#qi-hardware IRC log for Thursday, 2014-01-02

nicksydneynot sure about brightness but i've read the OSRAM LED are available that draw around 5mA 00:00
nicksydneysomething like this .. http://www.osa-opto.com/low-current_1206.html00:01
Action: whitequark nods00:01
nicksydneyi think it would be useful to provide feedback mechanism to the user to indicate low power maybe threshold if the battery is reaching 5-10% state00:05
whitequarkand charging via USB00:07
nicksydneywpwrak_: the MCU that you are using does it come in QFP packaging...was thinking of resizing the schematic and using QFP MCU to make it easier for me to solder :)00:08
DocScrutinizer05shoot 3 rounds in the air when you run out of ammunition00:12
wpwrak_nicksydney: we can try to get out an indication. not sure yet about how clearly we can detect a battery nearing the end of its life.00:13
DocScrutinizer05btw even sub-1ms flash is still easily visible00:14
wpwrak_nicksydney: you can get it in LQFP 64 or LQFP 80 (as opposed to QFN 48)00:15
DocScrutinizer05there are receptors in human eye that act strictly edge triggered00:15
wpwrak_DocScrutinizer05: (< 1 ms) visible, yes. but very weak. (i happen to have it flash for about 1-2 ms now, to mark progress for power measurements)00:16
wpwrak_(3 rounds) yeah ;-)00:16
DocScrutinizer05wpwrak_: of course you want to operate LED on high current when going ultrashort pulse00:17
DocScrutinizer05not at the usual <1mA you regularly use for low power LEDs in continuous mode00:18
DocScrutinizer05rather like 20 or even 50mA, for 0.5mS00:19
DocScrutinizer05bq24150 on N900 sgnals fault condition with a 128uS pulse on a pin which happens to ben hooked up to red/green LED of RGB indicator LED00:23
DocScrutinizer05the current is limited to continuous low power mode00:24
DocScrutinizer05i.e. <<5mA00:24
DocScrutinizer05the pulse is visible, not good but visible00:24
wpwrak_(high current) naw, peaks are bad. they mean that i would have to install huge caps just to feed the LED.00:24
DocScrutinizer05let's see00:25
DocScrutinizer05...00:25
DocScrutinizer0510uF?00:26
DocScrutinizer05make that 47uF and you're quite fine00:28
DocScrutinizer05http://it.mouser.com/Passive-Components/Capacitors/_/N-5g7r?P=1z0wrk9Z1z0x2u8Z1z0x24tZ1z0jvngZ1z0x6g0Z1yzmou600:33
DocScrutinizer05http://it.mouser.com/Passive-Components/Capacitors/Ceramic-Capacitors/_/N-5g8m?P=1z0wrk900:38
wpwrak_hmm, 47 uF may be good for about 0.1 ms, yes00:39
wpwrak_volate drop in the order of 200-300 mV00:40
wpwrak_but .. i still don't quite see the use case00:40
wpwrak_when the display is off, the device is in very low current mode anyway. so you're not likely to drain the battery there (and if you do, then the anelok is probably stowed away somewhere and you wouldn't see it blinking anyway)00:41
wpwrak_if you're using the device, it could just display a low battery indicator00:42
wpwrak_and if the battery is flat for good, then it just won't turn on :)00:42
whitequarkyes, that's true00:43
DocScrutinizer05yep00:45
DocScrutinizer05though http://product.tdk.com/capacitor/mlcc/en/documents/C3216X5R1E476M160AC.pdf driven correctly can also make nice klicking sounds ;-D00:46
DocScrutinizer05generally you want to use PWM for all sorts of brightness modulation00:50
DocScrutinizer05and a fat capacitor in parallel to your CR2032 will definitely help keep your digital circuitry operating stable00:51
DocScrutinizer05check above datasheet for incredible reduction of effective capacity with charging voltage00:53
DocScrutinizer05at 3.6V DC Bias you're alread ~50%00:53
DocScrutinizer05calling those critters capacitors is really feeling like oversimplifying matters00:55
DocScrutinizer05s/matters/things/00:56
DocScrutinizer05watch weird Z and Capacity vs freq diagrams00:57
DocScrutinizer05while this one is only -20% at 3.6V01:01
DocScrutinizer05http://psearch.murata.com/capacitor/product/GRM32ER71A476ME15%23.pdf01:01
wpwrak_wow. really suckish.01:11
wpwrak_is the difference because of X5R vs. X7R ? or is it an unrelated property ?01:13
nicksydneyDocScrutinizer05: "fat capacitor in parallel to your CR2032 will definitely help keep your digital circuitry operating stable" ... was googling about 'capacitor as voltage stabilizer' and found this .. http://electronics.stackexchange.com/questions/83892/how-to-select-an-appropriate-capacitor-for-input-voltage-stabilization is this the same what you mean as stabilizer ?01:20
nicksydneyDocScrutinizer05: this also "http://product.tdk.com/capacitor/mlcc/en/documents/C3216X5R1E476M160AC.pdf driven correctly can also make nice klicking sounds " ....  is this the same kind of clicking noise that i often hear when turning on a big amplifier product like one of those consumer based amp ?01:21
nicksydneywpwrak_: aren't these the same as your QFN http://www.digikey.com.au/product-search/en?pv16=1225&pv16=1005&pv16=1120&pv16=1109&FV=4002c9%2C4003ed%2C400455%2C400457%2C40045d%2C400460%2C4004c9%2C7e800c6&k=cortex+m0+freescale&mnonly=0&newproducts=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid=0&pageSize=500 ?01:26
wpwrak_that's a lot of different chips, all in QFP packages01:30
wpwrak_if you want "similar", get an MKL24, MKL25, or MKL2601:30
wpwrak_(resend) if you want "similar", get an MKL24, MKL25, or MKL2601:31
nicksydneywpwrak_: MKL24/25/26 you mean the dev board right ?01:32
wpwrak_i mean the chip for the board you want to make01:33
nicksydneyoahh ok you are referring to the part number on the 4th column from that lnk01:34
wpwrak_yeah. i cuold also have said "KL24", etc., but digi-key won't let you search for that :) and "KL2" may be a bit too generic01:36
nicksydneythought if there is 48 QFP it will have the same pinout as the QFN but now it's 64QFP the pinout will be different 01:36
wpwrak_well, actually no. kl2 is the same as { KL24, KL25, KL26 }01:36
wpwrak_yes, of course01:37
wpwrak_the data sheet is your friend :)01:37
nicksydneythought can get away without reading the dataasheet :)01:37
wpwrak_and there is no 48-LQFP for these chips. so your pinout WILL be different ;-)01:37
nicksydneythe painful part of ordering from digikey is the shipping ....OR shop $200 and you get free shipping01:41
nicksydneyeven Aliexpress or ebay does not have those MCU...seems like freescale is restricting sales01:41
kristianpaulzrafa_: hey02:25
kristianpaulzrafa_: I'm good. You?02:25
wpwrak_ah, thunderstorm arriving03:13
nicksydneywpwrak_: uh oh03:17
apeletegood morning08:48
wpwrak_yeah, nice. now that's what i call rain :)09:20
DocScrutinizer05freescale ARM M4 eval board 10 bucks: http://www.digikey.de/product-search/de?lang=de&site=de&keywords=FRDM-KE02Z&WT.z_slp_buy=Freescale_Freedom12:05
DocScrutinizer05http://www.digikey.de/product-highlights/de/de/freescale-freedom-development-platform/385112:06
DocScrutinizer05with USB-OTG and LCD: 12 bucks (ARM M0+ though, whatever that is)  http://www.digikey.de/product-search/de?lang=de&site=de&keywords=FRDM-KL46Z&WT.z_slp_buy=Freescale_Freedom12:09
DocScrutinizer05wpwrak_:  http://cache.freescale.com/files/32bit/doc/user_guide/FRDMKL25ZUM.zip powered by a CR2032 ;-)12:18
DocScrutinizer05just for an irrelevant sidenote12:18
DocScrutinizer05http://wstaw.org/m/2014/01/02/plasma-desktopGs3389.png12:22
DocScrutinizer05not long ago my mood would've been like "buy it! for that lil bit of money it's a mistake to not buy it, even when I have NFC what for"12:24
nicksydneyDocScrutinizer05: so now no mood ? :)12:47
DocScrutinizer05no, I try to stop that devastating habit12:47
nicksydneyDocScrutinizer05: http://wstaw.org/m/2014/01/02/plasma-desktopGs3389.png --> i have the same but it won't work with Linux need to flash the bootloader inside WI$ 12:47
nicksydneywpwrak_ has the same as that one too :)12:48
DocScrutinizer05heh :-)12:48
nicksydneyDocScrutinizer05 / wpwrak_ : this is interesting http://www.digikey.com/product-detail/en/XK-STK-A8DEV/880-1066-ND/448571012:51
nicksydneyocta core cortex M :)12:52
nicksydney(Eight 32bit processors  seven xCORE logical cores supporting DSP instructions, and an ARM Cortex-M3 processor )12:52
DocScrutinizer05https://www.xmos.com/en/download/public/Extending-startKIT-using-the-PCIe-slot-and-sliceCARDs(0.9).pdf  hah!12:58
nicksydneythat extension board allows several peripherals...WiFi, ethernet, etc13:01
nicksydneyLCD - 480 x 272 full color display .. 40-pin ZIF connector with ribbon cable to the display .... Resistive touch screen with 2 wire interface to xCORE ... yummy13:02
nicksydneyperhaps another mini tablet :)13:02
DocScrutinizer05whitequark: https://shop.klingler.net/product-details.en.php?productid=1001213:18
wpwrak_DocScrutinizer05: (cr2032) yes, i know :) and the board is nice. and pleasantly cheap. their flashing system is a little unusual but works quite well. the main problem seems to be the firmware - older versions didn't like linux.13:25
wpwrak_nicksydney: btw, did you manage to update it with windows ? after that, it may just work13:25
nicksydneywpwrak_: not yet need to visit a friend place to use a win machine...no windows at home 13:27
nicksydneywpwrak_: look what i found http://downloads.qi-hardware.com/people/werner/fisl13.pdf13:32
nicksydney:)13:32
nicksydneywas googling for "filled zone kicad" and yours came out :)13:32
wpwrak_(xcore) cute :) not quite what anelok needs but cute :)13:32
wpwrak_50 ml. wow !13:34
nicksydneylearning the difference between copper pour and keepout area...this is interesting13:45
nicksydneykicad 13:45
larscapelete: your MIPS patches have been queued for 3.14 http://git.linux-mips.org/?p=john/linux-john.git;a=shortlog;h=refs/heads/mips-next-3.1414:16
pcercueiparty!14:16
larscdefinitely ;)14:17
apeletew00t, it's christmas all over again \o/14:25
apeletelarsc pcercuei: thanks for the news and the cheering14:26
apeletelarsc: will soon resume working on the nanonote, will let you know when I start studying the MMC driver14:32
apelete(been busy with some adminsys work during the holidays, launched a blog among other things -> http://apelete.seketeli.net)14:33
apeletes/adminsys/sysadmin/14:34
DocScrutinizer05wpwrak_: btw this stuff might be interesting for you as well https://shop.klingler.net/product-details.en.php?productid=1001215:27
DocScrutinizer05wpwrak_: plus sth like http://www.conrad.de/ce/de/product/398434/?productname=AEG-Edelstahl-Fondue-Fritteuse-FFR-5551-Edelstahl-1200-W-52055115:28
DocScrutinizer05(needs a tad of thermostat readjustment)15:29
DocScrutinizer05(but I'd anyway place the thermo-sensor at *top* of the whole thing, making for a convenient automatic adjustment of vapor level15:30
DocScrutinizer05I guess the lid could use a huge heatsink so the HT230 goo can condensate and cycle instead of escape the can and vanish15:32
wpwrak_heh :) yeah, could be interesting at some point in the future. for now, i've made my peace with manual soldering. one day, i plan to be rich, buy me a bunch of CNC mills, then mod them to become pick and place machines and such. that way, such bulk-soldering techniques will become very interesting.15:34
DocScrutinizer05well, this ~60buck total VP soldering setup isn't exctly for bulk production ;-)15:39
DocScrutinizer05rather a convenient way to solder stuff you impossibly can solder any other way at home, like huge BGA chips etc15:40
wpwrak_yeah, i've still been able to avoid BGA :) but i know that luck will run out eventually ...15:41
DocScrutinizer05with the two items above, plus some good solder paste, even noobs should be able to create almost perfect SMD PCBs15:41
DocScrutinizer05I'd probably prefer that method over handsoldering even for extremely complex stuff like the android with glowing eyes ;-)15:42
DocScrutinizer05VP is *awesome*, judging by the physical basics of the whole process15:44
DocScrutinizer05no overheading, no problems to gheat up huge components and/or groundplanes15:45
DocScrutinizer05overheateing even15:45
DocScrutinizer05damn, typing sucks15:45
DocScrutinizer05time for coffee and a break15:46
DocScrutinizer05whitequark anyway will be pleased to learn there are 50ml bottles easily available15:46
DocScrutinizer05courtesy Nikolaus, who OTOH denied any option to get that stuff from fab, since they would rather see us getting our soldering done by them15:48
whitequarkDocScrutinizer05: OOOOO17:08
whitequarkthat is awesome17:09
wpwrak_DocScrutinizer05: well, for the volume you'll need, you may as well start developing relations with good fabs. find their flaws before you have 1000+ units in the pipeline :)17:11
DocScrutinizer05wpwrak_: sorry, you lost me17:15
DocScrutinizer05I think you don't know about GolDeliCo relations to fabs and status of evaluation of the yield and flaws17:17
DocScrutinizer05do you really think I whould've dared to start Neo900 project without GolDeliCo already having a decent fab for complete production, that we know can really handle that task?17:19
DocScrutinizer05wel, "complete" as far as PCB manuf goes. I think they don't do housing/mech17:20
DocScrutinizer05what do you think who soldered the modem onto the pcb:  http://neo900.org/news-0005-phase-i-results-ohsw-showcase17:23
wpwrak_ah, okay. thought you may still need to make that choice. i mean, in germany it shouldn't be that hard to find a fab run by decent engineers :)17:23
wpwrak_i think in china they learn this kind of soldering in kindergarten :)17:24
DocScrutinizer05re China, I'm pleased to state that our fab is magnitudes better than what FIQ offered to OM17:34
DocScrutinizer05I think in Chinal they learn completely different things in Kindergarden. E:G. to ignore all manuals since they are too complex anyway. Rather throw sufficient try&error effort on anything you do, and learn to smile at bitching customers17:36
DocScrutinizer05whitequark: yeah :-)17:40
whitequark... wow http://i.imgur.com/IUiSaDz.gif17:58
JewFro297...wat18:00
DocScrutinizer05whitequark: HIHI18:42
DocScrutinizer05whitequark: I probably had taken the whole contraption with me ;-)18:43
DocScrutinizer05the keyboard looks pathetic18:43
whitequarkthe mill has arrived to the remailer!19:36
whitequarkfedex is amazing. not only they worked through what is considered holidays, but also they were able to accurately predict the arrival time19:37
wpwrak_whitequark: the sequence makes the clip even nicer: looks as if there was a fake front on top of another fake front ;-)21:05
wpwrak_(fedex) they're great. if it can be done at all, they do it. i've had all the others drop the ball on occasion but never fedex.21:07
DocScrutinizer05anybody got a pointer for me to decode data from /dev/input/ts ?21:26
DocScrutinizer05http://privatepaste.com/aaf04ba9c921:26
DocScrutinizer05I gather it must follow some sort of standard for input devices21:30
nicksydneyDocScrutinizer05: this one is saying http://neo900.org/photos/neo900/pcb-a0/DSC01787.JPG ... I show you mine if you show me yours :)21:31
DocScrutinizer05hm?21:32
larscDocScrutinizer05: http://kerneldox.com/kdox-linux/d5/df7/structinput__event.html21:55
DocScrutinizer05larsc: many thanks21:56
DocScrutinizer05hah, struct timeval time I gathered already :-D21:57
DocScrutinizer05I bet it#s 32bit size21:57
larscon legacy 32bit systems at least21:57
DocScrutinizer05on mine :-)21:58
DocScrutinizer05God, that's conveniently self-referencing when you want to find out about enum type22:03
DocScrutinizer05\o/ 00172  * Event types22:06
DocScrutinizer05honestly C is about as poorly selfdocumenting a language as it gets22:07
whitequarkthe only good thing about C is that you have to deliberately go out of your way to make a language worse22:08
DocScrutinizer05indeed22:10
DocScrutinizer05\o/ 22:10
DocScrutinizer0500754  * Relative axes22:10
DocScrutinizer05now I *guess* this will show up *somewhere* in struct input_event {...}22:10
whitequarkDocScrutinizer05: http://talk.maemo.org/archive/index.php/t-68929.html22:11
DocScrutinizer05whitequark: great! Nicolai to the rescue. And python of course22:14
DocScrutinizer05though we again see an excellent example how python format depending syntax SUCKS!!!22:15
DocScrutinizer05let's see22:17
DocScrutinizer05aaah MUCH better http://talk.maemo.org/showthread.php?t=6892922:18
DocScrutinizer05I still wonder where from he got his event=0 and event=3 stuff22:19
DocScrutinizer05err type, but nevertheless22:19
larscprobably from the documentation ;)22:20
larsc3 is absolute position, 0 is sync22:21
DocScrutinizer050 is sync???22:26
DocScrutinizer05what's "sync"?22:26
DocScrutinizer05some sort of ETX?22:29
larsctells you that the event is complete22:29
DocScrutinizer05:nod:22:30
DocScrutinizer05that documentation you refer to is afaik usually decently hidden somewhere in /usr/share/docs/some-really-weird-filename22:31
DocScrutinizer05only if you're lucky and your system didn't suffer a purgedoc22:31
larschttps://www.kernel.org/doc/Documentation/input/22:33
DocScrutinizer05:-)))22:33
DocScrutinizer05event-codes.txt \o/22:33
DocScrutinizer05larsc: another noob's silly question: now that I have opened the /dev/input/ts for read, and I guess X11 did as well, will both processes see all events or are they 'randomly' handed out to only one of both, so both processes see statistically only every 2nd event?22:38
DocScrutinizer05I'd guess they should show up at all read()s in sync, no?22:40
DocScrutinizer05for a 'normal' file this situation is pretty defined, each read() has its own read pointer22:41
DocScrutinizer05for concurrent writes it already starts getting funny22:41
whitequarkDocScrutinizer05: naw it's duplicated for each process22:41
whitequarksame for /dev/input/eventX22:41
DocScrutinizer05thought as much. Thanks22:41
DocScrutinizer05will the driver's _get() function get called once for each event or once for each process reading an event?22:43
DocScrutinizer05IOW is the duplication done in driver's code or in the lib that handles this22:44
whitequarkfor that you'll need to look at the code...22:44
DocScrutinizer05:-)22:44
larscthe dupliucation is done in the kernels core input code22:45
DocScrutinizer05thanks to your first link you shared, I'm almost there already22:45
DocScrutinizer05thanks! stopping to annoy and pester you ;-)22:46
DocScrutinizer05epic!:22:58
DocScrutinizer05EV_PWR events are a special type of event used specifically for power22:58
DocScrutinizer05mangement. Its usage is not well defined. To be addressed later.22:58
whitequarkwpwrak_: you'll like reddit.com/r/nononono23:30
whitequarkit's a subreddit about expensive machinery breaking23:30
whitequarke.g. http://www.liveleak.com/view?i=0fc_138848986623:31
wpwrak_sweet ;-) i'm not quite sure what i saw but that machine crossing over while spilling white-hot steel looked rather interesting :)23:43
whitequarkapparently it's some kind of malfunction, for which the damage mitigation consists of dumping a thin layer of steel on the floor23:44
wpwrak_what, not this ? http://totalfratmove.s3.amazonaws.com/wp-content/uploads/2012/08/76431abd9675b83ba3b7b021956ba715408023247.jpg23:48
wpwrak_or at least this ? http://www.pics22.com/does-this-sign-mean-urinate-in-case-of-fire/23:49
whitequarkthe sign means exactly that23:49
whitequarkI'm not really sure in what context it appears though23:49
wpwrak_"hmm, we should put a sign here." "okay, i'll make one when i'm finished with the bottle."  http://static.someecards.com/someecards/usercards/1294089352451_1675164.png23:51
larscsave the planet with one bottle of vodka at a time?23:52
--- Fri Jan 3 201400:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!