| Wizzup | uvos__: lol I need even more d4 back covers | 08:15 |
|---|---|---|
| freemangordon | ...and my d4 has a new battery :) | 08:26 |
| freemangordon | and android is fine with charging it :) | 08:54 |
| uvos__ | freemangordon: androd ofc will charge it to 4.35V so be carefull with that | 09:13 |
| uvos__ | Wizzup: im pretty sure your buying the same ones that got removed from the d4's you also bought | 09:14 |
| freemangordon | uvos__: I rebooted @ about 70% full | 09:17 |
| freemangordon | uvos__: but, you said this battery is HV, no? | 09:17 |
| freemangordon | also, how to correctly calibrate? charge full and then discharge full? | 09:19 |
| freemangordon | or echo 2000000 > charge_full? | 09:20 |
| uvos__ | well if you know the capacity you can echo it shure | 09:20 |
| freemangordon | it was written on it :) | 09:20 |
| uvos__ | well | 09:20 |
| uvos__ | otherwise charge full and discarge fully, yes | 09:20 |
| freemangordon | what about 4.35? I ordered the battery you told me | 09:21 |
| freemangordon | and it is written 3.8V on it | 09:21 |
| uvos__ | pretty sure the nexus 4 battery is hv yeah | 09:21 |
| uvos__ | not that i use it that way | 09:21 |
| freemangordon | ah, I see | 09:21 |
| uvos__ | since leste never charges over 4,2 | 09:22 |
| freemangordon | "For LG E960" | 09:23 |
| freemangordon | this is what is written, I hope this is the correct one | 09:23 |
| freemangordon | exactly this https://www.ebay.com/itm/334581894280?hash=item4de6a28c88:g:DskAAOSwz65jO7SQ one | 09:24 |
| uvos__ | that should be the one | 09:29 |
| uvos__ | anyone know what implements a malinglist? | 09:30 |
| uvos__ | ie i need some deamon that recieves email and forwards it to a list of email addresses | 09:31 |
| freemangordon | uvos__: could you provide some details on "one wire protocol" that we currently have disabled? | 10:10 |
| freemangordon | I mean - how to enable and what is the issue? | 10:10 |
| uvos__ | modprobe ehci-omap | 10:11 |
| uvos__ | the issue is that the one wire interface framework supports hotplugging | 10:11 |
| uvos__ | the only way to support hotplugging on this bus is to scann the bus at all times | 10:11 |
| uvos__ | the omap driver implements this | 10:11 |
| uvos__ | scanning the bus at all times is ofc not ideal for battery life | 10:11 |
| freemangordon | what android does? | 10:12 |
| uvos__ | android just assumes that this bus has just one chip on it | 10:12 |
| uvos__ | and its location is hardcoded | 10:12 |
| uvos__ | this ofc isent going to fly in the mainline kernel | 10:12 |
| freemangordon | ok, lemme see how bad is it | 10:12 |
| freemangordon | hmm, what is supposed to happen? | 10:15 |
| uvos__ | a device appears in sys/bus/something | 10:16 |
| freemangordon | do I need omap_hdq too? | 10:16 |
| uvos__ | sorry it was omap_hdq not ehci | 10:17 |
| uvos__ | anyhow if the driver is not blacklisted | 10:17 |
| freemangordon | it is | 10:17 |
| freemangordon | but modprobe does not care about it | 10:17 |
| uvos__ | cpcap-battery will use it to determine what battery is inserted | 10:17 |
| freemangordon | mhm | 10:17 |
| uvos__ | *however there is a know issue with this code, that i wrote, that causes it to only work with some batteries | 10:18 |
| freemangordon | "omap_hdq 480b2000.1w: OMAP HDQ Hardware Rev 0.6. Driver in Interrupt mode" | 10:18 |
| freemangordon | it says interrupt mode | 10:18 |
| uvos__ | it sill scanns the bus | 10:18 |
| freemangordon | ok, lemme see what this code do | 10:19 |
| uvos__ | btw the known issue with the cpcap detection code | 10:19 |
| uvos__ | is that the device name of the chip changes (baste on what exact supplier motorola used for the eeprom i presume?), and i have no way of knowing what device on the bus is the right one | 10:20 |
| uvos__ | so it uses only the chip i hapen to have on my devce | 10:20 |
| freemangordon | hehe :) | 10:20 |
| freemangordon | ok, will see what we can do | 10:20 |
| freemangordon | uvos__: I see no polling there | 10:22 |
| freemangordon | why do you think it polls? | 10:23 |
| uvos__ | reading the protocoll spec | 10:23 |
| freemangordon | HW does polling? | 10:24 |
| uvos__ | it also uses a lot of power | 10:24 |
| uvos__ | its possible that hw dose poling, main symptom here is it draws a lot of power to have it on | 10:25 |
| freemangordon | ok, will see what happens, reading TRM atm | 10:26 |
| uvos__ | also theres a commen in android kernel somehwere | 10:27 |
| uvos__ | about avoiding the bus polling | 10:27 |
| uvos__ | *comment | 10:27 |
| freemangordon | uvos__: I think cpcap shall allow userspace to write the id of the battery | 10:35 |
| uvos__ | freemangordon: feals ugly, but maybe its the only option | 10:37 |
| freemangordon | unless we know how ids are coded | 10:37 |
| freemangordon | like - if it starts with 89, then it is the battery | 10:37 |
| uvos__ | right , but i dont, i wanted to figure out whre it comes from | 10:38 |
| uvos__ | but wasent a priority ofc with the code being dead due to the blacklist and all | 10:38 |
| freemangordon | it is programmed during manufacture afaik | 10:38 |
| uvos__ | ok | 10:39 |
| freemangordon | I don;t see it in nvram | 10:39 |
| uvos__ | well it might also help if everyone could post the ids they have | 10:39 |
| freemangordon | right | 10:39 |
| uvos__ | you have mine in the code | 10:39 |
| freemangordon | was about to ask you what is your id :) | 10:39 |
| freemangordon | ah, right | 10:39 |
| uvos__ | well mine is none now | 10:39 |
| uvos__ | since i dont use a battery with a chip :) | 10:39 |
| uvos__ | that id is from my bionic btw | 10:39 |
| freemangordon | hmm, still 98 | 10:40 |
| uvos__ | buZz: also posted his | 10:40 |
| freemangordon | (strcmp(dev_name(dev), "89-500029ba0f73") == 0) :D | 10:40 |
| uvos__ | its also 98- iirc | 10:40 |
| uvos__ | er 89 | 10:40 |
| uvos__ | freemangordon: :P | 10:41 |
| freemangordon | how did they not see that ?!? | 10:41 |
| uvos__ | eh | 10:41 |
| freemangordon | if it was me I would have NACK-ed that in a split second :p | 10:41 |
| uvos__ | i thought it might allways be the same | 10:41 |
| uvos__ | since my bionic has several bats | 10:41 |
| uvos__ | and they all have this | 10:41 |
| freemangordon | ah | 10:41 |
| freemangordon | lemme check mine | 10:41 |
| uvos__ | but buZz has a different id for sure | 10:42 |
| freemangordon | 89-50002aa3a5ae | 10:42 |
| uvos__ | on his d4 | 10:42 |
| freemangordon | Wizzup: could you post yours? | 10:42 |
| freemangordon | modprobe omap-hdq and then in /sys/bus/w1 | 10:42 |
| uvos__ | Wizzup could start a whole study | 10:43 |
| uvos__ | with his 100s of d4s xD | 10:43 |
| freemangordon | we need 1 only | 10:43 |
| freemangordon | to see if there is a common pattern | 10:43 |
| uvos__ | theres also other devices to consider | 10:43 |
| uvos__ | mz6xx xt860 etc | 10:44 |
| freemangordon | sure, but lets start form somewhere | 10:44 |
| freemangordon | also, I don;t see why hdq should draw power | 10:44 |
| freemangordon | any idea what "MOTOROLA E.P" could mean? | 10:48 |
| uvos__ | no idea about the contence of the rom | 10:48 |
| uvos__ | im pretty sure the bit i codes the voltage | 10:48 |
| uvos__ | *the bit i read | 10:48 |
| uvos__ | otherwsie no idea | 10:49 |
| freemangordon | ok | 10:50 |
| freemangordon | uvos__: are you sure hi current draw is still valid for 5.18? | 10:52 |
| freemangordon | the same for ehci | 10:52 |
| freemangordon | I see 55mA in idle with both modules loaded | 10:53 |
| freemangordon | which is the same as without them | 10:53 |
| uvos__ | well last time i tested this was probubly 5.10 or something | 10:53 |
| uvos__ | so no | 10:53 |
| freemangordon | could you test again when you have some spare time? | 10:54 |
| uvos__ | yes ofc | 10:54 |
| uvos__ | ill try to do it later today | 10:54 |
| freemangordon | ok | 10:54 |
| freemangordon | uvos__: yeah, it is the same | 11:04 |
| freemangordon | but, seems w1 bus has parameter how often to search | 11:04 |
| freemangordon | and also w1_max_slave_count | 11:04 |
| freemangordon | and we also have search_count | 11:05 |
| freemangordon | so we just have to set that to 3/1 and we are done | 11:07 |
| freemangordon | options wire search_count=1 max_slave_count=1 should do the trick | 11:12 |
| freemangordon | uvos__: 89(51) means battery monitor | 11:18 |
| freemangordon | so we can just check for that and then check in nvmem for MOTOROLA or something | 11:19 |
| freemangordon | hmm, actually no, my bad | 11:22 |
| freemangordon | this should be 128 bytes of memory :( | 11:23 |
| freemangordon | yep, with those parameters I see as low as 150mW | 11:32 |
| uvos__ | thats pretty high | 11:34 |
| uvos__ | but might be your system | 11:34 |
| freemangordon | yeah, this is usual here | 11:34 |
| uvos__ | thats 2x what my bionic idles at usually | 11:35 |
| freemangordon | no idea | 11:35 |
| uvos__ | anyhow ill check oo | 11:35 |
| uvos__ | *too | 11:35 |
| freemangordon | uvos__: "modprobe wire search_count=1 max_slave_count=1" | 11:35 |
| freemangordon | and then modprobe omap_hdq | 11:36 |
| uvos__ | as for identifieng idk if opening random devices to look for MOTOROLA is really acepptable | 11:36 |
| uvos__ | to the mainline folks | 11:36 |
| freemangordon | why random? it is nvmem | 11:36 |
| uvos__ | freemangordon: thanks | 11:36 |
| uvos__ | well random nvmem devices | 11:36 |
| freemangordon | ok, lets gather some IDs, will decide what to do | 11:37 |
| freemangordon | uvos__: 'your' idle is not as 'my' idle, I have ssh connections active | 11:38 |
| freemangordon | this is with 2 sessions, lemme check with 1 only | 11:39 |
| freemangordon | same | 11:41 |
| norayr | i think maaajordomo was a software to setup a mailing list. it was written in perl and had a web frontend. | 11:42 |
| norayr | majordomo | 11:43 |
| norayr | you also supposed to create a mail account for it so that it will be possible to subscribe or unsubscribe by sending an email. | 11:44 |
| norayr | uvos__ | 11:44 |
| uvos__ | norayr: yeah i know about that one | 11:45 |
| uvos__ | i was hoping for something simpler | 11:45 |
| norayr | it is sad that droid4 is not available on market (read ebay) any more for pretty long time already. i k last time i saw about a month ago the one which was in state 'read' and it had some problem. | 11:46 |
| norayr | uvos__ i think 20 years ago when i was searching for something that was the simplest. | 11:47 |
| uvos__ | norayr: ok yeah i might have to use it, i was thinking there would be some 300 line c program that just mails whatever it recives on pop3 to a static list. | 11:49 |
| norayr | yes it would be good to have such a thing. maybe that's a good idea to write something like that. i wish i had a bit of time, that's something that inspires me for work. | 12:06 |
| freemangordon | uvos__: do you know which blob reads nvmem on android? I may try to RE the structure | 12:24 |
| uvos__ | freemangordon: i dont quite remember, but it had an obvious name batd or something like that | 12:26 |
| freemangordon | ok | 12:26 |
| freemangordon | uvos__: nvram contents on 2 devices here is exactly the same | 12:49 |
| uvos__ | freemangordon: that is correct | 12:50 |
| uvos__ | its allways the same for eatch battery of one type | 12:50 |
| freemangordon | that's good | 12:50 |
| freemangordon | I guess | 12:50 |
| uvos__ | it dose _not_ contain any information on battery health or anything like that | 12:51 |
| uvos__ | or even a serial | 12:51 |
| freemangordon | right | 12:51 |
| freemangordon | this is just device id | 12:51 |
| freemangordon | so we can use to match | 12:51 |
| uvos__ | its lots of data for a device id | 12:51 |
| uvos__ | presumably its a voltage/current->state of charge table | 12:51 |
| freemangordon | maybe some cherging parameters | 12:51 |
| uvos__ | as this is how the batttery display works on android | 12:52 |
| freemangordon | could be | 12:52 |
| freemangordon | lemme feed IDA with batd | 12:52 |
| freemangordon | uvos__: so, blob checks for "CORP", "MOTOROLA E.P CHARGE ONLY", uid starting with 0x89 and last 2 bytes being 0x0050 | 13:31 |
| freemangordon | it also calculates 2 checksums, but I don;t think we shal care | 13:31 |
| uvos__ | that all? | 13:32 |
| uvos__ | what about all the other data | 13:32 |
| freemangordon | it is used co calculate "SBCM CORE convert_rom_data: bcap= %d, raw= 0x%x" | 13:33 |
| freemangordon | so most-probably you are right and this is voltage->% table | 13:33 |
| freemangordon | also it seems the capacity is there | 13:34 |
| freemangordon | uvos__: capacity = mul * batt_data.nvram.data[34]; | 13:39 |
| freemangordon | where mul is calculated based on some other data | 13:40 |
| freemangordon | I will provide the formula later on | 13:40 |
| uvos__ | something also has to tell the android what to charge too | 13:44 |
| freemangordon | what do you mean? | 13:44 |
| freemangordon | what is the type? | 13:44 |
| uvos__ | if you plug in a bw8x to bionic it suddenly knows to only charge to 4,2v | 13:44 |
| freemangordon | yeah, sure | 13:44 |
| uvos__ | while hw4x it will charge to 4,35 | 13:45 |
| freemangordon | ok | 13:45 |
| freemangordon | lemme finish with the capacity | 13:45 |
| buZz | uvos__: mine are here ; https://github.com/maemo-leste/bugtracker/issues/621 | 13:46 |
| buZz | freemangordon: ^ | 13:46 |
| freemangordon | uvos__: calculated value is 1740, sounds about right, no? | 13:47 |
| uvos__ | freemangordon: a bit low, but it bounces around for me alot too | 13:49 |
| uvos__ | it callibrates 1700 - 2000 for me | 13:49 |
| uvos__ | depending on day | 13:50 |
| uvos__ | (most likely this is voltage spikes triggering low early) | 13:50 |
| freemangordon | no, I mean - this is what is written in eeprom | 13:50 |
| uvos__ | oh | 13:51 |
| uvos__ | right | 13:51 |
| uvos__ | yes that is nominal for hw4x /eb41 | 13:51 |
| freemangordon | see https://pastebin.com/nJHHdeDt | 13:51 |
| freemangordon | this results in 1740 | 13:51 |
| freemangordon | but should work for all | 13:52 |
| uvos__ | i dont get why they go though this contortion | 13:52 |
| uvos__ | for a fixed value | 13:52 |
| uvos__ | but that sounds correct result wise | 13:52 |
| freemangordon | I bet this is used for all of their device ;) | 13:52 |
| freemangordon | *devices | 13:52 |
| tmlind | cool | 13:53 |
| freemangordon | hmm, voltage seems to be calculated with not-so-somple formula | 13:55 |
| freemangordon | *simple | 13:55 |
| uvos__ | freemangordon: btw | 13:55 |
| uvos__ | http://uvos.xyz/maserati/batterynvram/ | 13:55 |
| uvos__ | dumps | 13:55 |
| uvos__ | in case you want to check your formulars on other data | 13:55 |
| freemangordon | sure | 13:55 |
| uvos__ | like the bw8x | 13:56 |
| uvos__ | (should get 2200mAh iirc | 13:56 |
| uvos__ | ) | 13:56 |
| freemangordon | bionic_bw8x.bin - 2760 | 13:57 |
| uvos__ | that is correct | 13:58 |
| uvos__ | https://www.ebay.com/itm/255135819174?hash=item3b6747a5a6:g:8CwAAOSwCtlhQRZR | 13:58 |
| freemangordon | bionic_hw4x = 1740 | 14:00 |
| freemangordon | is that correct? | 14:00 |
| freemangordon | sorry, I have no idea what models are | 14:01 |
| uvos__ | yes hw4x is the same cell as eb41 in a different case/connector | 14:01 |
| freemangordon | so, our formula is correct :) | 14:01 |
| uvos__ | freemangordon: looks like :) | 14:01 |
| uvos__ | great work! | 14:01 |
| freemangordon | ok, will do some 'work' work first and then will get back to voltage | 14:01 |
| freemangordon | uvos__: please, if you have time, test hdq and ehci | 14:02 |
| freemangordon | maybe we can remove them from blacklist | 14:02 |
| freemangordon | I am sure for hdq though | 14:02 |
| freemangordon | ttyl | 14:02 |
| buZz | the code on them changed? | 14:03 |
| freemangordon | uvos__: it seems the voltage is taken from the table | 15:41 |
| freemangordon | does "16702 * nvmem[i] / 1000 + 1260" ring any bell? | 15:42 |
| uvos__ | not really | 15:48 |
| uvos__ | that would mean nvmem[i]=185 in eb41 case | 15:48 |
| freemangordon | 16702 * batt_data.nvram.data[65] / 1000 + 1260; | 15:48 |
| freemangordon | this is the charging voltage IIUC | 15:49 |
| uvos__ | well is [64] 185? | 15:49 |
| freemangordon | gives 4349 for d4 | 15:49 |
| freemangordon | lemme check | 15:49 |
| freemangordon | yes, 185 | 15:50 |
| uvos__ | i gues bw8x is 176 ish? | 15:50 |
| * freemangordon checks | 15:50 | |
| freemangordon | yes :) | 15:50 |
| uvos__ | great :) | 15:50 |
| freemangordon | the above formula calculates 4199 for 176 | 15:51 |
| freemangordon | what about 4349? | 15:51 |
| uvos__ | thats about right | 15:51 |
| freemangordon | ok | 15:51 |
| uvos__ | empiricly is charges to 4350 | 15:51 |
| freemangordon | right | 15:51 |
| uvos__ | but cpcap also wont have a register setting for 4349 | 15:51 |
| uvos__ | anyhow so it must convert again at some point | 15:52 |
| uvos__ | pretty wierd range they have there for voltage | 15:52 |
| freemangordon | some other value is calculated based on that | 15:52 |
| uvos__ | 1.2 ish volts to 5.5 ish votls maximum | 15:52 |
| freemangordon | well, why not? | 15:52 |
| uvos__ | so 2s lipo is out of the question motorola? | 15:52 |
| freemangordon | 2s lipo? | 15:52 |
| uvos__ | 2s lipos is pretty common in tables | 15:52 |
| uvos__ | 2 in series | 15:52 |
| freemangordon | ah | 15:52 |
| freemangordon | well... | 15:53 |
| freemangordon | lemme check the other one | 15:53 |
| freemangordon | hmm, maybe lowest is also somewhere there | 15:54 |
| Guest224 | Is hildon-connectivity-mobile package in any sd-card image yet? | 15:58 |
| Guest224 | or is it still unstable to put to image? | 16:05 |
| uvos__ | its still very wip | 16:10 |
| uvos__ | and no | 16:10 |
| uvos__ | im not sure why its important since you can just install it after | 16:11 |
| uvos__ | on d4 telephony and mobile data mostly works reliably now | 16:11 |
| Guest224 | how to install after? if wifi is disablet..ok..there is USB route if you have linux machine. | 16:12 |
| uvos__ | i mean wihout wifi | 16:12 |
| uvos__ | the devices are still fairly unusable | 16:12 |
| uvos__ | imo | 16:12 |
| freemangordon | uvos__: shall I try hard to find how low voltage is calculated or we can deduce from charge_voltage? | 16:13 |
| uvos__ | no | 16:13 |
| uvos__ | its essentaly constant anyhow | 16:13 |
| uvos__ | @freemangordon | 16:13 |
| uvos__ | Guest224: but you can also just download the packages and place them on the sdcard | 16:13 |
| freemangordon | I think most of the data in eeprom is some slope parameters | 16:13 |
| uvos__ | and then install via dpkg -i | 16:13 |
| freemangordon | but I think we have more than one | 16:13 |
| freemangordon | most-probably for different temperatures | 16:14 |
| uvos__ | we must have a 2 tabel really | 16:14 |
| uvos__ | one dimention for voltage and one for current | 16:14 |
| uvos__ | otherwise it makes lttle sense | 16:14 |
| uvos__ | *2d table | 16:14 |
| freemangordon | it makes sense to change charging current based on temp | 16:14 |
| freemangordon | otherwise it may overheat | 16:15 |
| uvos__ | sure | 16:15 |
| freemangordon | not sure if this data is in eeprom though | 16:15 |
| uvos__ | but i mean the state of charge table | 16:15 |
| uvos__ | that must have 2 dimentions (effectively) | 16:15 |
| freemangordon | I think I see voltages only | 16:15 |
| freemangordon | but... | 16:15 |
| freemangordon | not 100% sure | 16:15 |
| freemangordon | anyway, I think we have enough formulas | 16:16 |
| uvos__ | hmm no way they gain the accuracy you can observe on android from just voltage | 16:16 |
| uvos__ | it varys way to mutch with discarge current | 16:16 |
| freemangordon | accuracy for what is the charge? or what? | 16:16 |
| uvos__ | right | 16:16 |
| freemangordon | how do you know it is accurate? | 16:16 |
| uvos__ | by using it for a long time | 16:17 |
| freemangordon | ok | 16:17 |
| uvos__ | i used d4 as my primary device for several years around 2013-2016 or so | 16:17 |
| freemangordon | ok | 16:17 |
| freemangordon | got it :) | 16:17 |
| freemangordon | so, who is going to send tha patches? | 16:17 |
| freemangordon | *the | 16:17 |
| uvos__ | what patches, you re'ed the formulars | 16:18 |
| Guest224 | uvos: good idea. But I feeling stupid..where is .deb packages I tried from find from here http://maedevu.maemo.org/leste/dists/ | 16:18 |
| uvos__ | now we need to implement them in cpcap-battery | 16:18 |
| freemangordon | well, those formulas converted to patches | 16:18 |
| freemangordon | sure | 16:18 |
| uvos__ | freemangordon: i can do it if you like, so can you | 16:18 |
| uvos__ | ie dont care really | 16:18 |
| freemangordon | me neither | 16:18 |
| uvos__ | for this to be usefull imo | 16:19 |
| uvos__ | we first need figureout the chipid situation | 16:19 |
| uvos__ | i gues its 89-500 allways maybe? | 16:19 |
| freemangordon | yes | 16:19 |
| freemangordon | this is what blob checks | 16:19 |
| uvos__ | oh ok | 16:19 |
| freemangordon | along with those 2 strings ^^^ | 16:19 |
| uvos__ | dident see that if you mentioned it | 16:19 |
| freemangordon | "CORP" | 16:19 |
| freemangordon | and "MOTOROLA..." | 16:19 |
| uvos__ | ok well then its ready to be implemented indeed | 16:20 |
| freemangordon | right | 16:20 |
| freemangordon | going on a ride, bbl | 16:20 |
| uvos__ | Guest224: https://maedevu.maemo.org/pkgweb/ | 16:21 |
| uvos__ | https://maedevu.maemo.org/pkgweb/beowulf/main/binary-amd64/hildon-connectivity-mobile.html | 16:21 |
| uvos__ | just click through the depends and download them all | 16:21 |
| uvos__ | Guest224: another alternative is a usb-otg cable and a usb ethernet adapter | 16:22 |
| Guest224 | thanks | 16:22 |
| uvos__ | youll have to setup the network by hand in comandline | 16:22 |
| uvos__ | but then it should work fine | 16:22 |
| uvos__ | assuming d4/bionic/pp here | 16:22 |
| uvos__ | n900 cant do otg | 16:23 |
| Guest224 | Is pp with dockin-bar ethernet supported yet? | 16:23 |
| uvos__ | no idea | 16:24 |
| Wizzup | uvos__: I think so @ buying ones that got removed, kinda lame ;) | 16:46 |
| Wizzup | freemangordon: I can send it, how do I read it? | 16:47 |
| Wizzup | (just woke up, brb) | 16:47 |
| uvos__ | Wizzup: i can live with just one btw | 16:48 |
| uvos__ | if you are short | 16:48 |
| uvos__ | since i have just one d4 now anyhow | 16:48 |
| Wizzup | just short on time atm, not short on devices | 16:48 |
| uvos__ | (i want one to modify) | 16:48 |
| uvos__ | i mean doors | 16:48 |
| Wizzup | ah | 16:48 |
| Wizzup | right | 16:48 |
| Wizzup | I need coffee :p | 16:48 |
| uvos__ | what time zone are you in, Pacific? | 16:50 |
| freemangordon | Wizzup: umm, send what? | 17:01 |
| Wizzup | freemangordon: you wanted me to share some battery data | 17:20 |
| freemangordon | ah, no | 17:28 |
| freemangordon | we already figured it out | 17:29 |
| Wizzup | ok | 17:29 |
| freemangordon | like, I am writing patches to cpcap-battery to process eeprom data | 17:29 |
| Wizzup | I have a lot of d4's that I am bringing back | 17:29 |
| Wizzup | like, it has to be ~40 | 17:29 |
| freemangordon | :) | 17:29 |
| Wizzup | freemangordon: ok, cool | 17:29 |
| Wizzup | freemangordon: I wonder if we also ought to have a patch to charge the battery a bit more than 4.2 for folks who want it | 17:30 |
| freemangordon | I think we shall increase that | 17:30 |
| freemangordon | like, if battery is for 4.35, so be it | 17:30 |
| Wizzup | right, but we thought it maybe caused them to balloon quickly | 17:30 |
| Wizzup | this happened two at least two android phones I had that I kept on charger at all times | 17:31 |
| freemangordon | d4 phones? | 17:31 |
| Wizzup | yes | 17:31 |
| freemangordon | why keep them on charger though? | 17:31 |
| freemangordon | also, charger driver shall be smart enough and have some hysteresis, like, it should not restart charging before the voltage has dropped by lets say 2 or 3 percent | 17:32 |
| freemangordon | but in general, I would trust what manufacturers say | 17:33 |
| freemangordon | so, if they say 4.35, then... | 17:33 |
| freemangordon | Wizzup: also, were those batteries new? | 17:33 |
| freemangordon | I guess no | 17:33 |
| freemangordon | uvos__: did you check idle power usage with hdq enabled? | 17:39 |
| Wizzup | freemangordon: no, probably not new | 17:54 |
| tmlind | freemangordon: yeah d4 batteries kept on charger at higher than 4.2v will ballon after some weeks or months, no issues at 4.2v afaik | 18:03 |
| tmlind | you can already tweak the charge voltage via sysfs if you really want, i would not recommend it | 18:04 |
| freemangordon | I don't see a reason why anyone would want to keep the device on charger | 18:11 |
| freemangordon | but ok | 18:11 |
| freemangordon | though, I think default should be what battery is made for | 18:12 |
| tmlind | the battery was not made to last it seems | 18:31 |
| freemangordon | hmm | 18:34 |
| freemangordon | well, I guess we can at least increase a bit, like to 4.25, no? | 18:34 |
| uvos | im terribly convinced that there is anything wrong per say with charging eb41 to 4.35V | 18:34 |
| uvos | but hvlipos have lower cycle lifetime | 18:34 |
| uvos | which is why they are all very degraded by now | 18:35 |
| uvos | only charging them to 4.2v helps emensely | 18:35 |
| freemangordon | afaik even 'new' are degrated | 18:35 |
| uvos | sure yes | 18:35 |
| uvos | 10 years is old even for purley calendary ageing | 18:35 |
| ashley | ballooning batteries are such "fun"...I just had to throw out a brand new, unused Nokia XL battery a few weeks ago because of how swollen it had gotten, all on its own :-/ | 18:37 |
| uvos | shelf life of lipos is about 5 years so yes | 18:37 |
| freemangordon | so, I am not convinced reducing the charge voltage help much, but ok :) | 18:39 |
| uvos | all d4 batteries are degraded | 18:39 |
| uvos | degraded batteries tend to be most unsafe at full charge | 18:39 |
| uvos | it makes some sense to reduce the voltage | 18:40 |
| uvos | not that i agree we should | 18:40 |
| freemangordon | btw, why do we report li-ion and not li-po for genuine batteries? | 18:44 |
| uvos | a lipo is a type of lion | 18:44 |
| uvos | im not sure lipo is valid for this field | 18:45 |
| uvos | i dident set this either | 18:45 |
| freemangordon | tmlind: uvos: in the meanwhile https://pastebin.com/fwxVyf9f :) | 18:46 |
| sicelo | does anything in userspace actually use VOLTAGE_MIN_DESIGN? would be nice if upower did, i guess | 18:49 |
| sicelo | nice work there freemangordon | 18:49 |
| tmlind | sounds like a driver parsing eeprom patch is coming? :) | 18:57 |
| freemangordon | yes | 18:58 |
| tmlind | cool | 18:58 |
| * freemangordon is looking for volunteers to upstream 3 cpcap patches | 19:02 | |
| freemangordon | :) | 19:02 |
| freemangordon | hmm, why github takes so long to show a new branch? | 19:02 |
| freemangordon | oh, I pushed to wrong repo, again :( | 19:04 |
| Wizzup | uvos: I received 5 mugen batteries for the droid3, so that's nice | 19:06 |
| Wizzup | with battery doors | 19:06 |
| freemangordon | tmlind: https://github.com/maemo-leste/droid4-linux/commits/maemo-5.18.y-cpcap | 19:06 |
| tmlind | freemangordon: looks good to me, please check the default charge current stays the same to be sure.. | 19:12 |
| freemangordon | how? | 19:13 |
| freemangordon | tmlind: ^^^ | 19:30 |
| * buZz excited to try this new cpcap :) | 19:47 | |
| tmlind | freemangordon: hmm i guess that patch does not affect the default charge voltag | 20:30 |
| freemangordon | no | 20:31 |
| freemangordon | just design voltage | 20:31 |
| freemangordon | tmlind: please, what is the proper git command to send series? "git send-email -3 --subject="[PATCH 0/3] power: supply: cpcap-battery improvements" --compose"? | 20:32 |
| tmlind | freemangordon: i first use git format patch to /tmp/outgoing, then to send git send-email --confirm=always --suppress-cc=self --cc-cmd=~/bin/my-git-get-cover-cc --to "..." --cc "..." | 20:42 |
| tmlind | that git-get-cover-cc is some script that adds folks to cc | 20:43 |
| tmlind | not sure if that's even needed any longer to add folks to cc | 20:44 |
| freemangordon | ok, thanks | 20:45 |
| tmlind | the get cover i have is from some mailing list thread i can't find and contains: | 20:47 |
| tmlind | grep -e "Cc: " -e "Reported-by: " -e "Reviewed-by: " \ | 20:47 |
| tmlind | -e "Acked-by: " -e "Suggested-by: " -e "Signed-off-by: " /tmp/outgoing/* | \ | 20:47 |
| tmlind | cut -d':' -f3 | sort | uniq | xargs -n1 -i echo "{}" | 20:47 |
| tmlind | probably git already has that as some option available by now | 20:47 |
| freemangordon | well, that's too complicated for me :) | 20:47 |
| freemangordon | so I'll send whatever I can and will see | 20:47 |
| sicelo | mailing a patch is more complicated than producing the patch itself, heh | 20:50 |
| freemangordon | well, I want to send series with cover letter | 20:50 |
| * sicelo is sitting on some N900 patches for this same reason | 20:50 | |
| sicelo | dts patches | 20:50 |
| tmlind | freemangordon: you use git-format-patch --cover-letter to output the patches to /tmp/outgoing first, right? | 20:51 |
| freemangordon | no, I use the command ^^^ | 20:51 |
| freemangordon | "git send-email -3 --subject="[PATCH 0/3] power: supply: cpcap-battery improvements" --compose" | 20:52 |
| freemangordon | I sent to myself snd it looks ok | 20:52 |
| tmlind | with git send-email probably option --cc-cover does what i use --cc-cmd for.. | 20:52 |
| tmlind | hehe careful with the test emails, they easily get sent to the lists :) | 20:53 |
| freemangordon | sent :) | 20:54 |
| freemangordon | no, I am adding --to and --cc by hand | 20:54 |
| tmlind | ok | 20:54 |
| freemangordon | you're in cc | 20:55 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!