IRC log for #neo900 on 20170324

00:41.07*** join/#neo900 infobot (ibot@rikers.org)
00:41.07*** topic/#neo900 is http://neo900.org | CCCAMP15 lightning talks at http://neo900.org/stuff/cccamp15/ - major: http://neo900.org/stuff/cccamp15/ccc2015talk/neo900-wpwrak_CCC2015.webm | conversations are logged to http://infobot.rikers.org/%23neo900/ and http://irclog.whitequark.org/neo900
00:41.07*** mode/#neo900 [+v infobot] by ChanServ
01:00.35*** join/#neo900 herpderphurr (~afwang@c-98-234-221-193.hsd1.ca.comcast.net)
03:00.43*** join/#neo900 ashneo76 (~ashneo76@204.14.154.168)
03:21.08Joerg-Neo900mail sent \o/
03:21.17Joerg-Neo900oops ECHAN
04:57.13*** join/#neo900 knttl (~knttl@dyndsl-091-248-173-080.ewe-ip-backbone.de)
05:53.18*** join/#neo900 DocScrutinizer05 (~saturn@openmoko/engineers/joerg)
05:53.18*** mode/#neo900 [+v DocScrutinizer05] by ChanServ
05:53.29*** join/#neo900 neo900 (~office@neo900/coreteam/joerg)
05:53.29*** mode/#neo900 [+v neo900] by ChanServ
06:36.04*** join/#neo900 arcean (~arcean@131.228.32.166)
07:04.54*** join/#neo900 chomwitt (~chomwitt@ppp-94-67-220-213.home.otenet.gr)
07:38.21*** join/#neo900 jonsger (~Thunderbi@HSI-KBW-046-005-000-096.hsi8.kabel-badenwuerttemberg.de)
08:33.33*** join/#neo900 e (e@freenode/staff/spy.edk)
08:40.41*** join/#neo900 jonwil (~jonwil@27.33.80.219)
08:43.44*** join/#neo900 paulk-collins (~paulk@gagarine.paulk.fr)
09:23.15*** join/#neo900 saper (saper@m.saper.info)
09:23.15*** join/#neo900 saper (saper@wikipedia/saper)
10:02.00*** join/#neo900 kile21 (2ef62c4f@gateway/web/freenode/ip.46.246.44.79)
10:13.13*** join/#neo900 chomwitt (~chomwitt@94.67.220.213)
11:16.32*** join/#neo900 rootman (~rootman@83.164.147.60)
11:46.24*** join/#neo900 cc__ (~ac@2001:910:113f:1:6a05:caff:fe1c:1627)
12:05.17*** join/#neo900 Kabouik (~kabouik@158.38.1.70)
12:21.43*** join/#neo900 Pali (~pali@Maemo/community/contributor/Pali)
12:46.46*** join/#neo900 xes_ (~xes@unaffiliated/xes)
14:23.34*** join/#neo900 Kabouik (~kabouik@158.38.1.70)
15:08.48*** join/#neo900 galiven (~Andrew@50-205-116-131-static.hfc.comcastbusiness.net)
15:41.24*** join/#neo900 ecloud (quassel@nat/qt/x-srwiebxfnjphfhgg)
15:44.42*** join/#neo900 Kabouik (~kabouik@158.38.1.70)
15:48.12*** join/#neo900 wpwrak (~werner@13-145-231-201.fibertel.com.ar)
15:48.12*** mode/#neo900 [+v wpwrak] by ChanServ
15:55.30*** join/#neo900 drrz (~drrrrz@162.216.46.43)
15:57.39*** join/#neo900 wpwrak (~werner@13-145-231-201.fibertel.com.ar)
15:57.39*** mode/#neo900 [+v wpwrak] by ChanServ
16:38.54*** join/#neo900 Pali (~pali@Maemo/community/contributor/Pali)
17:01.40*** join/#neo900 Kabouik (~quassel@32.232.202.84.customer.cdi.no)
18:08.50*** join/#neo900 x29a (~x29a@unaffiliated/x29a)
18:16.05atkMy nokia N900 thinks my new earphones (which have a four band cable thing and an in-line microphone) are a TV out.
18:16.33atkAs in the 3.5mm jack has four metal bands.
18:16.37atknot the usual three.
18:17.11atkI imagine this behaviour can be somehow overridden
18:26.17sixwheeledbeastI imagine a OMTP AHJ issue that would only be "fixed" with a convertor
18:26.24sixwheeledbeast~omtp
18:29.24sixwheeledbeasthttp://www.dx.com/p/universal-3-5mm-omtp-to-ctia-earphone-adapter-black-red-339166
18:29.59sixwheeledbeastsomething like that is basically a 2nd ring to sleeve crossover cable.
18:34.16atkI waited 6 weeks for these earphones, now I will wait 6 weeks for a converter :P
18:34.20atkThe joys of buying from china.
18:36.54*** join/#neo900 chomwitt (~chomwitt@ppp-94-67-220-213.home.otenet.gr)
18:37.02sixwheeledbeastI just use the WH-205 provided...
18:37.49sixwheeledbeastYou may want to check that the headset is CTIA/AHJ with a multimeter or another device first
18:38.49atkI have a multimeter
18:39.31atkI didn't buy these IEMs for the microphone, I bought them for the sound producing bits
18:45.33sixwheeledbeasthttps://en.wikipedia.org/wiki/Phone_connector_(audio)#TRRS_standards
18:46.47atk<PROTECTED>
18:46.50sixwheeledbeastAs you can see you can confirm from the resistance between the connections.
18:47.49sixwheeledbeastawaits Doc to chime in on what a stupid idea having the ground signal on R2 is
18:47.56sixwheeledbeastand I fully agree.
18:48.47sixwheeledbeastN900 is OMTP, most modern stuff is CTIA/AHJ
18:52.12sixwheeledbeastthe reason for the tv out is the N900 will be looking for the mic resistance on R2 to detect if R2 should be mic or cvbs
18:55.37Joerg-Neo900yep, all correct
18:56.51sixwheeledbeastBTW the whitequark irclog cert expires today in case anyone here can do anything about that.
18:57.35Joerg-Neo900whitequark should :-)
18:58.25atkit seems ground is R2
18:58.37Joerg-Neo900AHJ
18:58.42sixwheeledbeast:nod:
18:58.59Joerg-Neo900not compatible to N900
18:59.04*** join/#neo900 b0b0t (~bushido@173-169-126-136.res.bhn.net)
18:59.10*** join/#neo900 pagurus (~user@p4FDC3F3C.dip0.t-ipconnect.de)
18:59.23Joerg-Neo900Neo900 has an automatic switch to use both standards
18:59.26*** join/#neo900 pagurus (~user@p4FDC3F3C.dip0.t-ipconnect.de)
18:59.31atkat least this is the test I can make: R2 to R1 and tip - beeps, sleeve to R1 and tip - NC, sleeve to R2 - NC
18:59.43atkwhen I say beeps, there's a connection
18:59.54Joerg-Neo900yep
19:00.01Joerg-Neo900correct
19:00.07sixwheeledbeastresistance would be a better test than continuity
19:00.11atkand sleeve to R1 or tip - nc
19:00.56*** join/#neo900 pagurus (~user@p4FDC3F3C.dip0.t-ipconnect.de)
19:01.47atkwhen I say NC (no connection) my meter shows over limit
19:02.26*** join/#neo900 pagurus (~user@p4FDC3F3C.dip0.t-ipconnect.de)
19:02.39Joerg-Neo900you can press and hold the button of mic before plugin
19:02.51Joerg-Neo900shorts sleeve and R2
19:02.54sixwheeledbeastI'd say <100ohm over each of the speakers and ~2k for the mic ?
19:03.02Joerg-Neo900yes
19:03.34Joerg-Neo900~ping
19:03.34infobot1 packet transmitted, 1 packet received, 0.0% packet loss
19:03.39Joerg-Neo900~omtp
19:03.51Joerg-Neo900~listvalues omtp
19:03.56*** join/#neo900 pagurus (~user@p4FDC3F3C.dip0.t-ipconnect.de)
19:04.00Joerg-Neo900o.O
19:04.03*** join/#neo900 ravelo (ravelo@neo900/community/supporter/ravelo)
19:04.22Joerg-Neo900~+status
19:04.22infobotSince Fri Mar 24 00:39:06 2017, there have been 0 modifications, 7 questions, 0 dunnos, 0 morons and 2 commands.  I have been awake for 18h 25m 16s this session, and currently reference 119652 factoids.  I'm using about 83232 kB of memory. With 0 active forks. Process time user/system 97.64/1.67 child 0/0
19:05.03*** join/#neo900 xmn (~xman@user-0cdft2n.cable.mindspring.com)
19:05.26*** join/#neo900 pagurus (~user@p4FDC3F3C.dip0.t-ipconnect.de)
19:05.39Joerg-Neo900~OMTP is http://www.omtp.org/OMTP_Local_Connectivity_Wired_Analogue_Audio_v1_0.pdf or https://en.wikipedia.org/wiki/Phone_connector_(audio)#TRRS_standards
19:05.40infobotokay, Joerg-Neo900
19:06.56*** join/#neo900 pagurus (~user@p4FDC3F3C.dip0.t-ipconnect.de)
19:08.47Joerg-Neo900~ahj
19:09.23atkah, I put my meter out of auto-ranging
19:09.25atkthat explains it
19:12.03atk1k6R S R2, 1k6R S R1, 1k6R S T, 18R R2 R1, 18R R2 T, 34R R1 T
19:12.11atkI think it's obvious R2 is ground.
19:12.55atkand you're right, if I hold the mic button as I plug in, N900 recognises as earphones
19:12.59atkwonderful
19:13.25Joerg-Neo900need button pressed for music playback
19:13.41atkhmm, yes, that makes sense
19:13.43Joerg-Neo900as soon as you release it, it fails
19:13.45atkgets some tape
19:14.11sixwheeledbeastthat crossover lead looking a potential option again
19:14.31atkYes, I'll get one from ebay for twice the price but which will get here in two days instead of a month
19:14.53atkAlthough tape works too
19:16.14Joerg-Neo900use some elastic object (syrofoam ball or the like) under tape, to have constant elastic pressure on button
19:16.28Joerg-Neo900styrofoam
19:16.43atkthat would slowly flatten itself
19:16.57atkI think some cardboard would work though
19:16.58Joerg-Neo900right, styrofoam suboptimal
19:16.59atkthin strip
19:17.40Joerg-Neo900rather paper, folded. It's more elastic than cardboard
19:18.01sixwheeledbeastelastic band?
19:18.17Joerg-Neo900or you use a rubber ring instead sticky tape, then a rigid object will do
19:20.45atkAlternatively I could get some aluminium foil and carefully short R1 and S with it, use a tiny bit if tape to hold it down
19:20.57atkexplores idea further
19:22.09Joerg-Neo900forget that, there's no space left between receptacle and plug for such contraption
19:22.23atkYe of little faith.
19:22.31atkWe shall see
19:22.46Joerg-Neo900do the math :-)
19:22.57atkthe math would take more time than trying it
19:23.34Joerg-Neo900also I's bet alu foil will break on jack contact springs sliding over it
19:24.04Joerg-Neo900trying it might result in you beeding to disassemble and clean your N900 jack
19:24.09Joerg-Neo900needing*
19:24.35atkworth the risk
19:25.11Joerg-Neo900though chances are low you will be able to insert that alufoil into jack together with the plug. However if you succeed, odds are it will stay nside
19:25.33Joerg-Neo900I strongly discourage you to do that
19:26.01Joerg-Neo900won't work for at least 3 very strong reasons
19:26.23atkall wrong, it won't work because my tape is too inflexible
19:26.31atkand not sticky enough
19:26.31Joerg-Neo900chances 5% * 3% * 10%
19:26.46Joerg-Neo900ooh, sticky tape
19:26.53Joerg-Neo900chances 5% * 3% * 10% * 2%
19:31.00sixwheeledbeastalu foil?... im out
19:31.45atkwell, it worked once, then the tape bunched up and it stopped working, you were right about the tolerances though, I didn't realise they were so tight
19:31.55atkeither way, I got to try it, and tape didn't get stuck inside my N900
19:32.05sixwheeledbeastthey need to be
19:32.09atkshrugs
19:32.27atkI am a "see it for myself" kind of person
19:32.36sixwheeledbeastlol
19:32.57atkWhen the opportunity presents itself, I'll do something stupid to see just how stupid it is
19:33.07sixwheeledbeastwe saw that from a miles away...
19:33.21atkWell yes, but I had faith
19:33.35atkand aluminium foil
19:33.37atkand tape
19:33.57atktime to tape some paper down
19:34.22Joerg-Neo900if you're so wildly determined to fix the issue, open up the microphone+button case, and heat up your soldering iron
19:34.49atkI was determined to find a quick temporary fix while the crossover cable arrives
19:35.01atkthat would be a permanent fix, and for some crazy reason I might want the mic
19:35.26Joerg-Neo900you just need to swap the cables for sleeve and R2
19:35.47atkI doubt I could work on such tiny wires
19:36.07Joerg-Neo900good point
19:36.09atkand the case looks like it was ultrasonically welded
19:36.22Joerg-Neo900quite possible, yes
19:42.27atktape and paper works
19:42.37atkactually, paper, blutack, paper and then tape
19:43.01atkprobably about as difficult to assemble as those Package on Package chips :P
19:45.42sixwheeledbeastI can't remember if you can do some temporary magic, editing devices files and making them read only. I have done it for cam button and prox etc
19:47.11Joerg-Neo900sixwheeledbeast: that won't fix the GND of audio circuitry connected to sleeve of jack, in N900
19:47.56sixwheeledbeastoh your right I didn't think about the connection...
19:48.08Joerg-Neo900headphone signal from amp-out is tip-sleeve and R1-sleeve no matter what you do on software
19:48.45sixwheeledbeast:nod:
19:48.55Joerg-Neo900and R2 is positive bias to sleeve, or disconnected (or TV-out)
19:49.56Joerg-Neo900for AHJ you need positive bias on sleeve wrt R2 (GND) for the mic to work
19:50.14Joerg-Neo900iirc
19:51.08Joerg-Neo900...and according to N900 detecting the R2-sleeve as open-circuit
19:51.18Joerg-Neo900not as ~2kR
19:51.29Joerg-Neo900mic has wrong polarity
19:52.02Joerg-Neo900amazing this isn't considered a possible cause for damage
19:52.17atkwhat is a possible cause for damage?
19:52.45Joerg-Neo900mic + integrated amp powered wrong polarity
19:52.53atkI see
19:53.03atkare mics susceptible to damage if they're reverse biased?
19:53.10Joerg-Neo900the amp usually is a FET, so low risk from that
19:53.48Joerg-Neo900https://en.wikipedia.org/wiki/Electret_microphone
19:54.25atksoon I will be undertaking the highly complex task of designing and making a regulated linear variable bench power supply for myself
19:55.37atkAs all the ones you can buy either cost an arm and two legs or look like they are going to fall apart from just looking at them for too long
19:55.46Joerg-Neo900got a switched mode lab PSU with 40V 10A and USB for programming. Best purchase ever, re PSUs
19:56.23atkAnd I get to spec it out in the way I specifically want
19:56.32Joerg-Neo900I just wish the current limiter had better than 0.1A granularity
19:57.23atkThis one will use some nice 10 turn pots or dual pots, I am hoping to get a bit better than 0.1A and 0.1V
19:57.37Joerg-Neo900sure :-)
19:57.51atkaiming for something 0-30V and 4A (adjustable current limit of course)
19:57.51Joerg-Neo900analog there's no granularity :-D
19:58.05atkIndeed
20:03.25Joerg-Neo900https://www.conrad.de/de/labornetzgeraet-einstellbar-voltcraft-pps-16005-1-36-vdc-0-10-a-360-w-usb-remote-programmierbar-anzahl-ausgaenge-2-x-513914.html
20:03.39atkAnyway, thank you for your assistance gentlemen.
20:03.43atkthe earphones now work
20:03.58Joerg-Neo900:-))
20:03.59atk250EUR
20:04.01atkholy cow
20:04.07atkonly 1 channel
20:04.37atkwhat does the programming let you do anyway?
20:05.05Joerg-Neo900all
20:05.07Joerg-Neo900:-)
20:05.27atk"all"
20:05.34Joerg-Neo900except for arbitrary waveform generator on frequencies >1Hz
20:05.43atkI see
20:06.13atkIf I throw a teensy into this upcoming PSU project and don't burn it, I could probably get even better than 1Hz :D
20:06.43Joerg-Neo900you can do all you also can do on human interface: set voltage and current, read out the meters. And then some on top
20:06.43atkI would need to write the firmware, sure, and that might take 3 years at the rate I'm working on this keyboard project on, sure
20:07.23atkon that note, I should finally implement this UART nonsense into this code and then try to connect to this MCU over serial
20:07.41atkThen I can debug things more easily.
20:07.53atkSo far i've gotten the USB SIE working with just a single LED for debugging :P
20:08.05atkIn retrospect, getting UART working would have made more sense before the USB
20:08.09atkbut that wouldn't have been as fun
20:09.05atkSame reason I'm not using any of that arduino nonsense
20:09.11atktakes all the fun out of programming MCUs
20:09.35atkIf it can't have some completely inexplicable bug which takes 3 weeks to track down, it's not fun enough.
20:09.53atkAlthough arduino is full of its own bugs.
20:10.15atkLike the PWM having an upper limit of 80% duty cycle for some insane reason
20:10.27atk(something around there, I can't remember the exact value)
20:11.34Joerg-Neo900LOL
20:14.34Joerg-Neo900the most nasty instrument it got which absolutely doesn't lend itself to decent control via USB (or even RS232) is my UPS
20:14.48Joerg-Neo900s/it/I/
20:15.42Joerg-Neo900damn HID
20:15.53atkuninterruptible power supply?
20:16.07atkhuman interface device?
20:16.13atkI have to say, the HID spec is a mess
20:16.41atkand the linux kernel was technically implementing a silly little quirk of it incorrectly
20:16.51atkSo I fixed it, now it's scheduled for 4.12
20:17.25atk4.12 HID will finally be slightly more conforming to the spec than it is now
20:17.43Joerg-Neo900yes, yes, and absolutely yes
20:18.38Joerg-Neo900don't ask me how a UPs is any sort of HID really
20:18.57atkthey make good bricks for beating people with
20:19.05atkso it's a kind of Human Interface Device
20:19.14Joerg-Neo900hehehe
20:19.25Joerg-Neo900too heavy, over 15kg
20:19.29atkIn the sense that it is a device which can be used to interface with the face of a human.
20:19.49atkYeah, they are awfully heavy, had to carry one from work to my house once, when I purchased one.
20:19.59atkNot only heavy, but the box it came in was unweildy
20:20.13atkfelt like my arms were going to fall off from the bizarre contorted holding position
20:20.27Joerg-Neo900:nod:
20:23.54Joerg-Neo900to keep you up to date with more on-topic stuff: since ahycka and ceene are busy with their day job and thus only partially available for layout, we're about to find more human resources to speed up the proto_v2 finilazing
20:24.39atkI see
20:25.29Joerg-Neo900a big thank you to ravelo for all the help with that
20:26.31atkIf you need any help with git, or editing a blog post for the website, or some linux related stuff, heck, if it's within my abilities I can even do hardware related stuff or kernel related stuff
20:27.09atk(Mind you, I am no EE though, I am just a hobbyist)
20:27.22Joerg-Neo900git, blog posts, all highly welcome. Git on demand though, just not right now
20:28.19Joerg-Neo900I *hope* soon i'll be glad to get your help with git, to create account for the new layouter
20:29.31Joerg-Neo900actually maybe even two new accounts
20:32.41Joerg-Neo900just a random screenshot of this morning's latest git commit: http://susepaste.org/78856565
20:33.57atkall the bits at the bottom have to be crammed into the N900 shaped board at the top?
20:34.21atkor I presume some go on the UPPER board at the top
20:34.27Joerg-Neo900yes
20:34.55Joerg-Neo900and LOWER we have S1 and S2 surface
20:35.31Joerg-Neo900by math estimation, we have a density ~70%
20:35.42Joerg-Neo900iirc
20:37.43Joerg-Neo900a tad on the crowded side, but still feasible
20:38.56Joerg-Neo900PCB might need more layers and blind vias to cope with such density, which makes it more expensive, but that's negiligble considering the other cost factors of the whole project
20:39.22Joerg-Neo900negligible*
20:42.05Joerg-Neo900http://susepaste.org/82019149 a less cluttered view
20:53.47*** join/#neo900 jonwil (~jonwil@27-33-80-219.tpgi.com.au)
21:00.19Joerg-Neo900(LOWER we have S1 and S2) on UPPER there's S3, S4, where S3 is only partially usable where the components on S2 leave enough headroom to place low profile components on S3
21:01.29Joerg-Neo900also several particularly large components (e.g. all kbd switches, SIM holder) are double, in the batch of spread-out components on bottom and a second time already placed and fixed on PCB
22:30.22enycooh things tappening
22:30.34enycbe good togcet a news article out soon
23:52.18*** join/#neo900 wpwrak (~werner@13-145-231-201.fibertel.com.ar)
23:52.18*** mode/#neo900 [+v wpwrak] by ChanServ

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.