based installer seems to work (although I note that the GUI based installer also complains about unauthenticated packages) Jul 14 22:23:18 well, maybe external components are not big enough to deliver that much Jul 14 22:23:32 they're, see backscroll :) Jul 14 22:23:47 tested that before attempting experiment Jul 14 22:24:08 PS doesn't have problem with 2A @ 5V, cable delivers 3A without problem Jul 14 22:24:26 I'm not insane, wouldn't risk burning anything Jul 14 22:24:41 not those components Jul 14 22:24:55 N900's ones? Jul 14 22:25:02 there is 1uH inductor in there Jul 14 22:25:15 Nope, the GUI based installer didn't work this time either, oh well, seems like I shouldn't do any maemo development for the moment Jul 14 22:25:16 it can only store so much energy Jul 14 22:25:47 so at whatever frequency bq24150 is running at it may be capable of only delivering 1000mA or whatever Jul 14 22:26:02 buck converters are really complex things Jul 14 22:26:03 I see. So, considering possibility of such limitation existing, should it be safe to left it at "max", without much risk of damaging anything in N900? Jul 14 22:26:17 or is it stressed too much anyway? Jul 14 22:26:27 well, somebody designed it for specific values Jul 14 22:26:36 working outside of those, you don't know what to expect Jul 14 22:26:49 increasing current means more heat disipated in all compoenents Jul 14 22:27:00 not only bq24150 but it's inductor as well Jul 14 22:27:06 and that has no temperature measurment on it Jul 14 22:27:10 sure, that's why I'm thinking if, for example, such inductor could be stressed working at max, from general EE point of view Jul 14 22:27:26 as well as battery Jul 14 22:27:43 charging it at higher current isn't the best thing for battery either Jul 14 22:27:51 lower currents are better for batteries Jul 14 22:28:07 as for battery itsewlf, I'm not worried about it, it's true beast. 1A is 0.3C anyway Jul 14 22:28:15 doesn't matter Jul 14 22:28:20 hm? Jul 14 22:28:26 less current is better for battery regardless of it's size Jul 14 22:29:07 sure, but isn't it like hardly-measurable difference, if it doesn't exceed certain fraction of C? Jul 14 22:29:17 or it doesn't matter at all? Jul 14 22:29:27 depends on many other factors Jul 14 22:29:56 but generally it's less important on low currents like that Jul 14 22:30:39 well, it seems that I have to be guinella pig again :) Jul 14 22:30:43 but thing is, lithium batteries tend to explode Jul 14 22:30:57 and i would prefer to be nice to it Jul 14 22:31:00 vent with flame, if anything :P (pun intended) Jul 14 22:31:51 sure, sure, I've put independent thermal sensor on it. BTW, it shows same value as chip's internal one Jul 14 22:32:25 reinforcing DocScrutinizer observation, that every bit of real heat during charging is generated inside cell only Jul 14 22:32:50 peaked @ 46 C on both sides, doesn't look like it's going to get higher Jul 14 22:33:28 funny, as during extensive discharges (on device) I've observed more than 60 C many times Jul 14 22:34:23 BTW, some times I'm pulling out of battery much more, (inside device, also) than, as it seems, I'm able to put inside, via N900's charging chip Jul 14 22:34:30 What's voltage measurement like when charging at 1.2A? Jul 14 22:34:48 btw temperature is reported by bq27200, not b24150 Jul 14 22:34:56 good point Jul 14 22:35:29 having in mind that it seems bo be charging @ 1050 anyway... Do You mean battery reported voltage? Jul 14 22:36:03 ...PS drop, or something else Jul 14 22:36:06 ? Jul 14 22:36:33 voltage reported by the charge script Jul 14 22:36:51 3955 mV ATM Jul 14 22:36:56 current? Jul 14 22:37:00 with 33 % state of charge Jul 14 22:37:10 current in screen off/idle etc if your xterm scrolls up Jul 14 22:37:25 when I'm using device, 860 mA, so it's about 1050 mA real Jul 14 22:37:39 on idle, 1040 and doesn't want go higher Jul 14 22:38:16 measuring via bq27200.sh it's ~25 mA less, due to additional power being pulled Jul 14 22:38:38 exact same datasource for the two Jul 14 22:38:42 (as by using, I mean wifi on and Xchat and screen on on moderate brightness) Jul 14 22:39:04 I estimate voltage at charging chip output is about 4150mV Jul 14 22:39:06 ShadowJK, not likely, bq27200.sh going in background in loop eat some power Jul 14 22:39:12 so it's probably limited there Jul 14 22:39:16 I see Jul 14 22:39:20 So is charge script :P Jul 14 22:39:26 true:P Jul 14 22:39:33 so it's probably normal fluctuation Jul 14 22:40:12 ok, so if voltage output from chip is limited, most important question - risk of it being stressed too much is high? Jul 14 22:40:12 The .13 Ohm resistance I've observed seems not entirely linear wrt current.. Jul 14 22:40:23 hm Jul 14 22:40:43 Well I'd look at what jacekowski wrote Jul 14 22:41:18 re "quite impossible to estimate it, as they have designed it for certain usage"? Jul 14 22:42:22 btw, is b24150 accesible on motherboard, after removing shield plates? would like to measure it's temp Jul 14 22:43:02 as You've made good point about temperature being measured on bq27200 Jul 14 22:43:58 hm, current raised to average 913 mA during usage, wtf? Jul 14 22:44:15 at 38% soc Jul 14 22:44:24 Estel: Am I correct in thinking that this discussion is related to the custom case you are building for the n900? Jul 14 22:44:52 TamyrRRRR, thanks for asking :) well, only slightly - I'm checking possibilities of faster native charging Jul 14 22:45:26 as it is possible, that custom case will have possibility for housing batteries as big as 6800 mAh Jul 14 22:45:46 (and even current batteries fitting in mugen cover are 3A) Jul 14 22:46:08 I hopes for at least 1.5A charging, but it seems 1050 mA is absolute maximum Jul 14 22:46:15 effective, at least Jul 14 22:46:33 Estel_, http://enivax.net/jk/forestel/charge21.sh.txt Jul 14 22:46:38 as for custom body, I've recorded some (nothing special) videos of cnc machine working on wood Jul 14 22:46:42 I'd lake paste of charging with that script Jul 14 22:47:00 ShadowJK, ok, getting it Jul 14 22:47:27 I copied and edited on webserver, untested on actual devices ;P Jul 14 22:47:43 wasn't available on server. Typo in urkl? Jul 14 22:47:57 ok, ok, I'll check it before running :P Jul 14 22:48:26 Estel: Neat. Good luck with both the case and the faster charging :) Jul 14 22:48:42 http://enivax.net/jk/n900/forestel/charge21.sh.txt Jul 14 22:48:46 thanks a lot :) Jul 14 22:48:48 ah, n900 bit was missing Jul 14 22:48:54 ShadowJK, figured it out at same moment ;) Jul 14 22:50:33 I'm ash-compliant only, doesn't matter? Jul 14 22:50:50 It does not Jul 14 22:50:52 nah, original is also for bash as per shebang Jul 14 22:50:56 yep Jul 14 22:51:27 I wrote this on device, with sdk tools nano (which crashes randomly), and on device busybox... Jul 14 22:51:32 originally, that is :P Jul 14 22:52:56 Also at the time I was under the impression and belief that neither bash nor ash supports binary arithmetic.. that's why all bit manipulation is in integer arithmetic, lol Jul 14 22:53:01 :) it's running, what to report? Jul 14 22:53:33 * Sicelo finally gives up on u-boot. will simply tag it 'incompatible' with this particular N900 Jul 14 22:53:38 ShadowJK: what is your USB-curlim in charge.sh? Jul 14 22:53:58 Load pastebin in web browser on N900. Switch to xterm. Press enter to make a mark. Lock screen, wait a minute, unlock screen, copy/paste xterm contents to pastebin :P Jul 14 22:54:16 DocScrutinizer, I dont remember charge.sh anymore, but in 21 it's unlimited Jul 14 22:54:16 since I-max-bat = I-max-usb / 5v * U-bat Jul 14 22:54:58 ShadowJK, ok. Am I mistaked, or does this version of charge separate current used by system, somehow? Jul 14 22:55:15 I thought you already had that version... Jul 14 22:55:39 But yes, it tries to calculate that. I just adjusted the parameters to assume 1250mA Jul 14 22:55:40 Hi. What's the name of the N900 backup-folder? Jul 14 22:55:46 Sicelo: might actually be the truth. Your NNAD might have too many bad blocks so uBoot+kernel doesn't fit into mtd Jul 14 22:55:48 ShadowJK, I told You You had some oldie, that charged with 1050 max as per default (0x50) Jul 14 22:55:53 ~/.backups or? Jul 14 22:55:57 "backups" Jul 14 22:56:01 on MyDocs or mmc1 Jul 14 22:56:13 ShadowJK, nice. Will paste zerobin in a minute or so, brb Jul 14 22:56:28 DocScrutinizer05: i hope it's not that. gawd :P Jul 14 22:58:45 ShadowJK, thanks Jul 14 22:58:50 ShadowJK: I think all that calculation of "true battery current" is kinda guesstimate Jul 14 22:58:55 Of course. Jul 14 22:59:55 *only* proper way to implement that: temprarily stop charge, read current flow _from_ battery as that's what system is jist ATM consuming Jul 14 23:01:07 maybe even use those real values to calibrate your guesswork-parameters Jul 14 23:01:20 It's far off when the current is fluctuating, but when I plotted against "known" consumption whilst charging, it was +-20mA at the start of the charge, and then drifted off to being 50% inaccurate near end of charge.. or other way around, I forget :-) Jul 14 23:02:46 well, you could at least try to either pick a period with low fluctuations, or calm the system ;-) Jul 14 23:03:16 Part of it is assumption of voltage drop across the circuit board.. Everytime an N900 has broken for me, I've reran calibration of this loss estimate on the new spare N900, which has as zero load as possible, and each time I've adjusted it upwards with a couple of mR :-) Jul 14 23:04:02 iirc I started somewhere at .125 ... .128, and now I'm at .132 Jul 14 23:04:14 statistic Jul 14 23:04:26 static? Jul 14 23:04:39 noise! Jul 14 23:04:48 However, the second and third assumption is that the charger chip sense resistor is .068, and that the bq27200 sense resistor is .021 :) Jul 14 23:05:29 and you also assume bq24150 has pretty exact voltage/current limit Jul 14 23:05:30 .021 obtained from the MeeGo boys with their test jigs and lab psus Jul 14 23:05:36 yes, that too Jul 14 23:06:27 well, .021 is correct for sure, since we get proper capacity readings with it Jul 14 23:06:43 errr Jul 14 23:06:55 do we? Jul 14 23:07:02 yep we do Jul 14 23:07:41 multiplier for charge is a function of RS Jul 14 23:08:04 s/charge/capacity_now/. Jul 14 23:08:05 There's probably another assumption one could make, assume offline mode idle drain is constant, and then gather data for voltage and current when charging, and solve for the 3 resistances... Jul 14 23:09:06 with bq27200 sense R as 21mR I was getting <1100mAh on my old nokia BL-5J (was 1250-ish new), and getting 1400 on dealextreme "Japod" BL-5J Jul 14 23:09:47 I got 1320..1350 on new bl-5j Jul 14 23:10:26 The japod is magic though, it maintains low Z for almost forever Jul 14 23:13:33 http://enivax.net/jk/n900/volt.png I wish I could figure out a way to represent what I mean better than this, but move your pointer to the start of the "Cliff", where the voltage drops like it fell off of a cliff, notice on the Y axis at what voltage the cliff happens Jul 14 23:14:14 It's significantly higher for the japod compared to the rest (let's ignore the brick-size mugen) Jul 14 23:14:50 So when b27200 says 1400mAh capacity for japod, the thing where bq assumes 3248mV is at 6% battery cap comes into play :) Jul 14 23:18:30 anyway, estel's concern about choke in stepdown converter bq24150 is valid. Depending on job sourcing did, it might run into saturation with 1200+mA and that's not really a good thing Jul 14 23:19:11 on GTA02 I even heard of chockes that desoldered themselves (or was it actually N900?) Jul 14 23:19:30 GTA02 lol. Grand Theft Auto Jul 14 23:19:54 Gotta Think Aboutit Jul 14 23:20:05 Zero Two Jul 14 23:20:26 scratch the zero Jul 14 23:20:31 I'm dizzy Jul 14 23:20:43 increase your font size Jul 14 23:21:27 won't help Jul 14 23:21:45 i'm usually not looking at what I type Jul 14 23:21:55 I look at the kbd Jul 14 23:22:11 you suck then. I can't look at my keyboard Jul 14 23:22:19 or at the text I'm answering Jul 14 23:22:23 all the keys wore out Jul 14 23:22:39 (Ok, i'm being dramatic. But it sure doesn't look like new.) Jul 14 23:22:52 when the keyboard led strikes in, it looks really ugly Jul 14 23:23:31 backspace key is a shame. It has no symbol in it, you can actually see the small dots of light in there. Jul 14 23:23:34 Where did estel go Jul 14 23:23:35 apropos, this afternoon I found N950-kbd-backlight is such a dim crap they could've saved it all together Jul 14 23:23:41 nfc Jul 14 23:24:13 i would love to get a spare keyboard but you know how it is, chinese keyboard clones suck Jul 14 23:25:00 DocScrutinizer05, I stopped using 1200mA charge rate after I went on a road trip with N900+2400mAh battery, my N900 streaming music over 3g, sygic mobile maps running, fm transmitter activated. It was nice, the music faded out and the voice nav told me where to go, music faded back in. When I arrived at destination my N900 was like 55C :s Jul 14 23:25:37 And the music/voicenav came through car speakers, not N900 speakers :D Jul 14 23:27:47 55ºC only? Ooh please, i forgot my N900 in my oven Jul 14 23:29:56 N900 sitting in 55C ambient isn't as bad as N900 disspating enough energy to heat itself to 55C, because in the latter case there'll be very small components that are considerably hotter than 55C :P Jul 14 23:30:56 and therefore frying eggs Jul 14 23:31:45 one day i should try frying eggs with my passive 8600GT gpu that is hanging somewhere Jul 14 23:32:01 that thing would reach 115ºC easily. the site says 105ºC is the limit Jul 14 23:32:32 ShadowJK, sorry for delay - my son waked up, and I needed to sing some lullababies :) Jul 14 23:32:37 here is the link Jul 14 23:32:56 http://sebsauvage.net/paste/?8fa5034cd4d8227f#gjq0EONBU95Mgx4blSA6bsoXO4hJwxbBP+7RK9Ehnss= Jul 14 23:34:18 it's funny, as now, when device is in use, it's rather: Jul 14 23:34:20 Status: 0x10 Mode: CHARGING Full: 0 Wall: 1 Voltage: 4077 NAC: 1786 level: 64 % Rate: 794 System: -137 Ch: 931 Compensated: Jul 14 23:34:33 lol, double negative... Jul 14 23:34:44 how accurate is system load measurement? Jul 14 23:34:53 it's a guess Jul 14 23:35:00 based on? Jul 14 23:35:04 You missed the conversation on it :) Jul 14 23:35:13 and why double negative? Jul 14 23:35:15 ah, I see Jul 14 23:35:21 will check logs Jul 14 23:35:59 It's based on assuming that the charging chip sense resistor is 68mR, that the battery gauge sense resistor is 21mR, that the circuit board losses between charging chip and battery is 132mR, and that the charging chip is able to charge at the full requested rate Jul 14 23:36:24 double negative Isn't Supposed to Happen(TM) :) Jul 14 23:36:37 Means one of the above assumptions are wrong Jul 14 23:36:43 ;) Jul 14 23:37:02 * Estel_ opens own link for investigation Jul 14 23:37:38 a-ha. This thing Jul 14 23:37:40 The expected power use when idle is something like 20mA, so it's off by like 30mA ;-) Jul 14 23:37:57 some lines later, it is single negative only, but generally, yep Jul 14 23:38:39 I wonder how it is in case of higher load, as in: Jul 14 23:38:45 Also, the CH value being less than 1250 means that the charge has reached the CV portion, and that hte voltage at the charging chip is close to 4.2V Jul 14 23:38:56 Status: 0x10 Mode: CHARGING Full: 0 Wall: 1 Voltage: 4077 NAC: 1786 level: 64 % Rate: 794 System: -137 Ch: 931 Compensated: Jul 14 23:39:07 it would be 1100 mA Jul 14 23:39:14 at this point of charge? Jul 14 23:39:28 yea Jul 14 23:40:13 but, but, also, remember that even on 3995 mV and almost empty battery it wasn't 1250, it seems Jul 14 23:40:29 It's guesstimating that the current rate provided by the charger is 932mA, the system uses 137mA, and that leaves the 794mA going into the battery. The 794mA is actually measured Jul 14 23:40:40 ough Jul 14 23:40:47 so I took it wrong Jul 14 23:41:02 931* Jul 14 23:41:32 Double negative system implies that energy is being added through some unknown path :) Jul 14 23:41:49 i though that 931 is charge after gusstimation, with additional 137 for system Jul 14 23:41:53 But that's not possible, so it's really just showing that the assumptions made are not correct. Jul 14 23:41:55 :) Jul 14 23:42:18 Estel_, pong? Jul 14 23:42:33 Also at 64%, it is expected that the battery is no longer accepting full charge rate. Jul 14 23:42:39 Any way to improve guesstimations basing on that data? Jul 14 23:42:55 GeneralAntilles, thanks, but I have no idea what I wanted few days ago :( Jul 14 23:43:05 ShadowJK, yea. Jul 14 23:43:14 Change CR=132 to CR=133 maybe Jul 14 23:43:28 wut? Jul 14 23:43:32 aaa Jul 14 23:43:33 I see Jul 14 23:43:38 Estel_, feel free to memoserv. Jul 14 23:43:50 thanks GeneralAntilles :) Jul 14 23:44:42 ShadowJK, as expected, temperature reported by bq27200 dropped to 42 C (from 46 C), same goes, of course, for cell temp Jul 14 23:45:11 what friggin cell temp? Jul 14 23:45:36 he has external thermometer on his battery Jul 14 23:45:42 duh Jul 14 23:46:06 well, it seems that I will charge few cycles with max current, then post results on TMO's battery's thread, with big red fat warning about concerns made by jacekowski, and that only guinea pigs should try it... Then, continue to use it myself for few months Jul 14 23:46:12 and report back again Jul 14 23:46:44 Estel_: you also should read chanlog for my concerns Jul 14 23:46:48 If all volunteer guinea pigs will be OK so far, we don't have, probably, much to fear ;) Jul 14 23:46:56 which are basically confirming jacekowski's Jul 14 23:47:01 DocScrutinizer, sure, will do it now Jul 14 23:47:08 Estel_, actually I'd be interested in a paste from the start of charge from "empty".. Something like ./charge21.sh | tee chargelog.txt Jul 14 23:47:18 And then paste/attach/email/dcc me the chargelog.txt :) Jul 14 23:47:23 ShadowJK, sure, I'll do it on next cycle Jul 14 23:48:06 how many lines from empty? as much as it will generate until full ;) or interested only in some first ones? Jul 14 23:48:25 Also mention to the guinea pigs that I used 1250mA charge rate on my first N900, and I also used a thermal throttle to limit charging when temperature got warm, and this thermal limit is not present in this script.. And, my first N900 died. Jul 14 23:48:45 ooook... BTW, how it died? Jul 14 23:48:46 actually if choke runs into saturation, temerature increase at start of charging empty battery will be *significantly* higer than on 950mA limited Jul 14 23:48:51 Estel_, if you usually charge it full, then yeah Jul 14 23:48:56 Estel_, sim card missing Jul 14 23:49:03 aka cellular modem fault Jul 14 23:49:32 I'm so worried my n900 will die some day and there will be no replacement :( Jul 14 23:49:48 My second N900 also died that way, and I hadn't abused that one at all.. But let's not mention that to the guinea pigs. They deserve a good warning. Jul 14 23:50:07 I see, that one. Sure, will report it too. BTW, considering all of this, increase from 950 to 1250 seems quite low gain, but I'll probably do it for sake of experiment and curiosity Jul 14 23:50:10 I've used it heavily but apparently gently enough that it still works fine Jul 14 23:50:29 joga, do like me and buy 8 of them Jul 14 23:50:32 ;) Jul 14 23:50:46 Estel_, yes, the reason I asked for log from start of charge, is to see how early the actual charge current drops below 950 with your megabattery :) Jul 14 23:50:53 Estel_ if only I had money for such, there's many available at the local auction site as we speak Jul 14 23:50:59 ShadowJK, sure, I'll try to scary them As much as humanly possible Jul 14 23:51:13 ShadowJK: heating by charging should only be a fraction of what a system at full cpu load creates Jul 14 23:51:41 DocScrutinizer, Well it's 1.2A over .5R Jul 14 23:51:57 o.O ??? Jul 14 23:52:34 .13 from PCB, .4 from battery? Jul 14 23:52:45 even it that's correct it only is a humble 0.6W Jul 14 23:52:59 yeah Jul 14 23:53:12 I honestly doubt that .130 from PCB Jul 14 23:54:08 DocScrutinizer, as for choke saturating, You mean temperature gain measurable via bq27200, or one exclusive to b42150? Jul 14 23:54:09 joga, I also don't have money for such - I just convinced whole family to switch their android shits for N900's, so I bought 15 devices at once. some sold on TMO; after reinforcing ports... Jul 14 23:54:11 = it turned out to be quite cheap per device :P Jul 14 23:54:11 well, maybe, just maybe, jolla will deliver. Jul 14 23:54:15 I rather think bq24150 never been built to be a reference voltage source Jul 14 23:54:53 Estel_ I sure hope something comes out of it...or similar Jul 14 23:54:53 Well, Li-Ion needs +-.05V precision Jul 14 23:55:01 Estel_: since they all are close to each other, you'll notice at bq27200 as well Jul 14 23:55:08 I just recently updated some stuff and browsed the repos again and found so much cool things Jul 14 23:55:33 DocScrutinizer, all right, will watch it closely when charging from extra-empty (like 100 mAh charge or so) Jul 14 23:55:46 ShadowJK: only on end of charge where current is low Jul 14 23:55:49 nice thing about this phone is that it seems to get a lot better in time instead of becoming obsolete :) Jul 14 23:56:04 Estel_, is this a dual lucialiu thingy Jul 14 23:56:09 joga, yeah :) if only ram would be upgradeable... :P Jul 14 23:56:20 Or dual japod or what Jul 14 23:56:21 yeah Jul 14 23:56:25 ShadowJK, yes, but lucialui doesn't sold real ones anymore :( Jul 14 23:56:31 no, not japod Jul 14 23:56:31 :( Jul 14 23:56:40 I forget what her batteries were called Jul 14 23:56:43 ShadowJK, scud Jul 14 23:56:46 ah Jul 14 23:57:06 well, scud is now only ~1300, like genuine, so rather overpriced Jul 14 23:57:43 BTW, some time ago I've sold 2 last dual ones on TMO, but one guy cancelled order due to some receiving problems, and battery will return soob. I'll sell it on TMO, probably Jul 14 23:57:56 aka last dual scud known to exist :P Jul 14 23:58:04 (last unused) Jul 14 23:58:37 Weirdly, I'm doing fine with single battery :P Jul 14 23:58:55 but I really hope for putting 2x 18650 3.4 Phillips'es into modified body. I've measured it, and it's only 2mm thicker than mugen cover on stock body Jul 14 23:59:12 ShadowJK, weren't You mugen battery owner? Jul 14 23:59:16 yes Jul 14 23:59:38 I only used it for like less than a year, heh Jul 14 23:59:48 well, due to load being cut in half, many people report ~3200 effective capacity of dual scud Jul 14 23:59:52 I see Jul 14 23:59:55 never used it with second (RIP) N900 or third (current) N900 :P Jul 15 00:00:13 Also it's slightly swollen :P Jul 15 00:00:16 wasn't it comfortable with bogger cover? Jul 15 00:00:18 I see Jul 15 00:00:25 yes Jul 15 00:00:33 I did like the bigger cover Jul 15 00:00:44 Well, I must say that I like bigger cover more - easier to write on hardware keyboard, + this better grip Jul 15 00:01:10 You know, material used for mugen cover is less slippery Jul 15 00:01:11 * ShadowJK has 3 of those covers actually, one of them home modified to support camera and microsd Jul 15 00:01:42 ShadowJK, support camera and microSD? does it mean You had old ones that have"'t supported it? Jul 15 00:01:48 Yes Jul 15 00:01:57 for some years, mugen covers have silly switch for camera and magnet Jul 15 00:02:11 The first batch of covers didn't have magnet or camera switch Jul 15 00:02:13 LOL Jul 15 00:02:27 I glued this magnet to N900 corpus, anyway, and I've modified cover to have real camera lens cover, like stock Jul 15 00:02:37 Mugen redesigned cover to add magnet and camera switch, and sent two covers to everyone with the old cover. Jul 15 00:02:39 ShadowJK, I was aware about camera, but magnet? :P Jul 15 00:02:49 sounds nice Jul 15 00:02:55 thats why You have 3 :P Jul 15 00:03:01 yeah Jul 15 00:03:17 well, protecting camera via own slide cover was kinda priority for me Jul 15 00:03:25 I agree that the less slippery surface of the mugen cover is more comfortable Jul 15 00:03:29 I've put tutorial somewhere on TMO for modding it like that Jul 15 00:03:32 do they actually exploit the additional space? Jul 15 00:03:43 DocScrutinizer, nope, only for battery Jul 15 00:03:54 yeah for battery Jul 15 00:03:56 advantage is that it have plenty room for mods Jul 15 00:04:05 As a bonus, the mugen cover gets rid of the problem with N900 flash blinding its own camera Jul 15 00:04:14 like getting multiple cells or custom shaped cell Jul 15 00:04:27 DocScrutinizer, as for battery, sure, it's 2x bigger. But capacity is crappy for 80 USD or how much it costs Jul 15 00:04:31 let me find the mugen cell, it's kinda silly Jul 15 00:04:33 custom shaped Jul 15 00:04:48 ShadowJK, yea, this flash thing ;) Jul 15 00:05:22 ooh flash - how did they handle that? Jul 15 00:05:32 DocScrutinizer, the funny fact is that dual-scud costed 2x less, and being dual cell, it had 3Ah power vs mugen'ish 2400 mAh Jul 15 00:05:44 also, isn't the camera door to small when so far away from lens? Jul 15 00:05:48 DocScrutinizer, simple - no chrome border, no blue parts :P Jul 15 00:05:58 DocScrutinizer, no, it's big enough Jul 15 00:06:11 no ill effects re photos, equally the opposite, better ones Jul 15 00:06:31 oh no, I found my original mugen cover, and the magnet has escaped! Jul 15 00:06:41 lol Jul 15 00:06:43 silly thing is that there is NO slide'able cover for camera, it's exposed all the time. Only switch for hardware detection Jul 15 00:06:47 but the white strip I added for camera is still there Jul 15 00:07:22 DocScrutinizer, but it's quite easy to mod it for adding real slide cover for camera, like stock, with auto detection working too Jul 15 00:07:47 white strip? So You had camera detected as "on" all the time? Jul 15 00:08:05 no ill effects on power usage? Jul 15 00:08:16 2400mAh, muge... er muhahahaha Jul 15 00:08:42 DocScrutinizer, ? :) Jul 15 00:08:48 no ill effects on power Jul 15 00:08:56 capacity isn't funny, only price is :P Jul 15 00:09:06 with a cover like mugen, i'd cram at least 5 cells in there, for ~4000mah Jul 15 00:09:26 btw, using custom shaped cell, they could, for holy hell's sake, upgrade it for better capacity, with time Jul 15 00:09:44 DocScrutinizer, kinda impossible re free volume Jul 15 00:09:55 that's what I meant Jul 15 00:09:57 though about using more smaller cells, or one bigger Jul 15 00:10:11 it doesn't turn out better than dual scud Jul 15 00:10:18 re flat cells available Jul 15 00:10:38 well, 18650 is gonna change everything, if it works out okey'ish :P Jul 15 00:10:51 severla smaller cells would for sure result in better capacity than mugen Jul 15 00:10:54 2 of them summing up for 6.8 A is kinda... well Jul 15 00:11:36 http://enivax.net/jk/n900/mugen.jpg Jul 15 00:11:39 DocScrutinizer, I've checked flat cells capacities on different size and volume available, and it was slightly better than mugen, but not better than dual scud. Not to mention price increase Jul 15 00:11:42 how much higher is the mugen? Jul 15 00:12:18 ooh fsck! Jul 15 00:12:20 .75mm higher than 2 BL-5J Jul 15 00:12:28 :P Jul 15 00:12:36 hmm Jul 15 00:12:39 ? Jul 15 00:12:43 ough, this N900 looks kinda unused Jul 15 00:12:54 well, it's almost like 2 bl-5j Jul 15 00:13:16 yes, and wasting all the empty space around battery Jul 15 00:13:19 that's why dual scud fits there fine Jul 15 00:13:23 DocScrutinizer, yea ;) Jul 15 00:13:51 Well, I use that space around battery how "grip" holding spare microSD card Jul 15 00:13:59 s/how/for/ Jul 15 00:14:01 Estel_ meant: Well, I use that space around battery for "grip" holding spare microSD card Jul 15 00:14:08 The new cover has a crossbeam that reaches down to the "hole" below the camera lens, this beam extens across teh width of the n900 Jul 15 00:14:23 DocScrutinizer, could You, kindly, modify bot to listen to me? It's kinda silly to do: Jul 15 00:14:30 ~n900 Jul 15 00:14:31 I'd find cells that are as long as N900 is high, and place several of them all on the back, covering all available area Jul 15 00:14:31 extra, extra, read all about it, n900 is a wooden box with stained cans inside and a beer Jul 15 00:14:46 ...just to use factoids :P Jul 15 00:14:54 ooh actually bigger than that Jul 15 00:15:52 * ShadowJK finds unused mugen cover in original plastic bag Jul 15 00:15:57 DocScrutinizer, the thing is that 18650, despite cylindrical size, offer better capacity per volume than flat ones Jul 15 00:15:57 still covered in protective plastic foil Jul 15 00:15:58 :P Jul 15 00:16:11 not to mention being much cheaper, even best ones Jul 15 00:16:17 due to popularity, etc Jul 15 00:17:09 single 18650 have - right now - reliable 3.4A capacity, so 2 of them, despite "wasted" space around cyllinders, are still better than flat ones using same volume and exploiting this space Jul 15 00:17:12 fully Jul 15 00:18:34 ShadowJK, btw, I'm planning to make "better grip" surface on custom backcover (for custom body, or seperate) too Jul 15 00:18:41 http://enivax.net/jk/n900/mugencover1.jpg http://enivax.net/jk/n900/mugencover2.jpg Jul 15 00:18:42 and better kickstand .P Jul 15 00:19:30 and this lovely port for external antenna, on custom body... Jul 15 00:19:39 antennas * Jul 15 00:20:19 ...and this lovely messing with ground plane... :P Jul 15 00:23:02 and contacts for a charger stand Jul 15 00:23:30 (or a coil .. inductive charging ftw) Jul 15 00:24:01 http://devel.japh.se/n950/fm/20120706_001.jpg Jul 15 00:24:29 fm antenna? :) Jul 15 00:24:35 yep :) Jul 15 00:26:27 docscrutinizer05, 11:02 ❌ doc|home japh: ping me later in the weekend <- not sure if that was you? Jul 15 00:26:53 http://maemo.cloud-7.de/share-service/20120715_003.jpg http://maemo.cloud-7.de/share-service/20120715_004.jpg Jul 15 00:27:10 :o Jul 15 00:27:13 nope Jul 15 00:27:34 oh he *was* here. doc|home, ping. Jul 15 00:27:59 3*1300mAh Jul 15 00:28:07 haha .. if you dont need the cam, i am sure you find 3 batteries which fit :P Jul 15 00:28:08 DocScrutinizer, anyway way to sensibly check if choke in our step down converter isn't going into saturation? Jul 15 00:28:18 You know, with shields took out Jul 15 00:28:19 docscrutinizer05, nice work! Jul 15 00:28:30 and some tool touching b42150 pins... Jul 15 00:28:54 Estel_: the choke is the black cube next to bq24150 Jul 15 00:29:00 DocScrutinizer, as for batteries, forget about place for camera? Jul 15 00:29:04 ouh Jul 15 00:29:05 made of ferrite ceramic Jul 15 00:29:24 place for camera? Jul 15 00:29:38 You can't use whole N900's lenght Jul 15 00:29:45 I didn't Jul 15 00:30:16 isn't the first photo clear enough? Jul 15 00:30:22 I see, but it is little too long to close cover Jul 15 00:30:34 batteries too long Jul 15 00:30:44 tried that already :P Jul 15 00:31:00 even with cells without heads with pins, no way to close cover (big one) Jul 15 00:31:06 yes, since I picked 'random' batteries to do those snapshots as we chat Jul 15 00:31:10 ah Jul 15 00:31:47 hm, dr_frost_dk have hundreds of batteries lying around, I'll forward idea to him Jul 15 00:31:49 and still I think with plastic cut off they would fit into mugen Jul 15 00:31:54 maybe some model is better suited Jul 15 00:32:33 as for choke black cube Jul 15 00:32:45 (those were GTA02 batteries) Jul 15 00:32:46 any way to measure something about it? or only temperature? Jul 15 00:32:53 http://www.hobbyking.com/hobbyking/store/uh_listCategoriesAndProducts.asp?idCategory=86&LiPoConfig=1&sortlist=&CatSortOrder=desc ..see if you find something :) Jul 15 00:33:19 with a scope you could find out about current or magnetic field Jul 15 00:34:10 but temperature is a pretty good indicator, as this thing should stay cool until it gets fsckng hot Jul 15 00:34:19 fsck Jul 15 00:35:10 btw, ShadowJK, here is tutorial for adding simple but effective camera slider to mugen cover v 2.0 Jul 15 00:35:11 http://talk.maemo.org/showpost.php?p=1038196&postcount=799 Jul 15 00:35:28 (i.e. one with fixed design, like those 2 You have) Jul 15 00:36:00 DocScrutinizer, I'll put thermal sensor to this cube, when charging from empty Jul 15 00:36:42 most significantly temp will drop sharply with battery pulling a lil bit less current Jul 15 00:37:01 in case it comes out of saturation eventually Jul 15 00:37:34 but honestly I doubt Nokia did SUCH botch Jul 15 00:37:48 ShadowJK, btw, what are those charging status'es shown when starting charge.sh, like 0x10? Jul 15 00:38:01 DocScrutinizer, I hope so :P Jul 15 00:38:12 finding a good choke isn't simple, and chip manufs suggest particular build&manuf they tested Jul 15 00:38:38 and that one is fit for chip max ratings Jul 15 00:39:06 so 1250 should be fine and happy, in theory. Jul 15 00:39:17 yep Jul 15 00:39:40 otherwise I wouldn't have stated "charging is safe no matter what you do" Jul 15 00:39:56 ok, going to sleep. ShadowJK, if You answer later, I'll check chanlogs :) Jul 15 00:40:04 I only later noticed there's *theoretical* possibility of Nokia building crap here Jul 15 00:40:26 DocScrutinizer, yea, well, we will test it, but for now, it looks good Jul 15 00:40:29 Estel_, they're the raw status value as returned by the charger chip Jul 15 00:40:35 their meaning is in the datasheet.. Jul 15 00:40:39 ShadowJK, ok, how to interpret them? Jul 15 00:40:41 ah Jul 15 00:40:44 thanks a lot Jul 15 00:41:10 bb and good night! Jul 15 00:41:28 one of the meansings is "overheat" Jul 15 00:41:39 meansong Jul 15 00:43:18 seems booson(2).sh script does proper decoding, and I thought i've seen similar thing in more recent charger scripts of ShadowJK? Jul 15 00:43:42 yes, decodes to /home/user/MyDocs/charge.log Jul 15 00:44:09 btw the meaning doesn't change much between boost and charge Jul 15 00:45:30 ShadowJK: oh wait, isn't the 0x10 at start the general status and not the error? Jul 15 00:45:43 It's "In Progress" Jul 15 00:45:50 lyep Jul 15 00:46:04 sth like that Jul 15 00:46:05 0x10 and 0x90, the other bit is for gpio status Jul 15 00:46:20 OTG aka fastcharger Jul 15 00:47:47 though not reliable as musbcore messing with 1707 config Jul 15 00:50:01 ShadowJK: btw if your script would monitor softstart, it could determine max USB current available for next retry Jul 15 00:50:35 yes Jul 15 01:03:23 yes this is rid Jul 15 01:15:22 DocScrutinizer05: well, slightly smaller inductor Jul 15 01:16:04 hmmm Jul 15 01:16:05 no Jul 15 01:18:01 hmm, capacitor is smaller than in reference design Jul 15 01:19:32 a lot smaller, 100n in n900 and 10uF in reference design Jul 15 01:23:54 on the output side? Jul 15 01:25:34 yes Jul 15 01:26:29 Are you familiar with the charger design before N900? :-) Jul 15 01:26:49 no Jul 15 01:26:54 not on phones Jul 15 01:27:02 but generally yes Jul 15 01:27:56 Before N900, the DC adapter was current limited to 900mA, and charge regulation was basically a single mosfet switched on and off at 1Hz. :-) Jul 15 01:28:30 on NITs or generally nokia phones? Jul 15 01:29:06 but well, those chargers were 3.7V for a reason Jul 15 01:29:08 All nokia phones of same generation as NITs :-) Jul 15 01:29:32 They actually output somewhere above 5.2V and less than 12V Jul 15 01:29:33 basically most of charging was done by a charger Jul 15 01:29:48 it says 3.7V on the charger Jul 15 01:29:49 Below 4.8V there was a linear regulation mode Jul 15 01:30:17 The "fast charger" above 5.2 was 1Hz pwm :- Jul 15 01:30:19 :-) Jul 15 01:30:43 well, bq isn't that much different Jul 15 01:30:43 It would accept 12V, if the supply was current limited to 100mA Jul 15 01:30:46 it's just much faster Jul 15 01:31:31 The first two batteris in my N810 caused reboots because bme aborted when the battery voltage during "ON" stage climbed above 4.55 :-) Jul 15 01:32:57 'ftd' gives insight into the voltages observed during charger on and charger off.. Also it's pretty amusing looking at a dc amp needle waving back and forth between .8A and 0A :-) Jul 15 01:33:12 boom doom bye Jul 15 01:33:51 My N800 backlight varies in brightness too when the charger is switched on and off... on my N810 it doesn't Jul 15 01:35:30 vlol Jul 15 01:35:40 LOL* Jul 15 01:36:26 jacekowski: 100n vs 10µ is basically irrelevant on output Jul 15 01:37:38 a) battery is supposed to have an own 100µF somewhere, for modem. b) as long as the C can take one choke power quantum, it's just fine Jul 15 01:40:11 My "server" pc has 3 different kind of USB ports. Back panel ones attached to motherboard, front panel ones in a 3.5" drive bay, and ports attached to a PCI backplate. The ones on the PCI backplate have extra capacitors on the power rails. The front usb port wont charge N900, at any current limit configured into bq24150, except when battery is near full. The PCI plate ports will charge N900 even when N900 battery is low, and N900 charger configured to no Jul 15 01:40:17 limit... Jul 15 01:43:45 the more interesting/relevant question regarding choke saturations is operation mode of stepdown converter: is it freq-regulated, or pulse pause regulated (waiting with next constant length charging step for a variable time, possibly quantisized), or will it actually do PWM to the amount/duration of energy stored to choke in one cycle? Jul 15 01:44:25 only the last mode is prone to resulting in possible saturation Jul 15 01:44:34 * ShadowJK knows nothing of step-down converters, but vaguely recalls something like pulse pause mentioned in datasheet Jul 15 01:44:49 yep, leaving out whole steps Jul 15 01:44:55 cycles Jul 15 01:44:59 it had a funny acronym Jul 15 01:45:03 something with p and q in it? Jul 15 01:45:19 pqm? Jul 15 01:45:26 I probably remember wrong and now you're thinking "wtf" and have to go read the datasheet Jul 15 01:45:28 pulse quantity modulation? Jul 15 01:45:33 perhaps Jul 15 01:46:43 usually called pulse pause modulation iirc Jul 15 01:48:09 and no, I won't try to find the datasheet on this mess in a bottle of this new laptop Jul 15 01:48:36 :P Jul 15 01:49:06 my whole old laptop is a dir in my ~ Jul 15 01:49:32 so I needed to dig that DS out there Jul 15 01:49:46 since no bookmarks in new clean browser Jul 15 01:53:35 ShadowJK: (old chargers tagged "4,8V" or whatever) they got some defined Z, so that's why it's somewhat not completely gaga to use them directly to charge a LiIon Jul 15 01:55:02 there's actually a spec for how to be compatible with them Jul 15 01:55:06 and I had a damn hard time to charge my 6210 from a stabilized universal PSU as on 4.5V it wasn't really charging, and at 6V it eventually blew chunks on too low battery charging Jul 15 01:56:57 I remembered name of document, weirdly after elopacalypse nokia.com isn't first hit anymore, but here it is: http://www.herjulf.se/solar/charge-station/Nokia_2-mm_DC_Charging_Interface_Specification_v1_2_en.pdf Jul 15 01:57:31 the funny bit: this 1Hz PWM charger concept with ancient BME kept my LiIon battery fresh for ~7 years of permanent usage and daily charging Jul 15 01:59:49 a low rate charger would be sufficient :) Jul 15 01:59:55 or low voltage (<5V) Jul 15 02:04:29 thanks Jul 15 02:06:46 Or then just occasionally Nokia sources good quality batteries Jul 15 02:07:16 It's weird that the Mugen battery in my N800 has outlasted 2 original Nokia batteries in N810 Jul 15 02:07:31 (original-formfactor mugen) Jul 15 02:07:59 You'd think Nokia would source better quality than random asian manufacturers :/ **** ENDING LOGGING AT Sun Jul 15 02:59:59 2012