IRC log of #maemo-ssu for Monday, 2013-04-08

*** NIN101 has quit IRC00:00
*** Estel_ has joined #maemo-ssu00:20
*** Estel_ has quit IRC00:20
*** Estel_ has joined #maemo-ssu00:20
*** xes has joined #maemo-ssu00:24
*** arcean_ has quit IRC00:27
*** arcean_ has joined #maemo-ssu00:28
Estel_kerio,  do you remember one of my devices that can't charge normally due to treating usb data pins as non-existent, so it require forcing dedicated mode in bme replacement?00:31
Estel_I did as suggested, i.e. forced it in /etc/event.d/bme, but it turned out, that it's too late for some situations00:31
Estel_for example, if battery get discharged completely (to the point of under 3000 mV), device doesnt event want to start booting, and when connected to charger, only *pretends* that it is emergency charging, by yellow diode00:32
Estel_now, where to modprobe bq2415x_battery, to get earliest charging possible?00:32
Estel_/sbin/preinit sounds ok, but...00:32
Estel_what about act_dead mode or emergency charging? anyone know what else should i edit to put charging module there?00:33
Estel_device uses u-boot, if that matters00:33
Estel_normally no problems, but if device user, accidentally, let it discharge too much, no way to charge it without using external charger, even if for 5 minutes (to allow device boot normally, at plug-in charger during boot). If possible, I would like to achieve emergency charging, too00:34
Estel_ShadowJK,  any ideas, how to force device into *really* charging in emergency mode (charger plugged in and battery under 3000 mV)00:34
*** Vlad_on_the_road has quit IRC00:34
Estel_or is it all hardcoded so without chip detecting crossed data, it will just light yellow diode, but do nothing? (as it does now)00:35
*** Martix has quit IRC00:37
*** xes has quit IRC00:39
*** xes has joined #maemo-ssu00:39
DocScrutinizer05amber LED is hardwired to bq24150 STAT output that indicates chip is in charging mode00:46
DocScrutinizer05D+/- detection by PHY is hardwired to bq24150 OTG input, that in POR state of schip determines whether chip is doing 500mA or 100mA charging00:47
kerioso Estel_'s stuck doing emergency charge at 100mA?00:47
DocScrutinizer05bottom line: when steady amber LED, bq24150 *is* charging battery, at least with 100mA00:48
Estel_DocScrutinizer05,  strange, as with this device, it doesn't work in emergency mode at all, while in normal mode it charger pefectly00:48
*** entitled has joined #maemo-ssu00:48
Estel_standard 1200 mA battery went from 2900 mV to 2600 mV00:48
Estel_during 10+ hours of "emergency charging"00:49
DocScrutinizer05a broken PHY will cause no D+/- short detection, so charging doesn't get forced to 500mA or even more, but stays at 100mA00:49
Estel_in that particular device00:49
Estel_i would be happy with 100 mA too00:49
Estel_I wonder why it doesn't charge at all. I'm not sure if it isn't even eating power due to that bright diode, lol00:49
DocScrutinizer05I don't know how much the system will consume when in pre-boot state at low voltage00:50
Estel_but that battery definitelly should get charged above 3000 mV pretty fast, @100mA00:50
Estel_just assumption - it shouldn't consume as much as 100mA, methinks00:50
DocScrutinizer05it *should* be really low power consumption00:50
* Estel_ nods00:51
Estel_btw on other devices I was able to emergency charge from dumb charger (100 mA), so something doesn't work as intended with broken phy00:51
Estel_ah00:51
Estel_from time to time, bme replacement detect this device as "boost mode", if not forced00:51
Estel_when connected to charger00:51
Estel_I wonder why, and how it can affect emergency charging?00:52
Estel_(btw, during non-charging, idle consumption is absolutely normal, so no power leaking)00:52
DocScrutinizer05maybe whatever damaged PHY (extreme overvoltage on USB?) also smoked up bq2415000:52
Estel_anyway, it seems that I can't affect the way emergency charging is happening via any means?00:52
Estel_DocScrutinizer05,  shouldn't be the case, as it's charging fine when device is booted using bq2415x_charger module :P00:53
DocScrutinizer05emergency charging is a purely hw-defined thing, as long as bq24150 is in POR state (hard to keep it in any other state)00:54
*** kolp has quit IRC00:54
Estel_although unknown reason for device being damaged (I got it in this state, for free, as un-repairable, revived it by preparing working backup to restore, that used modules for kernel matching one in-device)00:54
Estel_I see00:54
Estel_hm, I wonder what could trigger other state...00:54
DocScrutinizer05twl4030 might become sticky on some weird settings regarding power/usb00:55
DocScrutinizer05~full-reset00:55
DocScrutinizer05or somesuch00:55
Estel_anyway, leaving emergency mode - when battery is above 3000 mV and device boots into act_dead when charger plugged in, modprobing charging module in /sbin/preinit is enough?00:56
Estel_DocScrutinizer05,  if full reset is that thing including removing battery for extended period of time, sadly, I tried it00:56
DocScrutinizer05~listkeys #maemo00:56
infobotFactoid search of '#maemo' by key (16 of 44): #maemo bme-replacement ;; #maemo bmo ;; #maemo bnf ;; #maemo closed ;; #maemo closed_packages ;; #maemo db ;; #maemo ek ;; #maemo expired-key ;; #maemo fam ;; #maemo gmo ;; #maemo interim-dns ;; #maemo kp ;; #maemo mirror ;; #maemo mirrors ;; #maemo mo/* ;; #maemo multiboot.00:56
DocScrutinizer05~listkeys reset00:57
infobotFactoid search of 'reset' by key (15 of 19): password reset #DEL# ;; looks like the reset ;; hard reset ;; password reset ;; reset ;; firm reset ;; eq.preset ;; jargon 120 reset ;; jargon reset ;; only reason you absolutely must reset, ;; warm reset ;; trackerreset ;; connection reset by peer ;; soft reset ;; full reset.00:57
DocScrutinizer05~full reset00:57
infobotfrom memory, full reset is flip the battery switch to "replace" open the cover, press the little button with the stylus, close cover, flip the switch to normal, power on.  Warning: data in ram storage will be lost by doing a full reset00:57
Estel_For me it isn't problem, but device is used by my son on trips and somesuch, when LiPo charger or other N900's are not around. I'll wind up bancruting and buying external charger for 5$, probably. And modding it, cause ones sold here have risk of killing people by idiotic design, where *battery" pins have possibility to bend into touching 220V00:57
DocScrutinizer05meh!00:57
Estel_I just started to think I've missed some little button on N900 :P00:58
Estel_thanks for taking your time to explain emergency charger and searching for that factoid, btw00:58
*** gregoa has quit IRC00:58
*** gregoa has joined #maemo-ssu00:59
Estel_so, anyone remember how fckn act_dead is called? it doesn't skip /sbin/preinit in actdead mode, I hope? actually, it should be called from /sbin/preinit if certain conditions are meet?00:59
DocScrutinizer05full reset for twl4030 is done by holding powerbutton for 15s while inserting battery, or somesuch01:01
DocScrutinizer05~listkeys reset01:04
infobotFactoid search of 'reset' by key (15 of 19): trackerreset ;; password reset ;; looks like the reset ;; spa2000_factory_reset ;; jargon 120 reset ;; full reset ;; eq.preset ;; connection reset by peer ;; hard reset ;; soft reset ;; jargon reset ;; reset switch ;; warm reset ;; only reason you absolutely must reset, ;; firm reset.01:04
merlin1991freemangordon: use http://community-testing.merlin1991.at/ ;)01:04
freemangordonmerlin1991: it is missing there too01:04
merlin1991dafuq01:04
freemangordonand in all of the mirrors01:05
Estel_DocScrutinizer05,  thanks01:05
freemangordonmerlin1991: oh, wait01:05
Estel_one should hold it for at least 15 seconds and then enter battery, or hold and insert, then keep holding?01:05
freemangordonmerlin1991: it is in upstart, WTF?01:05
merlin1991http://community-testing.merlin1991.at/pool/free/u/upstart/system-services_0.3.8-68%2b0cssu1_armel.deb01:05
merlin1991yeah silly upstart package01:05
freemangordonyeah, my bad, I was looking in s directory01:06
merlin1991I did aswell01:06
freemangordonmerlin1991: however, repo should be fixed01:06
Estel_kerio, how one force u-boot to boot from sd, from u-boot terminal? it proposes some scripts like sdboot or emmcboot, but none of them works01:06
freemangordonEstel_: run sdboot01:06
Estel_(when one enters u-boot console, it informs about it, but none of them works)01:07
freemangordonor run mmcboot01:07
Estel_freemangrdon, it gives shitload of errors01:07
DocScrutinizer05I think it been like "hold for 12s at most 12s after inserting battery" or whatever01:07
Estel_any of them01:07
kerioEstel_: manual parameters01:07
freemangordonEstel_: then you don't have fat first partition. afaik01:07
Estel_freemangordon,  I have fat as first partition on emmc and it works from u-boot menu entries01:08
Estel_*just* mmcboot command doesn't work01:08
merlin1991DocScrutinizer05: are you planning on staying up longer today?01:08
Estel_it complains something about ext4, maybe it get confused by mmcblk0p2 being eext4? but why it cares, it should look at mmcblk0p1 only01:08
freemangordonEstel_: no idea then01:09
Estel_freemangordon,  considering that I have only one partition on sd and that partition is vfat, and using sdboot command result in same ext4 errors...01:09
Estel_I assume u-boot browse through all partitions and get pissed off by finding something that it doesn't like, even when he shouldn't care01:10
Estel_need to pester Pali, probably, but if he is going to say "won't fix, try to push upstream patch", I'm stopping to report anything to Pali :P01:10
Estel_if I would be able to push upstream patch, I wouldm fuck in fix it myself, instead of asking here or reporting anything to him01:10
Estel_s/fuck in/fucking/01:11
infobotEstel_ meant: if I would be able to push upstream patch, I wouldm fucking fix it myself, instead of asking here or reporting anything to him01:11
DocScrutinizer05merlin1991: probably01:16
DocScrutinizer05merlin1991: why?01:16
merlin1991might need some sudo on garage/repo later on01:16
DocScrutinizer05np01:16
Estel_DocScrutinizer05,  now the question is, does hard reset of charging chip work if battery inserted during performing hard reset is on unholy low voltage?01:31
Estel_or holy low voltage, like < 3000mV but > than 2500 mV01:32
Estel_(or unholy, like something like 2200 mV - never happened, but thinking theoretically). I wonder what kind fo voltage from battery is required to perform full_reset?01:32
Estel_"from different barrel" - I have 4 packages ready, but no time to polish packaging and release into repositories, ffs :P01:38
Estel_irony hits - all of them are for non-free section (kon-boot, and 3 games)01:38
Estel_especially funny, considering, how I used to ignore annoucements for  non-open source things for Maemo01:39
Estel_especially for non-FOSS game projects, heh01:39
*** arcean_ has quit IRC01:51
*** arcean_ has joined #maemo-ssu01:53
*** M4rtinK has quit IRC02:01
ShadowJKEstel_; afaik it's always limited to 100mA hardware-only. When I had uBoot+meego on spare device, bme would not start charging empty battery (nolo charging missing), but there was still enough time for me to launch xterm and launch meegocharge21.sh02:14
DocScrutinizer05ShadowJK: according to schematics and bq24150 datasheet it's forced into 500mA mode on OTG=1 in POR mode02:17
DocScrutinizer05OTG  D4 I02:20
DocScrutinizer05Boost mode enable control or input current limiting selection pin. When OTG is in active status, bq24150/1 is forced to operate in boost mode. It has higher priority over I2C control and can be disabled through control register. The logic voltage level at OTG active status can also be controlled. At POR, the OTG pin is default to be used as the input current limiting selection pin. When OTG = High, Iin – limit = 500mA and when OTG02:20
DocScrutinizer05= Low, Iin – limit = 100mA, see the Control Register for details.02:20
DocScrutinizer05while02:22
DocScrutinizer05STA TC 4 O  Charge status pin. Pull low when charge in progress. Open drain for other conditions. During faults,  a128-mS pulse is sent out. STAT pin can be disabled by the EN_STAT bit in control register. STAT ca nbe used to drive a LED or communicate with a host processor.02:23
*** arcean_ has quit IRC02:26
ShadowJKDocScrutinizer05; current balance and the time I had to start charge indicated 100mA02:26
ShadowJKI didn't always manage to start charging before voltage dropped too low02:27
DocScrutinizer05sorry, I can't follow02:31
DocScrutinizer05NB the OTG function is defined for POR (non I2C) mode only02:31
DocScrutinizer05in I2C mode you can configure OTG input to anything you like02:31
DocScrutinizer05btw difference of 1704 and 1707 is in level of charge output, so it matches what bq24150 needs for emergency fastcharging to work02:32
DocScrutinizer05iirc02:32
*** nox- has quit IRC02:49
*** xes has quit IRC03:31
Estel_thanks ShadowJK and DocScrutinizer05. ShadowJK, I don't get from where you've run meegocharge21.sh - I'm talking about situation, when device doesn't boot into anything system-related (for example, 2900 mV battery inserted)03:40
Estel_DocScrutinizer05,  tried reset of chip, haven't helped. I wonder if it may be related to some fubar with OTG pin?03:40
Estel_physically, that pin is ok, but who knows what happened inside broken phy (or whatever is broken there)03:41
Estel_anyway the result is, that while emergency "charging", that particular device is, in fact, discharging, at natural or even a little increased level03:42
Estel_another interesting thing - if device connected to charger *without* battery, led is yellow/orange the same way, like emergency charging in progress03:42
Estel_for now, I've ditched LiIon charging pcb from small bluetooth headphones, that take 5V as input, into inside of backcover...03:45
Estel_with wires attached to battery cell and wires from that charging PCB attached to battery wires via mini clips, + charging pcb attached via cables to device 5V input (charger port aka microUSB), with hardware switch in between03:47
Estel_= when emergency charging needed on that device, I plug in charger like normal, then close switch, resulting in recycled charging circuit being powered and charging cell. After few minutes I open switch, de-attach nokia charger, boot device, and attach charger again, quickly - bq2415x_battery module takes care of normal maemo charging, then (bme repl.)03:49
*** LauRoman has quit IRC04:08
ShadowJKEstel_; with uboot + meego, actdead did not exist, it booted into meego straight04:51
*** amiconn has quit IRC05:05
*** amiconn_ has joined #maemo-ssu05:05
*** amiconn_ is now known as amiconn05:05
*** BCMM has quit IRC05:43
*** DocScrutinizer05 has quit IRC06:03
*** DocScrutinizer05 has joined #maemo-ssu06:03
*** Phuzzed has joined #maemo-ssu06:42
*** Phuzzed has quit IRC06:43
*** M13 has joined #maemo-ssu06:47
*** M13 has quit IRC07:20
*** luf has joined #maemo-ssu08:13
*** dhbiker has joined #maemo-ssu09:02
*** unclouded has joined #maemo-ssu09:17
*** Mihanizat0r has joined #maemo-ssu10:01
*** LauRoman has joined #maemo-ssu10:09
*** M4rtinK has joined #maemo-ssu10:27
*** M4rtinK has quit IRC11:10
*** futpib has joined #maemo-ssu11:42
*** LauRoman has quit IRC13:22
*** Martix has joined #maemo-ssu13:40
*** futpib has quit IRC14:03
*** futpib has joined #maemo-ssu14:05
*** unclouded has quit IRC14:22
*** Mihanizat0r has quit IRC14:25
*** Martix has quit IRC15:09
*** M13 has joined #maemo-ssu15:22
*** andre__ has joined #maemo-ssu15:32
*** andre__ has joined #maemo-ssu15:32
*** BCMM has joined #maemo-ssu15:39
*** DocAvalanche has joined #maemo-ssu15:50
*** DocScrutinizer51 has quit IRC15:51
*** futpib has quit IRC15:51
*** DocAvalanche is now known as DocScrutinizer5115:51
*** futpib has joined #maemo-ssu15:52
*** lizardo has joined #maemo-ssu15:53
*** kolp has joined #maemo-ssu16:17
*** MohammadAG has quit IRC17:04
*** arcean has joined #maemo-ssu17:17
Estel_ShadowJK,  I see17:20
*** M4rtinK has joined #maemo-ssu18:24
*** NIN101 has joined #maemo-ssu18:53
*** dafox has joined #maemo-ssu19:35
*** sunny_s has quit IRC20:02
*** sunny_s has joined #maemo-ssu20:06
*** dafox has quit IRC20:21
*** xes has joined #maemo-ssu20:53
*** Vlad_on_the_road has joined #maemo-ssu21:22
*** dafox has joined #maemo-ssu21:26
*** dafox is now known as Guest7024421:26
*** Mihanizat0r has joined #maemo-ssu21:39
*** M13 has quit IRC21:42
*** Mihanizat0r has quit IRC21:44
*** BCMM has quit IRC22:06
*** Guest70244 has quit IRC22:11
*** Guest70244 has joined #maemo-ssu22:11
*** sunny_s has quit IRC22:46
*** NIN101 has quit IRC22:52
*** futpib has quit IRC23:17
*** andre__ has quit IRC23:20
*** andre__ has joined #maemo-ssu23:20
*** andre__ has joined #maemo-ssu23:20
*** LauRoman has joined #maemo-ssu23:41
*** dhbiker has quit IRC23:42
*** Martix has joined #maemo-ssu23:48
*** nox- has joined #maemo-ssu23:51
*** Martix has quit IRC23:52
*** Martix has joined #maemo-ssu23:53
*** lizardo has quit IRC23:56
Estel_DocScrutinizer05,  I got interesting thing to discuss with you. That device wit borked usb phy started to act strangely - after boot, it uses ~ 500 mA, idle. Powertop and CPU usage are fine. When I try booston and boostoff from HEN, it becomes OK23:57
Estel_same happens when I plug in charger and unplug23:57
Estel_it started when I did hard reset of chip, btw23:57
Estel_I assume that resetting bq24150 by i2c commands from booston and it's trap, are fixing that23:58
Estel_so, it seems, that device start a day, after boot, in boost mode23:58
Estel_ironically, if I disable charging at all, plugging in charger result in *increase* of power usage (until I unplug it, then, everything is normal, and next plug-ins doesn't affect power usage)23:59
Estel_I wonder, what sits in that usb chip/phy?23:59

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!